JP3860803B2 - Control device for gas supply servomechanism for ventilator - Google Patents

Description

【００７０】
すなわちオブザーバ５４が有する呼吸器官モデルは、自発呼吸圧力Ｐｍｕｓがゼロとした場合の患者の呼吸器官のモデルである。このモデルでは、演算気道圧力＾Ｐａｗから演算肺胞圧力＾Ｐａｌｖを減算した値は、推定流量＾Ｆに推定気道抵抗＾Ｒを乗算した値と等しい。また推定流量＾Ｆを支援ガス供給開始時刻から積分した値が演算体積＾Ｖと等しい。また演算肺胞圧力＾Ｐａｌｖは、肺の推定エラスタンス＾Ｅに演算体積＾Ｖを乗算した値と等しい。
【００７１】
したがって演算気道圧力＾Ｐａｗを入力値とし、推定流量＾Ｆを出力値とした場合には、オブザーバ５４の伝達関数は、次式によって表わされる。
【００７２】
【数２】

【００７３】
ここで、＾Ｒは、推定気道抵抗を示し、＾Ｅは、肺の推定エラスタンスを示す。また他の式についても、上式に示す記号について同様の意味を表わす。このようなオブザーバ５４が有する呼吸器官モデルは、実施の一例であって、患者の呼吸器官をモデル化した他のモデルであってもよい。
【００７４】
制御量演算手段５３は、偏差演算手段５２によって演算される流量偏差ΔＦに予め設定される係数である流量ゲインＫ_ＦＧを乗算した第１演算値（Ｋ_ＦＧ・ΔＦ）と、支援ガス供給開始時刻から流量偏差ΔＦを積分演算した値に予め設定される係数である体積ゲインＫ_ＶＧを乗算した第２演算値（Ｋ_ＶＧ・ΔＦ／ｓ）とを求め、第１演算値および第２演算値を加算して支援圧力Ｐｖｅｎｔに関連する目標圧力Ｐｉｎを演算する。流量偏差ΔＦを入力値とし、目標圧力Ｐｉｎを出力値とした場合には、制御量演算手段５３の伝達関数は、次式によって表わされる。
【００７５】
【数３】

【００７６】
ここで、Ｋ_ＦＧは、流量ゲインを示し、Ｋ_ＶＧは、体積ゲインを示す。また他の式についても、上式に示す記号について同様の意味を表わす。たとえば流量ゲインＫ_ＦＧは、推定した気道抵抗＾Ｒに予め定める流量増幅ゲインβ_ＦＧを乗算した値に設定され、体積ゲインＫ_ＶＧは、推定した肺のエラスタンス＾Ｅに予め定める体積増幅ゲインを乗算した値に設定される。なお、前記流量増幅ゲインβ_ＦＧと、体積増幅ゲインβ_ＶＧとを同じ値に設定した場合には、それらを単に増幅ゲインβと称する。さらに＾Ｒ＝Ｒでかつ＾Ｅ＝Ｅである場合の増幅ゲインβをＢとする。
【００７７】
上述した流量推定手段５１、偏差演算手段５２、制御量演算手段５３は理解を容易にするために、個別に説明したが、等価変換されて整理されてもよい。また流量推定手段５１、偏差演算手段５２および制御量演算手段５３は、数値演算可能なコンピュータが、予め定める動作プログラムを実行することによって実現されてもよい。
【００７８】
本発明の実施の一形態では、サーボ機構２０の伝達関数は、むだ時間要素を含んでいる。図２には、サーボ機構２０の伝達関数のうち、むだ時間要素を除いた伝達関数Ｇｃ(ｓ)と、むだ時間要素の伝達関数ｅ^−τ・ｓとを個別に図示する。すなわち目標圧力Ｐｉｎを入力値とし、吐出圧力Ｐｖｅｎｔを出力値とした場合には、サーボ機構２０の伝達関数は、次式によって表わされる。
Ｇｃ(ｓ)・ｅ^−τ・ｓ …（６）
【００７９】
ここでＧｃ(ｓ)は、むだ時間要素を除いたサーボ機構２０の伝達関数を示す。またｅは、自然対数を示し、τは、目標圧力Ｐｉｎが与えられてからサーボ機構２０が支援圧力Ｐｖｅｎｔの調整を開始するまでに要するむだ時間を示す。また他の式についても、上式に示す記号について同様の意味を表わす。
【００８０】
また実際の患者の呼吸器管においては、支援圧力Ｐｖｅｎｔの他に自発呼吸圧力Ｐｍｕｓが与えられることが、オブザーバ５４の呼吸器官モデルと異なる。なお、本発明の実施の形態においては、吸気管路での圧力損失が小さいとして、サーボ機構２０の吐出圧力となる支援圧力Ｐｖｅｎｔと、実際の患者の気道圧力Ｐａｗとが等しいとする。
【００８１】
図３は、本発明の全体の系１４における自発呼吸時の換気量Ｖｍｕｓとアシスト呼吸時の換気量Ｖａｓｔとの関係を示すグラフである。支援圧力Ｐｖｅｎｔは、自発呼吸圧力Ｐｍｕｓの時間変化に応じて、自発呼吸圧力Ｐｍｕｓの（１＋Ｂ）倍の増幅率で増幅される。換気量は、支援ガスが肺に流れた体積と等しい。ここで、Ｂは、上述したように＾Ｒ＝Ｒでかつ＾Ｅ＝Ｅ出ある場合の増幅ゲインβを示す。本実施の形態の人工呼吸器１７によるアシスト呼吸時の換気量Ｖａｓｔは、自発呼吸時の換気量Ｖｍｕｓの（１＋Ｂ）倍に増幅される。
【００８２】
患者の状態が呼気期間から吸気期間に切換ると、患者は横隔膜などの呼吸筋を動作させる。これによって自発呼吸時の換気量Ｖｍｕｓおよび自発呼吸圧力Ｐｍｕｓは、時間経過とともに徐々に増大し、あるピーク値Ｐ１に達すると徐々に減少する。そして患者の状態が吸気期間から呼気期間に切換る。
【００８３】
通常、吸気期間において自発呼吸時の患者の換気量Ｖｍｕｓおよび自発呼吸圧力Ｐｍｕｓは、時間に対する波形として、まず穏やかな漸増カーブを描き、次に極大値から急速な減少カーブを描く。ただし患者の状態によって、患者の換気量Ｖｍｕｓおよび自発呼吸圧力Ｐｍｕｓは、大幅に変動しそのピーク値Ｐ１および吸気期間Ｗ１が変動する。
【００８４】
制御装置２１に制御されるサーボ機構２０は、患者の自発呼吸圧力Ｐｍｕｓに予め定める増幅ゲインβに基づいて比例増幅した気道圧力Ｐａｗとなるように、支援圧力Ｐｖｅｎｔで支援ガスを吐出する。たとえば、自発呼吸圧力Ｐｍｕｓのピーク値Ｐ１が小さく、吸気期間Ｗ１が短い場合には、気道圧力Ｐａｗのピーク値Ｐ２が小さく、支援ガスが供給される期間Ｗ２が短くなるように、支援圧力Ｐｖｅｎｔが制御される。同様に、自発呼吸圧力Ｐｍｕｓのピーク値Ｐ１が大きく、吸気期間Ｗ１が長い場合には、気道圧力Ｐａｗのピーク値Ｐ２が大きく、支援ガスが供給される期間Ｗ２が長くなるように、支援圧力Ｐｖｅｎｔが制御される。
【００８５】
以上のように本実施の形態の制御装置２１によれば、流量偏差ΔＦに基づいて、支援圧力Ｐｖｅｎｔを決定する。検出流量Ｆは、患者の自発呼吸圧力Ｐｍｕｓによって変化するが、推定流量＾Ｆは、患者の自発呼吸圧力Ｐｍｕｓの影響を受けない。したがって流量偏差ΔＦは、自発呼吸圧力Ｐｍｕｓの変化を抽出した値となる。すなわち検出が通常困難な自発呼吸圧力Ｐｍｕｓを推測することができ、自発呼吸圧力Ｐｍｕｓを外乱とみなした場合の外乱オブザーバとして構成することができる。
【００８６】
このように、自発呼吸圧力Ｐｍｕｓに関係する流量偏差ΔＦに応じて目標圧力Ｐｉｎを演算することによって、自発呼吸圧力Ｐｍｕｓにほぼリアルタイムで追従する支援圧力Ｐｖｅｎｔで、支援ガスを患者に供給することができる。
【００８７】
図４は、図２の本発明の全体の系１４を等価変換して整理して示すブロック線図である。本発明の全体の系１４は、自発呼吸圧力Ｐｍｕｓを入力値とし、自発呼吸圧力Ｐｍｕｓと支援圧力Ｐｖｅｎｔとの加算値を出力値とすると、その伝達関数は、次式によって表わされる。
【００８８】
【数４】

【００８９】
ここで、各記号については、上述する記号にそれぞれ対応する。
Ｇｃ（ｓ）＝１、＾Ｒ＝Ｒ、＾Ｅ＝Ｅ、Ｋ_ＦＧ＝＾Ｒ・Ｂ、Ｋ_ＶＧ＝＾Ｅ・Ｂとすると、本発明の全体の系１４では、自発呼吸圧力Ｐｍｕｓと支援圧力Ｐｖｅｎｔとを加算した圧力（Ｐｍｕｓ＋Ｐｖｅｎｔ）が、自発呼吸圧力Ｐｍｕｓの（１＋Ｂ）倍に増幅される。すなわち正確に気道抵抗ＲおよびエラスタンスＥを正確に推定可能な場合には、Ｂが０より大きい限り、必ず増幅することができる。
【００９０】
図５は、図２４に示す従来技術の全体の系５を等価変換して示すブロック線図である。従来技術の全体の系５は、自発呼吸圧力Ｐｍｕｓを入力値とし、自発呼吸圧力Ｐｍｕｓと支援圧力Ｐｖｅｎｔとの加算値を出力値とすると、その伝達関数は、次式によって表わされる。
【００９１】
【数５】

【００９２】
ここで各記号は、上述する記号にそれぞれ対応する。
Ｇｃ（ｓ）＝１、＾Ｒ＝Ｒ、＾Ｅ＝Ｅ、Ｋ_ｆａ＝＾Ｒ・Ａ、Ｋ_ｖａ＝＾Ｅ・Ａ、とすると、従来技術の全体の系５では、自発呼吸圧力Ｐｍｕｓと支援圧力Ｐｖｅｎｔとを加算した圧力（Ｐｍｕｓ＋Ｐｖｅｎｔ）が、自発呼吸圧力Ｐｍｕｓの１／（１−Ａ）倍に増幅される。この場合、Ａ＜０またはＡ＞１となると、自発呼吸圧力Ｐｍｕｓを増幅することができない。
【００９３】
上述したように本発明の全体の系１４では、Ｂが０よりも大きい限り増幅することができるが、従来技術の全体の系５では、Ａが０＜Ａ＜１となる必要がある。したがって本発明の全体の系１４は、ゲイン選択の自由度を高くすることができる。
【００９４】
なお、推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅが、実際の気道抵抗Ｒおよび実際のエラスタンスＥに対して、全く同一の値にすることは不可能であり、一般的には、これらの値にずれがある場合（＾Ｒ≠Ｒ、＾Ｅ≠Ｅ）が通常である。さらに人工呼吸器１７を実現するサーボ機構には、何らかの遅れが存在するのが通常である場合（Ｇｃ（ｓ）≠１）であり、このような一般的な場合については後述する。
【００９５】
従来技術および本発明の全体の系５，１４において、患者の状態の変化、各パラメータの設定誤差、外乱などの要因によっては、その過渡応答が不安定となる場合がある。この場合、過度の圧力の支援ガスが患者に供給されてランナウェイを生じる場合がある。
【００９６】
しかしながら本発明の全体の系１４では、後述するように、従来技術の全体の系５に比べて、安定となる余裕度が高い。したがって患者の状態変化が生じても、各パラメータの設定誤差があっても、ゲインを大きく設定しても、外乱などが作用しても、ランナウェイを生じにくくすることができる。
【００９７】
図６は、モデルの安定度を説明するためのグラフである。全体の系において、安定度余裕、すなわち不安定状態になりにくさについて判定するために、ナイキスト線図が用いられる。
【００９８】
たとえば全体の系における安定余裕の判定方法の一つとして、一巡伝達関数のベクトル軌跡４７と安定限界点Ｌとが最も近接する距離∇Ｍに基づいて安定余裕を判定方法がある。この距離∇Ｍは、モジュラスマージン∇Ｍと称される。モジュラスマージン∇Ｍは、一巡伝達関数がＧ_ＯＬ(ｊω)で表わされる場合、次式によって表わされる。なお、モジュラスマージン∇Ｍが大きければ大きいほど、全体の系は不安定になりにくいと判定される。
∇Ｍ＝｜１＋Ｇ_ＯＬ(ｊω)｜_ｍｉｎ …（９）
【００９９】
従来技術の全体の系５におけるモジュラスマージン∇Ｍｙは、（１）式から、次式によって表わされる。
【０１００】
【数６】

【０１０１】
また本発明の実施の一形態の全体の系１４におけるモジュラスマージン∇Ｍｋは、（８）式から、次式によって表わされる。
【０１０２】
【数７】

【０１０３】
ただし、（１０）式および（１１）式は、比較を容易にするために、Ｋ_ｆａ＝＾Ｒ・α、Ｋ_ｖａ＝＾Ｅ・α、Ｋ_ＦＧ＝＾Ｒ・β、Ｋ_ＶＧ＝＾Ｅ・βとする。
【０１０４】
また呼吸器系の推定誤差が本発明の場合と従来の場合とで同一であり、かつ＾Ｒ／Ｒ＝＾Ｅ／Ｅ＝ａとして仮定すると、モジュラスマージン∇Ｍは、次式で表わせる。
【０１０５】
【数８】

【０１０６】
ここでａは、推定値と実際の値とのずれを示し、αおよびβは、自発呼吸圧力Ｐｍｕｓに対して、支援圧力Ｐｖｅｎｔを増幅する増幅率に関係する値である。
【０１０７】
サーボ機構２０の伝達関数Ｇｃ（ｊω）は、一般的に一次遅れ要素とむだ時間要素とを含む。すなわちサーボ機構２０の伝達関数Ｇｃ（ｊω）を次式に示す。
【０１０８】
【数９】

【０１０９】
ここでＴｃは、サーボ機構２０の時定数である。またｅは、自然対数を示し、τは、サーボ機構２０のむだ時間を示す。このような伝達関数をサーボ機構２０が有する場合における、全体の系のナイキスト線図を以下に示す。
【０１１０】
図７は、ａ＜１である場合の本発明の全体の系１４のナイキスト線図を示す。また図８は、ａ＜１である場合の従来技術の全体の系５のナイキスト線図を示す。
【０１１１】
（１２）式から、従来技術の全体の系５の場合、α・ａは、０＜α・ａ＜１となる必要がある。この場合には、一巡伝達関数のベクトル軌跡は、正帰還構成（ポジティブフィードバック）のサーボ機構２０の伝達関数−Ｇｃ（ｊω）の要素を有する。これによって図８に示すように、角周波数ωがゼロの状態から角周波数ωが増加するにつれて、実軸の負の領域から原点Ｏに向かうように描かれる。したがって従来の全体の系５のベクトル軌跡は、角周波数ωがゼロのときに安定限界点Ｌに最も近接する。これによって全体の系５が安定であったとしてもモジュラスマージン∇Ｍｙは小さく、アシスト率αを大きくすると不安定になりやすい。
【０１１２】
（１３）式から、本発明の全体の系１４の場合、ａ＜１であるならば、一巡伝達関数のベクトル軌跡は、負帰還構成（ネガティブフィードバック）のサーボ機構２０の伝達関数Ｇｃ（ｊω）の要素を有する。したがって図７に示すように、角周波数ωがゼロの状態から角周波数ωが増加するにつれて、実軸の正の領域から原点Ｏに向かうように描かれる。したがって本発明の全体の系１４は、角周波数ωが０よりも進んだときに安定限界点Ｌに最も近接する。これによって従来の全体の系５に比べて、本発明の全体の系１４のモジュラスマージン∇Ｍｋは大きくなる。したがって増幅ゲインβを大きくしたとしても不安定になりにくい。また具体的に述べれば、｜＾Ｒ・ｓ＋＾Ｅ｜＜｜Ｒ・ｓ＋Ｅ｜と設定できる限り、必ず負帰還機構に構成することができる。
【０１１３】
たとえば図７に示すβ＝９の本発明の全体の系１４のベクトル軌跡４８と、図８に示すα＝０．９の従来技術の全体の系５のベクトル軌跡４９とは、同じ増幅率に設定した場合を示す。図７および図８からも明らかのように、本発明の全体の系１４のほうが、従来技術の全体の系５に比べてモジュラスマージン∇Ｍが大きく、安定余裕が大きいことが分かる。
【０１１４】
図９は、ａ＞１である場合の本発明の全体の系１４のナイキスト線図を示す。また図１０は、ａ＞１である場合の従来技術の全体の系５のナイキスト線図を示す。本発明の全体の系１４は、ａ＞１となる場合には、正帰還構成の伝達関数Ｇｃ（ｊω）の要素を有する。このような場合であっても、（１２）式および（１３）式から、本発明の全体の系１４のモジュラスマージン∇Ｍｋは、従来技術の全体の系のモジュラスマージン∇Ｍｙよりも大きい。
【０１１５】
たとえば図９に示すβ＝５の本発明の全体の系１４のベクトル軌跡１４８と、図１０に示すα＝０．８３３３の従来技術の系５のベクトル軌跡１４９とは、同じ増幅率に設定した場合を示す。図９および図１０からも明らかのように、ａ＞１であっても、本発明の全体の系１４のほうが従来技術の全体の系５に比べてモジュラスマージン∇Ｍが大きく、安定余裕が大きいことが分かる。
【０１１６】
以上のように、本発明の実施の形態の全体の系１４では、安定余裕を向上することができ、不安定となりにくくすることができる。すなわち各パラメータの設定誤差、患者の状態変化、外乱などによって、制御装置２１が模擬した全体の系１４に対して、実際の全体の系が異なる場合であっても、正帰還構成になりにくく、ランナウェイを防止することができる。これによって患者の負担をさらに低減したサーボ機構の制御方法を実現することができる。
【０１１７】
さらに推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅが、実際の気道抵抗ＲおよびエラスタンスＥに対して少々ずれた場合であっても、上述したように全体の系の安定余裕が大きいので、全体の系１４が不安定となることが防がれ、増幅ゲインβを調整することで、支援圧力Ｐｖｅｎｔを増幅することができる。特に上述したようにａ＜１すなわち、＾Ｒ＜Ｒ、＾Ｅ＜Ｅとなることによって、必ず負帰還のフィードバック構成となり、安定余裕をより大きく設定することができる。したがって医師などが患者の状態を確認して、＾Ｒ＜Ｒ、＾Ｅ＜Ｅとなる推定気道抵抗＾Ｒと推定エラスタンス＾Ｅを設定することによって、正確な気道抵抗ＲおよびエラスタンスＥを求めなくても、ランナウェイが生じる可能性が小さくなり、サーボ機構を制御することができる。
【０１１８】
たとえば現実に起こりうる例として、患者の気道に痰が詰まった場合、一般的に想定する気道抵抗Ｒに比べ、実際の気道抵抗Ｒは大きくなる。このような場合は、本発明の全体の系は、＾Ｒに対してＲが大きくなるので、より安定側に推移することになる。
【０１１９】
また、本発明の実施の形態では、流量偏差ΔＦに応じて目標圧力Ｐｉｎを演算することによって、ほぼリアルタイムの自発呼吸圧力Ｐｍｕｓに基づいて、支援圧力Ｐｖｅｎｔが決定される。したがって、患者の自発呼吸圧力Ｐｍｕｓが逐次その波形のパターン、ピーク値および発生期間が変化したとしても、そのときの自発呼吸圧力Ｐｍｕｓに基づいた支援圧力Ｐｖｅｎｔを与えることができる。
【０１２０】
これによって自発呼吸圧力Ｐｍｕｓに比例増幅した支援圧力Ｐｖｅｎｔの支援ガスを患者の気道に供給するという圧力支援式人工呼吸器の本来の目的をより確実に実現することができる。またサーボ機構から患者を離脱させていく経過処置を好適に実現することができる。
【０１２１】
図１１は、気道圧力演算器５５を示すブロック線図である。たとえば気道圧力演算器５５は、サーボ機構２０を模擬してモデル化したサーボ機構モデルを有する。この場合、気道圧力演算器５５は、目標圧力Ｐｉｎが与えられると、サーボ機構２０が吐出するであろう吐出圧力、すなわち支援圧力Ｐｖｅｎｔを演算する。具体的には、気道圧力演算器５５に設定される伝達関数Ｇｐ(ｓ)が、サーボ機構２０の過渡特性を模擬した伝達関数Ｇｃ(ｓ)・ｅ^−τ・ｓに設定される。
【０１２２】
このように気道圧力演算器５５の伝達関数Ｇｐ（ｓ）が設定されることによって、サーボ機構２０の過渡特性に起因して時間経過とともに変化する演算気道圧力＾Ｐａｗを演算することができる。
【０１２３】
気道圧力演算器５５は、演算した演算気道圧力＾Ｐａｗを、オブザーバ５４に与える。また気道圧力演算器５５は、吸気管路の管路抵抗、制御装置のサンプリング時間、圧力の伝播遅れなどを考慮することによって、気道圧力＾Ｐａｗをさらに正確に推定することができる。
【０１２４】
図１２は、制御量演算手段５３を示すブロック線図である。制御量演算手段５３は、体積演算器６４と、流量演算器６５と、体積ゲイン乗算器６６と、流量ゲイン乗算器６７と、第１加算器６８と、第２加算器６９とを含む。
【０１２５】
体積演算器６４は、支援ガス供給開始時刻から流量偏差ΔＦを順次積算した値に推定エラスタンス＾Ｅを乗算した体積演算値（＾Ｅ・ΔＦ／ｓ）を演算する。体積ゲイン乗算器６６は、前記体積演算値に予め定める体積増幅ゲインβ_ＶＧを乗算し、その値を第１加算器６８に与える。
【０１２６】
また流量演算器６５は、偏差演算手段５２によって演算される流量偏差ΔＦに、推定気道抵抗＾Ｒを乗算した流量演算値（ΔＦ・＾Ｒ）を演算する。流量ゲイン乗算器６７は、前記流量演算値に予め定める流量ゲインβ_ＦＧを乗算し、その値を第１加算器６８に与える。第１加算器６８は、流量ゲイン乗算器６７から与えられる演算値と、体積ゲイン乗算器６６から与えられる演算値とを加算し、目標圧力Ｐｉｎとして、サーボ機構２０に与える。これによって体積増幅ゲインβ_ＶＧおよび流量増幅ゲインβ_ＦＧをそれぞれ個別に設定することができ、利便性を向上することができる。
【０１２７】
また図１２に示すように、流量演算器６４および体積演算器６５は、演算結果を第２加算器６９にそれぞれ与える。第２加算器６９は、流量演算器６４の演算結果と体積演算器６５の演算結果を加算し、自発呼吸圧力＾Ｐｍｕｓとして演算することができる。なお、体積増幅ゲインβ_ＶＧおよび流量増幅ゲインβ_ＦＧが等しい値βとなる場合、演算した自発呼吸圧力＾Ｐｍｕｓに対して（１＋β）倍に増幅した目標圧力Ｐｉｎを演算することができる。
【０１２８】
また制御装置２１は、表示手段６３を有していてもよい。表示手段６３は、自発呼吸圧力推定器６１の第２加算器６９が演算した自発呼吸圧力＾Ｐｍｕｓを取得し、取得した自発呼吸圧力＾Ｐｍｕｓを表示することができる。これによって医師などは、無侵襲で患者の呼吸の強さである自発呼吸圧力Ｐｍｕｓを確認することができる。
【０１２９】
図１３は、本発明の実施の一形態の全体の系１４のシミュレーション結果である。図１３には、制御装置２１が、図１１および図１２に示す構成を有する場合のシミュレーション結果である。また、推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅが、実際の気道抵抗Ｒおよび実際のエラスタンスＥに対してずれがある場合（＾Ｒ≠Ｒ、＾Ｅ≠Ｅ）で、かつサーボ機構２０の伝達関数が、一次遅れ要素とむだ時間要素とを含む場合（Ｇｃ（ｓ）・ｅ^−τ・ｓ≠１）について示す。
【０１３０】
【表１】

【０１３１】
表１には、図１３の全体の系の各パラメータの設定値を示す。シミュレーションでは、模擬的に自発呼吸圧力Ｐｍｕｓを設定し、その設定される自発呼吸圧力Ｐｍｕｓが与えられる場合における、支援ガスの検出流量Ｆ、支援ガスの検出体積Ｖ、推定自発呼吸圧力＾Ｐｍｕｓ、支援圧力Ｐｖｅｎｔの時間応答を求めている。
【０１３２】
図１４は、従来技術の全体の系５のシミュレーション結果である。図１４は、図１３と対応するように各パラメータが設定される。すなわち流量ゲインＫ_ｆａが＾Ｒ・αに設定され、体積ゲインＫ_ｖａが＾Ｅ・αに設定され、増幅率が１０倍に設定される。また他のパラメータについては、図１３と同様である。
【０１３３】
図１４に示すように、従来技術の全体の系５では、速応性が悪い結果として、患者の吸気期間が終了する吸気終了時刻Ｔ１で、支援ガスの流量Ｆがゼロとならず、図１４（２）にＦ１で示す流量の支援ガスが患者の気道に供給されてしまう。すなわち、患者の状態が、吸気状態から呼気状態に切換った後も、支援ガスを患者に供給することになり、いわゆる非同期状態となる。非同期状態となると、患者の負担が大きくなる。
【０１３４】
また図１４（１）および図１４（４）に示すように、全体の系５の速応性の不足によって、支援圧力Ｐｖｅｎｔが自発呼吸圧力Ｐｍｕｓに対して比例した波形とならず、これによっても患者の負担が大きくなる。
【０１３５】
これに対して、本発明の全体の系１４は、サーボ機構２０の過渡特性に応じて支援圧力Ｐｖｅｎｔが設定されるので、図１３（２）に示すように、吸気終了時刻Ｔ１で、支援ガスの検出流量Ｆがゼロとなる。すなわち本発明の全体の系１４では、非同期状態が生じる可能性を小さくする設定、または調整が可能であり、患者の自発呼吸圧力Ｐｍｕｓに応じた吸気期間でのみ、支援ガスを供給することができる。
【０１３６】
また図１３（１）および図１３（５）に示すように、自発呼吸圧力Ｐｍｕｓに比例した支援圧力Ｐｖｅｎｔを与えることができる。本発明の全体の系１４では、増幅ゲインβを増大させても、非同期状態となることがない。また推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅが、実際の推定気道抵抗ＲおよびエラスタンスＥに対して誤差がある場合であっても、非同期状態となることがない。このように非同期状態となることを防止することによって、患者の負担をさらに低減して、人工呼吸を行うことができる。
【０１３７】
このように非同期状態を解消し、自発呼吸圧力Ｐｍｕｓに比例した支援圧力Ｐｖｅｎｔを与えることができるのは、制御装置２１が、サーボ機構２０および空気回路による圧力伝達の遅れを予測したモデルを内方し、そのモデルに基づいて、流量推定値＾Ｆを推定しているからである。
【０１３８】
図１５は、表１に示すパラメータのうち、増幅ゲインβを１９に変更した場合のシミュレーション結果である。図１５に示すように、増幅ゲインβを極端に大きくした場合、本実施の形態の全体の系１４では、図１５（２）に示すように、支援ガスの流量Ｆが振動的となる場合がある。このように支援圧力Ｐｖｅｎｔが振動的となることは、あまり好ましくない。
【０１３９】
図１６は、図１５に示す状態から流量ゲイン乗算器６７によって流量増幅ゲインβ_ＦＧを流量演算値を５０％に減少させた場合のシミュレーション結果である。支援ガスの流量Ｆが振動的な場合には、流量増幅ゲインβ_ＦＧを減少させて、その値を第１加算器６８に与えることによって、図１６に示すように支援ガスの流量が振動的になることを防止することができる。これによって支援ガスの流量が振動的になることなく、増幅率を増幅させることができる。
【０１４０】
また、流量増幅ゲインβ_ＫＧは、ＰＩ制御における比例ゲインに相当する。したがって流量増幅ゲインβ_ＫＧを調整することで、自発呼吸圧力Ｐｍｕｓに対する速応性を向上することができる。また体積増幅ゲインβ_ＶＧは、ＰＩ制御における積分ゲインに相当する。したがって体積増幅ゲインβ_ＶＧを調整することで、目標圧力Ｐｉｎの定常ゲインを調整することができる。
【０１４１】
このように体積増幅ゲインβ_ＶＧおよび流量増幅ゲインβ_ＦＧを調整することによって、定常ゲインと合わせて、速応性および減衰性などの制御特性を向上して目標圧力Ｐｉｎを設定することができる。なお、本発明の全体の系１４は、上述したように、安定性が向上されているので、パラメータ選択の自由度が大きく、増幅ゲインβを増幅したり、体積増幅ゲインβ_ＶＧおよび流量増幅ゲインβ_ＦＧを変更しても、ランナウェイが生じにくく、好適に調整することができる。
【０１４２】
図１７は、人工呼吸器１７の一例を示すブロック図である。制御装置２１は、コンピュータを含む制御装置本体３３と、流量検出手段５０と、入力手段３９と、表示手段４０と、サーボアンプ４８とを含む。また制御装置２１は、気道圧力検出手段６１をさらに含んでいてもよい。
【０１４３】
流量検出手段５０は、サーボ機構２０の吸気管路２５を流れる気体の流量を電気信号に変換し、その電気信号を制御装置本体３３に与える。入力手段３９は、医師および看護士などのサーボ機構２０を管理する管理者からの推定気道抵抗Ｒ＾および推定エラスタンス＾Ｅ、増幅ゲインβ、サーボ機構２０の時定数Ｔｃおよびむだ時間τなどが入力される。入力手段３９は、入力された情報を示す信号を制御装置本体３３に与える。
【０１４４】
表示手段４０は、患者の気道圧力を報知する報知手段である。表示手段４０は、制御装置本体３３から受ける表示指令信号に基づいて、患者の自発呼吸圧力Ｐｍｕｓの時間的変化を示す波形を表示画面に表示する。
【０１４５】
サーボアンプ４８は、制御装置本体３３が演算した目標圧力Ｐｉｎを示す信号をポンプ用アクチュエータ３１に与える。ポンプ用アクチュエータ３１は、目標圧力Ｐｉｎを示す信号に基づいてポンプを制御し、サーボ機構２０の吐出圧力がフィードバック制御される。
【０１４６】
制御装置本体３３は、インターフェース１０１と、演算部１０２と、一時記憶部１０３と、記憶部１０４とを含む。インターフェース１０１は、接続される流量検出手段５０からの信号が入力されて、その信号を演算部１０２に与える。記憶部１０４は、制御装置本体３３が実行すべきプログラムが記憶され、演算部１０２が記憶部１０４に記憶されるプログラムを読み出して実行することによって、前記流量推定手段５１、偏差演算手段５２、制御量演算手段５３を実現することができる。これによって制御装置本体３３は、前述するサーボ機構２０の制御を行うことができる。また記憶部１０４は、コンパクトディスクなどのコンピュータ読取可能な記録媒体であってもよい。
【０１４７】
サーボ機構２０は、制御装置２１によって、吐出する支援ガスの圧力が制御可能なものであり、患者の気道１５に支援ガスを導く吸気管路２５が形成されていれば、特に限定されない。たとえば図１７に示すようにベローズ型ポンプを有する人工呼吸器であってもよい。また配管を介して支援ガスを供給する人工呼吸器であってもよい。
【０１４８】
サーボ機構２０の過渡特性は、患者の状態に比べて大きく変動することがなく、予めその特性を求めることができる。たとえばサーボ機構２０の伝達関数が予め求められて、その伝達関数の要素が前記気道圧力演算器５５に設定される。同様に流量検出手段の検出遅れなども予め求められて、制御装置２１に与えられる。
【０１４９】
図１８は、本発明のさらに他の実施の形態の全体の系１３を示すブロック線図である。図１８に示す全体の系１３は、図２に示す全体の系１４に対して、流量推定手段５１の構成の一部が異なる以外は、同一の構成を有する。したがって同様の構成については、説明を省略し、図２の全体の系１４に対応する符号を付する。
【０１５０】
流量推定手段５１は、検出遅れ演算器６０をさらに有する。検出遅れ演算器６０は、流量検出手段５０を模擬してモデル化した検出手段モデルを有する。この場合、検出遅れ演算器６０は、オブザーバ５４から推定流量＾Ｆが与えられると、流量検出手段５０の検出遅れに基づいた推定流量＾Ｆを演算し、この演算結果を偏差演算手段５２に与える。偏差演算手段５２は、検出遅れ演算器６０が演算した推定流量＾Ｆから、流量検出手段５０によって検出された検出流量を減算して、流量偏差ΔＦを演算する。これによって流量推定手段５１は、患者の気道に供給された支援ガスの流量Ｆを検出してから、検出手段５０が検出結果を出力するまでの時間変化に基づいて、患者の気道に供給されるべき支援ガスの流量＾Ｆをさらに精度よく演算することができる。したがって非同期状態をさらに確実に防止することができる。
【０１５１】
図１９は、本発明に関連する実施の形態の全体の系１２を示すブロック線図である。図１９に示す全体の系１２は、図２に示す全体の系１４に対して、流量推定手段５１の構成の一部が異なる以外は、同一の構成を有する。したがって同様の構成については、説明を省略し、図２の全体の系１４に対応する符号を付する。
【０１５２】
流量推定手段５１に換えて、圧力検出手段６１を備えてもよい。圧力検出手段６１は、患者の気道内の圧力である気道圧力Ｐａｗを検出する。そして圧力検出手段６１は、検出した気道圧力Ｐａｗをオブザーバ５４に与える。このようにしても、上述した効果を達成することができる。また気道圧力Ｐａｗを検出することで、サーボ機構２０の遅れの影響を考慮することなく、気道圧力Ｐａｗを取得することができ、自発呼吸圧力Ｐｍｕｓに正確に対応した支援圧力Ｐｖｅｎｔで支援ガスを患者の気道に与えることができる。
【０１５３】
図２０は、本発明のさらに他の実施の形態の全体の系１１を示すブロック線図である。図２０に示す全体の系１１は、図２に示す全体の系１４の気道圧力演算器５５に設定される伝達関数Ｇｐ(ｓ)が、次式で表わされた場合について、等価変換したモデルである。
【０１５４】
【数１０】

【０１５５】
ここで、γは、０＜γ＜１に設定される係数であり、他の記号については、上述する記号にそれぞれ対応する。この伝達関数は、実際の気道抵抗Ｒ、実際の肺エラスタンスＥが既知である場合を示しており、実際の気道抵抗Ｒ、実際の肺エラスタンスＥが不明の場合には、（１５）式は、次式によって代用される。
【０１５６】
【数１１】

【０１５７】
このように気道圧力演算器５５の伝達関数を設定して、図２０に示す全体の系１１に設定することによって、サーボ機構２０のむだ時間要素による遅れを、γ・Ｇｃ（ｓ）・ｅ^−τ・ｓの要素によって補償することができ、好適にサーボ機構のむだ時間の遅れを補償することができる。
【０１５８】
このように気道圧力演算器５５は、目標圧力Ｐｉｎに基づいて気道圧力Ｐａｗを推定できればよい。したがって気道圧力演算器５５の伝達関数Ｇｐ(ｓ)が、サーボ機構２０の特性を模擬した伝達関数Ｇｃ(ｓ)以外に設定されていてもよい。
【０１５９】
図２１は、本発明のさらに他の実施の形態の全体の系１０を示すブロック線図である。図２１に示す全体の系１０は、図２に示す全体の系１４に対して、流量推定手段５１に設定される推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅの設定が異なる以外は、同一の構成を有する。したがって同様の構成については、説明を省略し、図２の全体の系１４に対応する符号を付する。
【０１６０】
図２２は、気道抵抗Ｒを説明するためのグラフである。気道内を流れる支援ガスの流れが層流となる場合には、気道圧力Ｐａｗに比例して、その流速が直線的に変化する。しかし実際には、気道は、分岐を繰り返し太さも均一でないので、支援ガスの流れは乱流となる。したがって乱流抵抗を考慮して推定気道抵抗＾Ｒを設定する。
【０１６１】
具体的には、流量推定手段５１に設定される前記推定気道抵抗＾Ｒは、支援ガスの流量にかかわらず一定に設定される第１抵抗係数＾Ｒ_Ｔと、前記推定流量演算器で演算される支援ガスの流量＾Ｆに基づく第２抵抗係数＾Ｋ_Ｔとを加算した値である。第１抵抗係数＾Ｒ_Ｔおよび第２抵抗係数＾Ｋ_Ｔは、患者の気道抵抗に応じた係数に設定される。また肺のみならず胸郭などを含めた呼吸器官全体の抵抗を、推定気道抵抗＾Ｒとして設定してもよい。他の近似式で表わされる推定気道抵抗＾Ｒによって気道抵抗Ｒを近似してもよい。
【０１６２】
図２３は、肺のコンプライアンスを説明するためのグラフである。流量推定手段５１に設定される推定エラスタンス＾Ｅは、前記支援ガス体積演算器で演算される支援ガス体積＾Ｖに基づく値であり、肺のコンプライアンスＣの逆数となる。コンプライアンスＣは、患者の吸気期間中においては、支援ガスの体積Ｖの増加とともに非線形的に増加し、飽和特性とヒステリシス特性とを有する。
【０１６３】
肺胞圧力演算器５９が、予めコンプライアンスＣと支援ガスの体積との関係を示す情報を予め取得することによって、コンプライアンスが非線形である場合を考慮したであっても、正確に肺胞圧力Ｐａｌｖを演算することができる。
【０１６４】
このようにより非線形となる呼吸器管のモデルをオブザーバが有することによって、より精度よく推定流量＾Ｆを推定することができる。これによって自発呼吸圧力Ｐｍｕｓを精度よく推定することができるとともに、推定した自発呼吸圧力Ｐｍｕｓに応じて、支援圧力Ｐｖｅｎｔを決定することができる。
【０１６５】
上述した本発明の実施の形態は、本発明の一例示であって、発明の範囲内において、構成を変更することができる。たとえば上述したブロック線図は、本発明の例示に過ぎず、同様の効果を得ることができるならば、等価変換されてもよい。また検出流量Ｆが与えられたコンピュータが目標圧力Ｐｉｎを演算する場合、各手段がソフトウエアによって実現されてもよい。
【０１６６】
また推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅは、医師が適切に設定してもよいが、予め計測器器によって測定した気道抵抗ＲおよびエラスタンスＥを用いてもよい。また特許文献２に開示される推定方法によって求められる気道抵抗＾Ｒおよびエラスタンス＾Ｅを用いてもよい。
【０１７０】
【発明の効果】
以上のように請求項１記載の本発明によれば、制御量演算手段が自発呼吸圧力Ｐｍｕｓを表わす値である流量偏差ΔＦに基づいて、目標圧力Ｐｉｎを決定する。したがって自発呼吸圧力Ｐｍｕｓに比例増幅した支援圧力Ｐｖｅｎｔの支援ガスを患者の気道に供給するという圧力支援式人工呼吸器の本来の目的を確実に実現することができる。
【０１７１】
さらに検出した支援ガスの流量Ｆにゲインを直接付与して圧力情報値を作成する従来技術に比べて、全体の系の安定限界に対する余裕を大きくすることができる。これによってランナウェイが生じにくい系を構成することができる。したがって外乱が生じたり、時間遅れがあったり、患者の呼吸器官の状態が変化したりして、実際の全体の系が変動した場合であっても、正帰還構成になりにくく、ランナウェイを防止することができる。ランナウェイを防止することによって、患者の負担をさらに低減したサーボ機構の制御方法を実現することができる。
また気道圧力演算器によって患者の気道内の圧力＾Ｐａｗを演算することによって、精度よく支援ガスの流量＾Ｆを求めることができる。たとえば気道圧力演算器は、検出流量の検出遅れ、制御装置の演算遅れなどを考慮して気道内の圧力＾Ｐａｗを演算することによって、人工呼吸における非同期状態を防ぐことができる。呼吸器官モデルは、比較器と、推定流量演算器と、支援ガス体積演算器とを有し、推測される患者の気道抵抗＾Ｒおよびエラスタンス＾Ｅが設定される。これによって患者の呼吸器官モデルを模擬してモデル化することができる。
【０１７２】
また請求項２記載の本発明によれば、サーボ機構の時間特性に応じて支援ガスの流量＾Ｆを演算することで、より精度よく支援ガスの流量＾Ｆを演算することができる。また患者が支援ガスを吸引する呼気期間と、サーボ機構が支援ガスを患者の気道に供給する供給期間とがずれることを防止することができる。いわゆる非同期状態を防ぐことができる。これによって患者の呼吸動作における負担をさらに低減することができる。
【０１７３】
また請求項３記載の本発明によれば、検出手段の時間特性に応じて支援ガスの流量＾Ｆを演算する。たとえば流量推定工程では、むだ時間要素を考慮した検出手段の伝達関数に基づいて支援ガスの流量＾Ｆを演算する。これによってより精度よく支援ガスの流量＾Ｆを算出することができる。また検出手段の検出遅れに起因して生じる非同期状態を防ぐことができる。したがって患者の呼吸動作における負担をさらに低減することができる。
【０１７４】
また請求項４記載の本発明によれば、流量偏差ΔＦを前述するように流量ゲインＫ_ＦＧおよび体積ゲインＫ_ＶＧ を付与することによって、患者の自発呼吸圧力Ｐｍｕｓの大きさに比例する支援圧力Ｐｖｅｎｔを与えることができる。これによって自発呼吸圧力の時間変化に応じた支援圧力Ｐｖｅｎｔで支援ガスを患者に供給することができ、患者の負担を低減することができる。
【０１７５】
また各流量ゲインＫ_ＦＧおよび体積ゲインＫ_ＶＧを調整することによって、自発呼吸圧力Ｐｍｕｓに対する速応性を向上するとともに、支援圧力Ｐｖｅｎｔが振動的になることを防ぐことができる。
【０１７７】
また請求項５記載の本発明によれば、推定気道抵抗＾Ｒとして第１抵抗係数と第２抵抗係数とを加算した値に設定される。また推定エラスタンス＾Ｅとして前記支援ガス体積演算器で演算される支援ガス体積＾Ｖに基づいて設定される。これによって患者の呼吸器系のモデルを精度よく模擬してモデル化することができ、患者の状態に精度よく対応した支援圧力Ｐｖｅｎｔで支援ガスを供給することができる。これによって患者の負担をさらに低減することができる。
【０１７８】
また請求項６記載の本発明によれば、前記推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅは、制御中に変更可能に設定されることで、利便性を向上することができる。たとえば患者の状態、サーボ機構の種類などに応じて、推定気道抵抗＾Ｒおよび推定エラスタンス＾Ｅを変更することで、患者の呼吸器系のモデルを実際の気道抵抗および肺のエラスタンスの変化に追従させることができる。
【０１７９】
またたとえば支援ガスの流量変化が振動的となる場合には、推定気道抵抗Ｒに相当するゲインを減少させることで、支援ガスの流量変化が振動的となることを防止することができる。これによってさらに患者の負担を軽減することができる。
【図面の簡単な説明】
【図１】本発明の実施の一形態の人工呼吸器１７と患者１８を示すブロック図である。
【図２】本発明の実施の一形態の全体の系１４を具体的に示すブロック線図である。
【図３】本発明の全体の系１４における自発呼吸時の換気量Ｖｍｕｓとアシスト呼吸時の換気量Ｖａｓｔとの関係を示すグラフである。
【図４】図２の本発明の全体の系１４を等価変換して整理して示すブロック線図である。
【図５】図２４に示す従来技術の全体の系５を等価変換して示すブロック線図である。
【図６】モデルの安定度を説明するためのグラフである。
【図７】ａ＜１である場合の本発明の全体の系１４のナイキスト線図を示す。
【図８】ａ＜１である場合の従来技術の全体の系５のナイキスト線図を示す。
【図９】ａ＞１である場合の本発明の全体の系１４のナイキスト線図を示す。
【図１０】ａ＞１である場合の従来技術の全体の系５のナイキスト線図を示す。
【図１１】気道圧力演算器５５を示すブロック線図である。
【図１２】制御量演算手段５３を示すブロック線図である。
【図１３】本発明の実施の一形態の全体の系１４のシミュレーション結果である。
【図１４】従来技術の全体の系４のシミュレーション結果である。
【図１５】表１に示すパラメータのうち、増幅ゲインβを１９に変更した場合のシミュレーション結果である。
【図１６】図１５に示す状態から流量ゲイン乗算器６７によって流量増幅ゲインβ_ＦＧを流量演算値を５０％に減少させた場合のシミュレーション結果である。
【図１７】人工呼吸器１７の一例を示すブロック図である。
【図１８】本発明のさらに他の実施の形態の全体の系１３を示すブロック線図である。
【図１９】 本発明に関連する実施の形態の全体の系１２を示すブロック線図である。
【図２０】本発明のさらに他の実施の形態の全体の系１１を示すブロック線図である。
【図２１】本発明のさらに他の実施の形態の全体の系１０を示すブロック線図である。
【図２２】気道抵抗Ｒを説明するためのグラフである。
【図２３】肺のコンプライアンスを説明するためのグラフである。
【図２４】従来技術の人工呼吸器１と患者２とを含む全体の系５を示すブロック線図である。
【図２５】従来技術の全体の系５における自発呼吸時の換気量Ｖｍｕｓとアシスト呼吸時の換気量Ｖａｓｔとの関係を示すグラフである。
【符号の説明】
１４ 全体の系
１７ 人工呼吸器
２０ ガス供給サーボ機構
２１ 制御装置
５０ 流量検出手段
５１ 流量推定手段
５２ 偏差演算手段
５３ 制御量演算手段
５４ オブザーバ
Ｐｍｕｓ 自発呼吸圧力
Ｐｉｎ 目標圧力
Ｐｖｅｎｔ 支援圧力
Ｆ 実際の支援ガスの流量
＾Ｆ 推定される支援ガスの流量
ΔＦ 流量偏差[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to a gas supply servo mechanism used for a ventilator.System ofRelated to the device.
[0002]
[Prior art]
Proportional Assist Ventilation (Proportional Assist Ventilation, abbreviated PAV method) as a control method for the ventilator gas supply servomechanism (hereinafter simply referred to as the servomechanism) during the inspiratory period when the patient is breathing spontaneously )
[0003]
FIG. 24 is a block diagram showing the entire system 5 including a prior art ventilator 1 and a patient 2. A ventilator 1 that realizes the PAV method includes a servo mechanism 3 and a control device 4 that controls the servo mechanism 3. The control device 4 detects the flow rate of the support gas and determines the discharge pressure of the servo mechanism 3 based on the detected flow rate.
[0004]
The control device 4 has a flow rate gain K_faIs multiplied by the flow rate F of the support gas to the first calculated value (K_faF) and volume gain K_vaIs multiplied by the volume V of the support gas supplied into the patient's lung (K_vaF / s) is calculated, and the target pressure Pin is calculated by adding the first calculation value and the second calculation value. Flow rate gain K_faIs a value obtained by multiplying the estimated airway resistance ^ R by the assist rate α, and the volume gain K_vgIs a value obtained by multiplying the estimated lung elastance ^ E by the assist rate α. Here, let A be the assist rate α when ^ R = R and ^ E = E.
[0005]
The control device 4 gives the calculated target pressure Pin to the servo mechanism 3. The servo mechanism 3 to which the target pressure Pin is given discharges the support gas at the support pressure Pvent based on the target pressure Pin. Flow rate gain K_faAnd volume gain K_vaIs appropriately set, the servo mechanism 3 gives a support pressure Pvent obtained by proportionally amplifying the patient's spontaneous breathing pressure Pmus (see, for example, Patent Document 1).
[0006]
  FIG. 25 is a graph showing the relationship between the ventilation volume Vmus during spontaneous breathing and the ventilation volume Vast during assist breathing in the entire system 5 of the prior art. As described above, when the assist rate α in the case of ^ R = R and ^ E = E is A, the support pressure Pvent is 1 / (the spontaneous breathing pressure Pmus according to the time change of the spontaneous breathing pressure Pmus. 1−A) It is amplified at a double amplification factor. As a result, the ventilation volume Vast during assisted breathing by the ventilator is 1 / (1 of the ventilation volume Vmus during spontaneous breathing.−A) It is amplified twice.
[0007]
In the above-described prior art, the patient's airway resistance ^ R and lung elastance ^ E are accurately determined and the flow rate gain K is determined._faAnd volume gain K_fgNeed to be determined. Therefore, as another conventional technique, a method for determining a patient's airway resistance R and lung elastance E is disclosed (for example, Patent Document 2).
[0008]
[Patent Document 1]
Japanese Patent Publication No. 2714288
[Patent Document 2]
Japanese National Patent Publication No. 11-502755
[0009]
[Problems to be solved by the invention]
The entire system 5 of the prior art supplies the support gas to the patient at a pressure (Pvent + Pmus) obtained by adding the support pressure Pvent and the spontaneous breathing pressure Pmus. Therefore, a positive feedback mechanism is included. This positive feedback mechanism may cause the entire system to become unstable, and may cause over assist called runaway.
[0010]
For example, the runaway is a flow gain K_faAnd volume gain K_vaIs not appropriate, when there is a delay in the servo mechanism 3, when the patient's condition fluctuates, or when other disturbances are applied. When runaway occurs, the flow rate of the support gas diverges and amplifies without converging. This may destroy the patient's lungs and airways, and there is a case where the assist is forced to stop in the conventional technique.
[0011]
The reason why runaway is likely to occur is that the stability margin of the entire system 5 of the prior art is small and the transient response tends to become unstable. If the stability margin of the entire system 5 is small, even if the entire system is slightly changed, the stability limit may be exceeded and a runaway may occur. The flow rate gain K is set so that the entire system 5 does not become unstable._faAnd volume gain K_vaTherefore, there is a problem that the degree of freedom in gain selection is low and it is difficult to adjust an appropriate gain.
[0012]
  Therefore, an object of the present invention is to provide a gas supply servo mechanism for a ventilator that can increase the stability margin in the overall shape including the ventilator and the patient.System ofIs to provide a control device.
[0019]
[Means for Solving the Problems]
  ContractClaim1The described invention is based on the target pressure Pin corresponding to the patient's spontaneous breathing pressure Pmus, and supplies the assist gas containing oxygen to the patient's airway at the assist pressure Pvent corresponding to the patient's spontaneous breathing pressure Pmus. A control device for controlling a gas supply servomechanism of a assisted ventilator,
  Flow rate detection means for detecting a flow rate F of support gas supplied to the patient's airway;
  A flow rate estimation means for estimating a flow rate F of the support gas to be supplied to the patient's airway when the support gas is supplied at the support pressure Pvent, the respiratory organ model having the patient's respiratory organ modeled;
  Deviation calculating means for calculating a flow deviation ΔF between the detected flow rate F and the estimated flow rate ^ F;
  Control amount calculation means for calculating the target pressure Pin based on the flow rate deviation ΔF and providing the target pressure Pin to the servo mechanism.See
  The flow rate estimation unit further includes an airway pressure calculator that calculates a pressure ^ Paw in the patient's airway based on the target pressure Pin calculated by the control amount calculation unit,
  The respiratory organ model includes a comparator that subtracts the alveolar pressure ^ Palv generated by the elastic restoring force of the lung from the pressure ^ Paw in the airway calculated by the airway pressure calculator when the spontaneous breathing pressure Pmus does not exist;
  An estimated flow rate calculator for estimating the flow rate of support gas to be supplied to the patient's airway by dividing the subtracted value subtracted by the comparator by a preset estimated airway resistance ＲR;
  Support for calculating the volume of support gas ^ V to be supplied to the patient's airway from the support gas supply start time by sequentially integrating the support gas flow rate ^ F calculated by the estimated flow rate calculator from the support gas supply start time A gas volume calculator,
  An alveolar pressure calculator that calculates the alveolar pressure ^ Palv by multiplying the calculated volume of assisting gas ^ V by a preset lung estimated elastance ^ E, and supplies it to the subtractor.This is a control device for a gas supply servomechanism of a ventilator.
[0020]
  Claim1According to the described invention, the flow rate detection means detects the flow rate F of the support gas, and the flow rate estimation means estimates the flow rate F of the support gas by the flow rate estimation means. The detected support gas flow rate F varies depending on the patient's spontaneous breathing pressure Pmus, but the estimated support gas flow rate ^ F is not affected by the patient's spontaneous breathing pressure Pmus. Therefore, by obtaining the flow rate deviation ΔF between the support gas flow rate F detected by the deviation calculating means and the estimated support gas flow rate F, information related to the patient's spontaneous breathing pressure Pmus can be acquired.
[0021]
The control amount calculation means calculates a target pressure Pin for controlling the servo mechanism based on the flow rate deviation ΔF. Therefore, the target pressure Pin is also a pressure corresponding to the patient's spontaneous breathing pressure Pmus. By applying the target pressure Pin calculated in this way to the servo mechanism, the support gas can be supplied to the patient's airway at the support pressure Pvent corresponding to the spontaneous breathing pressure Pmus that changes sequentially.
[0022]
  Further, the control amount calculation means calculates the target pressure Pin based on the flow rate deviation ΔF, so that compared with the conventional technique in which the target pressure Pin is calculated based only on the detected flow rate F of the support gas, the control amount calculation means The entire system including the respirator can be made less likely to have a positive feedback configuration, and a margin for the stability limit of the entire system can be increased. If this causes a disturbance, if there is a time delay in the servo mechanism, if the patient's respiratory organ model cannot be set accurately, or if the patient's condition changes, etc., runaway can be made difficult to occur. . As a result, the burden on the patient can be reduced.
  Further, based on the target pressure Pin calculated by the airway pressure calculator, the pressure ^ Paw in the patient's airway is calculated, and the support gas flow rate ^ F is estimated from the pressure ^ Paw in the airway. Specifically, the flow rate of support gas ^ F is obtained by subtracting the alveolar pressure Palv from the pressure ^ Paw in the airway and dividing the subtracted value by the estimated airway resistance ^ R by each calculator and comparator. . The alveolar pressure Palv is obtained by integrating the calculated support gas flow rate {circumflex over (F)} F to obtain the support gas volume {circumflex over (V)}, and multiplying the support gas volume {circumflex over (V)} by the estimated elastance ^ E. Thus, a model of the patient's respiratory tract can be realized.
  For example, as long as | ^ R · s + ^ E | <| R · s + E | is set, the entire system including the patient, the servo mechanism, and the control device can always be configured as a negative feedback mechanism. As a result, the margin for the stability limit can be increased. In the above equation, ^ R is the estimated airway resistance, ^ E is the estimated elastance, R is the actual patient's airway resistance, E is the actual patient's lung elastance, and s is the Laplace operator.
[0023]
  And claims2In the described invention, the flow rate estimation means has a servo mechanism model in which the servo mechanism is modeled, and after the target pressure Pin is given to the servo mechanism, the servo mechanism supplies the support gas at the support pressure Pvent. On the basis of the time characteristic until it is performed, the flow rate F of the support gas to be supplied to the patient's airway is estimated.
[0024]
  Claim2According to the described invention, the flow rate F of the support gas is calculated according to the time characteristic of the servo mechanism. For example, in the flow rate estimation step, the flow rate {circumflex over (F)} of the support gas is calculated based on the transfer function of the servo mechanism that takes into account the dead time factor. As a result, the flow rate F of the support gas can be calculated with higher accuracy. Further, it is possible to prevent the intake period during which the patient sucks the support gas from the supply period during which the servo mechanism supplies the support gas to the patient's airway. A so-called asynchronous state can be prevented. This can further reduce the burden on the patient's breathing motion.
[0025]
  And claims3In the present invention described above, the flow rate estimation unit has a detection unit model that models the flow rate detection unit, and the detection unit detects the detection result after detecting the flow rate F of the support gas supplied to the patient's airway. Based on the time characteristic until output, the flow rate {circumflex over (F)} of assisting gas to be supplied to the patient's airway is estimated.
[0026]
  Claim3According to the described invention, the flow rate F of the support gas is calculated according to the time characteristic of the detection means. For example, in the flow rate estimation step, the flow rate {circumflex over (F)} F of the support gas is calculated based on the transfer function of the detection means considering the dead time element. As a result, the flow rate F of the support gas can be calculated with higher accuracy. Further, an asynchronous state caused by the detection delay of the detection means can be prevented. This can further reduce the burden on the patient's breathing motion.
[0027]
  And claims4According to the present invention described above, the control amount calculation means is configured so that the flow rate gain K is set in advance._FGIs multiplied by the flow rate deviation ΔF, and a preset volume gain K_VGAnd a second calculated value obtained by multiplying the integral value of the flow rate deviation ΔF by
  The target pressure Pin is calculated by adding the first calculation value and the second calculation value.
[0028]
  Claim4According to the described invention, the flow rate deviation ΔF is a change in flow rate caused by the spontaneous breathing pressure Pmus. By calculating the flow rate deviation ΔF as described above, the target pressure Pin corresponding to the spontaneous breathing pressure Pmus can be calculated. In this case, flow rate gain K_FGAnd volume gain K_VGIs set appropriately, the support gas can be supplied to the patient's respiratory tract with a pressure proportionally amplified to the spontaneous breathing pressure Pmus.
[0029]
For example, if the patient's respiratory organ model can be obtained accurately, the flow gain K_FGIs B times greater than the patient's airway resistance R and the volume gain K_VGIs set to be B times larger than the lung elastance E, the support gas can be supplied to the patient's respiratory tract at a pressure (1 + B) times the spontaneous breathing pressure Pmus.
[0030]
  Further claims1As shown in the present invention, since the stability of the entire system can be improved, the degree of freedom in gain selection can be increased, and an appropriate gain can be selected. For example, flow rate gain K_FGCorresponds to a proportional gain in PI control. Therefore flow rate gain K_FGBy adjusting, the quick response to the spontaneous breathing pressure Pmus can be improved. Volume gain K_VGCorresponds to an integral gain in PI control. Therefore volume gain K_FGThe steady gain of the target pressure Pin can be adjusted by adjusting.
[0035]
  And claims5According to the present invention, the estimated airway resistance R is a first resistance coefficient that is set to be constant regardless of the flow rate of the support gas, and a first flow rate based on the flow rate F of the support gas calculated by the estimated flow rate calculator. It is a value obtained by adding 2 resistance coefficients,
  The estimated elastance ^ E is a value based on the assist gas volume ^ V calculated by the assist gas volume calculator.
[0036]
  Claim5According to the described invention, the estimated airway resistance R is set to a value obtained by adding the first resistance coefficient and the second resistance coefficient. The estimated elastance ^ E is set based on the support gas volume ^ V. As a result, the respiratory organ model of the flow rate estimation means can realize an actual patient's respiratory tract more accurately. As a result, the target pressure Pin corresponding to the spontaneous breathing pressure Pmus with high accuracy can be calculated.
[0037]
Estimated airway resistance ^ R and estimated elastance are set to variables representing coefficients related to the patient's respiratory tract. For example, the estimated airway resistance {circumflex over (R)} is determined based on Roehl's general formula considering turbulent resistance. The estimated elastance ^ E is determined on the basis of the compliance saturation characteristic and hysteresis characteristic, which are the reciprocals thereof.
[0038]
  And claims6According to the present invention described above, at least one of the estimated airway resistance ^ R and the estimated elastance ^ E is determined based on either the flow rate F of the support gas supplied to the patient's airway or an input value input from the outside. It further comprises changing means for changing.
[0039]
  Claim6According to the described invention, the estimated airway resistance ^ R and the estimated elastance ^ E are set to be changeable during the control, thereby improving convenience. For example, by changing the estimated airway resistance ^ R and the estimated elastance ^ E according to the patient's condition, the type of servomechanism, etc., the patient's respiratory system model can be changed in the actual airway resistance and lung elastance. Can be followed.
[0045]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a block diagram showing a ventilator 17 and a patient 18 according to an embodiment of the present invention. The ventilator 17 includes a ventilator gas supply servo mechanism (hereinafter simply referred to as a servo mechanism) 20 and a control device 21 that controls the servo mechanism 20. The servomechanism 20 supplies a support gas 16 containing oxygen to the patient's airway 15. The support gas 16 is, for example, air in the atmosphere that is appropriately pressurized. The servo mechanism 20 can control gas supply means such as a pump, for example, and can control the pressure of the support gas discharged from the gas supply means.
[0046]
When the patient 18 performs spontaneous breathing, a proportional assist ventilation method (proportional assist ventilation,
Proportional Assist Ventilation (abbreviated as PAV method). The control device 21 of the present invention controls the servo mechanism 20 in accordance with the original purpose of the PAV method. The servo mechanism 20 supplies the support gas 16 to the patient's airway 15 at a support pressure Pvent proportional to the spontaneous breathing pressure Pmus. Spontaneous breathing pressure Pmus is a pressure acting from the outside of the lungs caused by movement of respiratory muscles such as the diaphragm. In the present embodiment, it is assumed that the support pressure Pvent is substantially equal to the discharge pressure of the servo mechanism 20.
[0047]
The servomechanism 20 controlled according to the PAV method supplies the support gas 16 to the patient at a higher pressure as the patient 18 sucks the support gas 16 more strongly. Further, as the suction force of the patient 18 becomes weaker, the pressure of the support gas 16 to be supplied is lowered, and the supply of the support gas 16 is stopped while the patient finishes sucking the support gas.
[0048]
By controlling the servo mechanism 20 in this manner, the support gas 16 can be supplied at a pressure corresponding to the respiratory effort of the patient 18, and the burden on the patient 18 in the breathing operation can be reduced.
[0049]
The control device 21 calculates a target pressure Pin corresponding to the patient's spontaneous breathing pressure Pmus, and gives the target pressure Pin to the servo mechanism 20. The servo mechanism 20 to which the target pressure Pin is applied supplies the support gas 16 to the patient's airway 15 at the support pressure Pvent corresponding to the patient's spontaneous breathing pressure Pmus.
[0050]
In the embodiment of the present invention, a symbol with “(s)” indicates a transfer function in the Laplace region, and a symbol with “(jω)” is a frequency transfer function. It shows that. Also, a value with “^” indicates that it is an estimated value or an operation value, not an actual value, and “s” indicates a Laplace operator.
[0051]
The control device 21 includes a flow rate detection unit 50, a flow rate estimation unit 51, a deviation calculation unit 52, and a control amount calculation unit 53. The flow rate detection means 50 detects the flow rate F of the support gas 16 actually supplied to the patient's airway 15. Hereinafter, the flow rate detected by the flow rate detection means 50 may be referred to as a detected flow rate F. The detected flow rate F is a flow rate of the respiratory system in a state where the spontaneous breathing pressure Pmus is applied.
[0052]
The flow rate detection means 50 measures the flow rate of the support gas 16 flowing through the intake pipe line 25. The intake line 25 is a line that guides the support gas 16 from the pressure source of the servo mechanism 20 to the patient's airway. For example, the flow rate detection means 50 is realized by a differential pressure type flow meter. When the flow rate detection unit 50 detects the flow rate F of the support gas, the flow rate detection unit 50 gives the detected flow rate F to the deviation calculation unit 52.
[0053]
The flow rate estimation means 51 has an observer 54 that is a respiratory organ model that is modeled by simulating a patient's respiratory tract. Based on the information corresponding to the target pressure Pin calculated by the control amount calculation means 53, the observer 54 calculates the flow rate ^ F of the support gas that will be supplied to the patient when the spontaneous breathing pressure Pmus does not exist. To do.
[0054]
  Hereinafter, the flow rate estimated by the flow rate estimation means 51 is referred to as the estimated flow rate.^Sometimes referred to as F. The estimated flow rate {circumflex over (F)} is the respiratory system flow rate at the calculated airway pressure value {Paw} corresponding to the support pressure Pvent. When the flow rate estimation means 51 estimates the flow rate {circumflex over (F)} of the support gas, the flow rate estimation means 51 gives the estimated flow rate {circumflex over (F)} to the deviation calculation means 52.
[0055]
The deviation calculation means 52 calculates a flow rate deviation ΔF that is a value obtained by subtracting the estimated flow rate ^ F from the detected flow rate F, and gives the calculation result to the control amount calculation means 53. The control amount calculation means 53 gives a preset gain to the flow rate deviation ΔF, and calculates a target pressure Pin related to the support pressure Pvent.
[0056]
The control amount calculation unit 53 gives the calculated target pressure Pin to the flow rate estimation unit 51 and the servo mechanism 20, respectively. The servo mechanism 20 supplies the support gas 16 to the patient's airway 15 at the discharge pressure based on the target pressure Pin given from the control amount calculation means 53, that is, the support pressure Pvent. The flow rate estimating unit 51 sequentially calculates the estimated flow rate ^ F based on the target pressure Pin given from the control amount calculating unit 53.
[0057]
FIG. 2 is a block diagram specifically showing the entire system 14 according to the embodiment of the present invention. The flow rate estimation means 51 further includes a delay compensation unit 55 in addition to the observer 54. The delay compensation unit 55 compensates for delay elements such as a primary delay element and a dead time element of each component constituting the entire system 14 such as a delay element of the servo mechanism 20 and a delay element of the air circuit. In the present embodiment, the delay compensation unit 55 is the airway pressure calculator 55.
[0058]
The airway pressure calculator 55 calculates the patient's airway pressure ^ Paw based on the target pressure Pin. Hereinafter, the patient's airway pressure calculated by the airway pressure calculator 55 may be referred to as a calculated airway pressure ^ Paw, and the actual patient's airway pressure may be simply referred to as an airway pressure Paw. The airway pressure calculator 55 gives the calculated airway pressure ^ Paw to the observer 54.
[0059]
Based on the patient's respiratory organ model, the observer 54 estimates the estimated flow rate ^ F of the support gas when the support gas is supplied to the patient's airway at the calculated airway pressure ^ Paw. The observer 54 includes a comparator 56, an estimated flow rate calculator 57, a support gas volume calculator 58, and an alveolar pressure calculator 59.
[0060]
The comparator 56 receives the calculated airway pressure ^ Paw from the airway pressure calculator 55 and the calculated alveolar pressure ^ Palv from the alveolar pressure calculator 59. The comparator 56 subtracts the calculated alveolar pressure ^ Palv from the calculated airway pressure ^ Paw and gives the value to the estimated flow rate calculator 57. The calculated alveolar pressure ^ Palv will be described later.
[0061]
The estimated flow rate calculator 57 divides the subtracted value subtracted by the comparator 56 by a preset estimated airway resistance ^ R, and calculates the divided value as an estimated flow rate ^ F.
[0062]
The estimated airway resistance ^ R is a value obtained by estimating the airway resistance of the patient, and is set in advance by, for example, a medical staff. In addition, the estimated airway resistance R may be set in advance by a detection value detected by the measuring device. Also, as will be described later, in the overall system 14 of the present invention, the estimated airway resistance R does not have to be exactly matched to the actual patient's airway resistance R.
[0063]
The support gas volume calculator 58 sequentially accumulates the estimated flow rate ^ F calculated by the estimated flow rate calculator 57 from the support gas supply start time, and calculates the integrated value as the support gas volume ^ V. The assist gas volume calculator 58 is a so-called integrator. Hereinafter, the volume of the support gas calculated by the support gas volume calculator 58 is referred to as a calculation volume ^ V, and may be distinguished from the actual volume V of the support gas.
[0064]
The alveolar pressure calculator 59 multiplies the calculation volume {circumflex over (V)} by a preset lung estimated elastance {circumflex over (E)} E, and calculates the multiplication value as the calculated alveolar pressure ^ Palv. The alveolar pressure calculator 59 gives the calculated calculated alveolar pressure ^ Palv to the comparator 56. The calculated alveolar pressure ^ Palv is a value obtained by estimating the pressure in the alveoli, and may be referred to as distinguished from the actual alveolar pressure Palv.
[0065]
The estimated lung elastance ^ E is a value obtained by estimating the lung elastance of the patient, and is preset by, for example, a medical staff. Further, the estimated lung elastance ^ E may be set in advance by a detection value detected by a measuring instrument such as a ventilation mechanics inspection apparatus. Also, as will be described later, in the overall system 14 of the present invention, the lung elastance ^ E does not have to be exactly matched to the actual patient's lung elastance E.
[0066]
When the support gas flows through the airway, a pressure loss almost proportional to the flow rate F of the support gas occurs, and the pressure in the lung becomes lower than the airway pressure. The airway resistance R represents the relationship between the flow rate F of this support gas and the pressure loss. A value (F · R) obtained by multiplying the flow rate F of the support gas by the airway resistance R is a loss pressure caused by the resistance of the airway. For example, a general airway resistance R is 5 to 30 (cmH₂0) / (liter / second). However, the airway resistance R varies greatly depending on the patient's condition.
[0067]
When the support gas is supplied into the lung, the alveolar pressure Palv increases almost in proportion to the increase in the volume V of the support gas supplied into the lung. The lung elastance E represents the relationship between the volume V of the assisting gas and the alveolar pressure Palv. The value (V · E) obtained by multiplying the volume V of the support gas by the elastance of the lung is the alveolar pressure Palv. The alveolar pressure Palv is a pressure against the inflow of the support gas. For example, general lung elastance E is 1/20 to 1/50 (milliliter) / (cmH₂O). However, lung elastance E varies greatly depending on the patient's condition.
[0068]
Based on such characteristics of the respiratory tract, a respiratory organ model possessed by the observer 54 is set. The respiratory organ model possessed by the observer 54 is set to the following relationship.
[0069]
[Expression 1]

[0070]
That is, the respiratory organ model possessed by the observer 54 is a model of the respiratory organ of the patient when the spontaneous respiratory pressure Pmus is zero. In this model, the value obtained by subtracting the calculated alveolar pressure ^ Palv from the calculated airway pressure ^ Paw is equal to the value obtained by multiplying the estimated flow rate ^ F by the estimated airway resistance ^ R. Further, the value obtained by integrating the estimated flow rate ^ F from the support gas supply start time is equal to the calculation volume ^ V. The calculated alveolar pressure ^ Palv is equal to the value obtained by multiplying the estimated lung elastance ^ E by the calculated volume ^ V.
[0071]
Therefore, when the calculated airway pressure ^ Paw is an input value and the estimated flow rate ^ F is an output value, the transfer function of the observer 54 is expressed by the following equation.
[0072]
[Expression 2]

[0073]
Here, ^ R represents the estimated airway resistance, and ^ E represents the estimated elastance of the lung. The other expressions have the same meaning with respect to the symbols shown in the above expressions. The respiratory organ model possessed by such an observer 54 is an example of implementation, and may be another model obtained by modeling the respiratory organ of a patient.
[0074]
  The control amount calculation means 53 is a flow rate gain K that is a coefficient preset for the flow rate deviation ΔF calculated by the deviation calculation means 52._FGThe first operation value (K_FGΔF) and the flow rate deviation ΔF from the support gas supply start timeIntegral operationVolume gain K which is a coefficient preset to the value_VGThe second operation value (K_VGΔF / s) is obtained, and the first calculation value and the second calculation value are added to calculate the target pressure Pin related to the support pressure Pvent. When the flow rate deviation ΔF is an input value and the target pressure Pin is an output value, the transfer function of the control amount calculation means 53 is expressed by the following equation.
[0075]
[Equation 3]

[0076]
Where K_FGIndicates the flow gain and K_VGIndicates volume gain. The other expressions have the same meaning with respect to the symbols shown in the above expressions. For example, flow rate gain K_FGIs a flow amplification gain β determined in advance for the estimated airway resistance ^ R._FGIs set to the value multiplied by the volume gain K_VGIs set to a value obtained by multiplying the estimated lung elastance ^ E by a predetermined volume amplification gain. The flow rate amplification gain β_FGAnd volume amplification gain β_VGAre set to the same value, they are simply referred to as amplification gain β. Further, let B be the amplification gain β when ^ R = R and ^ E = E.
[0077]
The flow rate estimating means 51, the deviation calculating means 52, and the control amount calculating means 53 described above have been individually described for easy understanding, but they may be equivalently converted and arranged. Further, the flow rate estimating means 51, the deviation calculating means 52, and the control amount calculating means 53 may be realized by a computer capable of numerical calculation executing a predetermined operation program.
[0078]
In one embodiment of the present invention, the transfer function of the servo mechanism 20 includes a dead time element. FIG. 2 shows the transfer function Gc (s) excluding the dead time element in the transfer function of the servo mechanism 20, and the dead time element transfer function e.^{-Τ ・ s}And are shown separately. That is, when the target pressure Pin is an input value and the discharge pressure Pvent is an output value, the transfer function of the servo mechanism 20 is expressed by the following equation.
Gc (s) · e^{-Τ ・ s}                                    ... (6)
[0079]
Here, Gc (s) represents a transfer function of the servo mechanism 20 excluding the dead time element. E represents a natural logarithm, and τ represents a dead time required for the servo mechanism 20 to start adjusting the support pressure Pvent after the target pressure Pin is given. The other expressions have the same meaning with respect to the symbols shown in the above expressions.
[0080]
Further, in the actual patient's respiratory tract, the spontaneous breathing pressure Pmus is given in addition to the support pressure Pvent, which is different from the respiratory organ model of the observer 54. In the embodiment of the present invention, it is assumed that the pressure loss in the intake pipe is small and the support pressure Pvent that is the discharge pressure of the servo mechanism 20 is equal to the actual patient airway pressure Paw.
[0081]
FIG. 3 is a graph showing the relationship between the ventilation volume Vmus during spontaneous breathing and the ventilation volume Vast during assist breathing in the entire system 14 of the present invention. The support pressure Pvent is amplified with an amplification factor of (1 + B) times the spontaneous breathing pressure Pmus according to the time change of the spontaneous breathing pressure Pmus. Ventilation is equal to the volume of support gas flowing into the lungs. Here, B represents the amplification gain β when ^ R = R and ^ E = E as described above. The ventilation volume Vast during assist breathing by the ventilator 17 of the present embodiment is amplified to (1 + B) times the ventilation volume Vmus during spontaneous breathing.
[0082]
When the patient's condition is switched from the expiration period to the inspiration period, the patient operates respiratory muscles such as the diaphragm. As a result, the ventilation volume Vmus and the spontaneous breathing pressure Pmus during spontaneous breathing gradually increase with time, and gradually decrease when reaching a certain peak value P1. Then, the patient's condition is switched from the inspiration period to the expiration period.
[0083]
Normally, the patient's ventilation volume Vmus and spontaneous breathing pressure Pmus during spontaneous breathing during the inspiratory period first draw a gentle increasing curve as a waveform with respect to time, and then draw a rapid decreasing curve from the maximum value. However, depending on the patient's condition, the patient's ventilation volume Vmus and spontaneous breathing pressure Pmus vary greatly, and the peak value P1 and the inspiration period W1 vary.
[0084]
The servo mechanism 20 controlled by the control device 21 discharges the support gas at the support pressure Pvent so that the airway pressure Paw is proportionally amplified based on the amplification gain β predetermined for the spontaneous breathing pressure Pmus of the patient. For example, when the peak value P1 of the spontaneous breathing pressure Pmus is small and the inhalation period W1 is short, the support pressure Pvent is set so that the peak value P2 of the airway pressure Paw is small and the period W2 in which the support gas is supplied becomes short. Be controlled. Similarly, when the peak value P1 of the spontaneous breathing pressure Pmus is large and the inhalation period W1 is long, the support pressure Pvent is set such that the peak value P2 of the airway pressure Paw is large and the period W2 during which the support gas is supplied becomes long. Is controlled.
[0085]
As described above, according to the control device 21 of the present embodiment, the support pressure Pvent is determined based on the flow rate deviation ΔF. The detected flow rate F varies depending on the patient's spontaneous breathing pressure Pmus, but the estimated flow rate ^ F is not affected by the patient's spontaneous breathing pressure Pmus. Accordingly, the flow rate deviation ΔF is a value obtained by extracting a change in the spontaneous breathing pressure Pmus. That is, the spontaneous breathing pressure Pmus, which is normally difficult to detect, can be estimated, and can be configured as a disturbance observer when the spontaneous breathing pressure Pmus is regarded as a disturbance.
[0086]
In this way, by calculating the target pressure Pin according to the flow rate deviation ΔF related to the spontaneous breathing pressure Pmus, the assisting gas can be supplied to the patient at the assisting pressure Pvent that follows the spontaneous breathing pressure Pmus in almost real time. it can.
[0087]
FIG. 4 is a block diagram showing the entire system 14 of FIG. In the entire system 14 of the present invention, when the spontaneous breathing pressure Pmus is an input value and the sum of the spontaneous breathing pressure Pmus and the support pressure Pvent is an output value, the transfer function is expressed by the following equation.
[0088]
[Expression 4]

[0089]
Here, each symbol corresponds to the symbol described above.
Gc (s) = 1, ^ R = R, ^ E = E, K_FG= ^ R ・ B 、 K_VG= ^ E · B In the entire system 14 of the present invention, the pressure (Pmus + Pvent) obtained by adding the spontaneous breathing pressure Pmus and the support pressure Pvent is amplified to (1 + B) times the spontaneous breathing pressure Pmus. In other words, if airway resistance R and elastance E can be accurately estimated, amplification can always be performed as long as B is greater than zero.
[0090]
FIG. 5 is a block diagram showing equivalent conversion of the entire system 5 of the prior art shown in FIG. In the entire system 5 of the prior art, when the spontaneous breathing pressure Pmus is an input value and the sum of the spontaneous breathing pressure Pmus and the support pressure Pvent is an output value, the transfer function is expressed by the following equation.
[0091]
[Equation 5]

[0092]
Here, each symbol corresponds to the symbol described above.
Gc (s) = 1, ^ R = R, ^ E = E, K_fa= ^ R ・ A 、 K_va= ^ E · A, in the entire system 5 of the prior art, the pressure (Pmus + Pvent) obtained by adding the spontaneous breathing pressure Pmus and the support pressure Pvent is 1 / (1-A) times the spontaneous breathing pressure Pmus. Amplified. In this case, if A <0 or A> 1, the spontaneous breathing pressure Pmus cannot be amplified.
[0093]
As described above, the entire system 14 of the present invention can amplify as long as B is larger than 0, but in the entire system 5 of the prior art, A needs to satisfy 0 <A <1. Accordingly, the overall system 14 of the present invention can increase the degree of freedom in gain selection.
[0094]
Note that it is impossible for the estimated airway resistance ^ R and the estimated elastance ^ E to be exactly the same values as the actual airway resistance R and the actual elastance E. When there is a deviation in the values (^ R ≠ R, ^ E ≠ E), it is normal. Furthermore, it is normal (Gc (s) ≠ 1) that there is some delay in the servo mechanism for realizing the ventilator 17, and such a general case will be described later.
[0095]
In the conventional system and the entire system 5 or 14 of the present invention, the transient response may become unstable depending on factors such as changes in the patient's condition, setting errors of parameters, disturbances, and the like. In this case, excessive pressure of support gas may be supplied to the patient to cause a runaway.
[0096]
However, the overall system 14 of the present invention has a higher margin of stability than the overall system 5 of the prior art, as will be described later. Therefore, even if there is a change in the state of the patient, there is an error in setting each parameter, a large gain is set, or a disturbance or the like acts, it is possible to make it difficult to generate a runaway.
[0097]
FIG. 6 is a graph for explaining the stability of the model. A Nyquist diagram is used to determine the stability margin, that is, the difficulty of becoming unstable in the entire system.
[0098]
For example, as one method for determining the stability margin in the entire system, there is a method for determining the stability margin based on the distance ∇M where the vector locus 47 of the loop transfer function and the stability limit point L are closest to each other. This distance ∇M is referred to as a modulus margin ∇M. The modulus margin ∇M has a round transfer function of G_OLWhen represented by (jω), it is represented by the following equation. It is determined that the larger the system margin ∇M, the less the entire system becomes unstable.
∇M = | 1 + G_OL(jω) |_min                      ... (9)
[0099]
The modulus margin ∇My in the entire system 5 of the prior art is expressed by the following equation from the equation (1).
[0100]
[Formula 6]

[0101]
Further, the modulus margin ∇Mk in the entire system 14 according to the embodiment of the present invention is expressed by the following equation from the equation (8).
[0102]
[Expression 7]

[0103]
However, Equation (10) and Equation (11) are expressed as K_fa= ^ R ・ α 、 K_va= ^ E ・ α 、 K_FG= ^ R · β, K_VG= ^ E · β.
[0104]
Assuming that the estimation error of the respiratory system is the same between the case of the present invention and the conventional case, and ＾ R / R = ＥE / E = a, the modulus margin ∇M can be expressed by the following equation.
[0105]
[Equation 8]

[0106]
Here, a represents a deviation between the estimated value and the actual value, and α and β are values related to an amplification factor for amplifying the support pressure Pvent with respect to the spontaneous breathing pressure Pmus.
[0107]
The transfer function Gc (jω) of the servo mechanism 20 generally includes a first-order lag element and a dead time element. That is, the transfer function Gc (jω) of the servo mechanism 20 is shown in the following equation.
[0108]
[Equation 9]

[0109]
Here, Tc is a time constant of the servo mechanism 20. E represents the natural logarithm, and τ represents the dead time of the servo mechanism 20. A Nyquist diagram of the entire system when the servo mechanism 20 has such a transfer function is shown below.
[0110]
FIG. 7 shows a Nyquist diagram for the entire system 14 of the present invention when a <1. FIG. 8 also shows a Nyquist diagram of the entire prior art system 5 when a <1.
[0111]
From the equation (12), in the case of the entire system 5 of the prior art, α · a needs to satisfy 0 <α · a <1. In this case, the vector locus of the round transfer function has an element of the transfer function −Gc (jω) of the servo mechanism 20 having a positive feedback configuration (positive feedback). As a result, as shown in FIG. 8, the angular frequency ω is drawn from the negative region of the real axis toward the origin O as the angular frequency ω increases from the state where the angular frequency ω is zero. Therefore, the vector trajectory of the entire conventional system 5 is closest to the stability limit point L when the angular frequency ω is zero. As a result, even if the entire system 5 is stable, the modulus margin ∇My is small, and if the assist rate α is increased, it tends to become unstable.
[0112]
From the equation (13), in the case of the entire system 14 of the present invention, if a <1, the vector locus of the one-round transfer function is the transfer function Gc (jω) of the servo mechanism 20 in the negative feedback configuration (negative feedback). It has the element of. Therefore, as shown in FIG. 7, the angular frequency ω is drawn from the positive region of the real axis toward the origin O as the angular frequency ω increases from the state where the angular frequency ω is zero. Thus, the entire system 14 of the present invention is closest to the stability limit point L when the angular frequency ω is greater than zero. This increases the modulus margin ∇Mk of the entire system 14 of the present invention as compared to the conventional entire system 5. Therefore, even if the amplification gain β is increased, it is difficult to become unstable. More specifically, as long as it can be set as | ^ R · s + ^ E | <| R · s + E |, a negative feedback mechanism can always be configured.
[0113]
For example, the vector locus 48 of the entire system 14 of the present invention with β = 9 shown in FIG. 7 and the vector locus 49 of the entire system 5 of the prior art with α = 0.9 shown in FIG. Indicates the case of setting. As is apparent from FIGS. 7 and 8, the entire system 14 of the present invention has a larger modulus margin ∇M and a larger stability margin than the entire system 5 of the prior art.
[0114]
FIG. 9 shows the Nyquist diagram of the entire system 14 of the present invention when a> 1. FIG. 10 also shows a Nyquist diagram of the entire system 5 of the prior art when a> 1. The entire system 14 of the present invention has elements of a transfer function Gc (jω) having a positive feedback configuration when a> 1. Even in such a case, from the equations (12) and (13), the modulus margin ∇Mk of the entire system 14 of the present invention is larger than the modulus margin ∇My of the entire system of the prior art.
[0115]
For example, the vector locus 148 of the entire system 14 of the present invention with β = 5 shown in FIG. 9 and the vector locus 149 of the prior art system 5 with α = 0.8333 shown in FIG. 10 are set to the same amplification factor. Show the case. As is clear from FIGS. 9 and 10, even if a> 1, the entire system 14 of the present invention has a larger modulus margin ∇M and a larger stability margin than the entire system 5 of the prior art. I understand that.
[0116]
As described above, in the entire system 14 according to the embodiment of the present invention, the stability margin can be improved and it is difficult to become unstable. That is, even if the actual overall system differs from the overall system 14 simulated by the control device 21 due to setting errors of each parameter, patient state change, disturbance, etc., it is difficult to achieve a positive feedback configuration. Runaway can be prevented. As a result, a servo mechanism control method that further reduces the burden on the patient can be realized.
[0117]
Furthermore, even if the estimated airway resistance ^ R and the estimated elastance ^ E are slightly deviated from the actual airway resistance R and elastance E, the overall system has a large stability margin as described above. The system 14 is prevented from becoming unstable, and the assist pressure Pvent can be amplified by adjusting the amplification gain β. In particular, as described above, by satisfying a <1, that is, ^ R <R, ^ E <E, a negative feedback feedback configuration is always provided, and a larger stability margin can be set. Therefore, doctors confirm the patient's condition and set the estimated airway resistance ^ R and estimated elastance ^ E such that ^ R <R and ^ E <E. Even if not required, the possibility of runaway is reduced, and the servo mechanism can be controlled.
[0118]
For example, as an example that may actually occur, when the patient's airway is clogged, the actual airway resistance R is larger than the generally assumed airway resistance R. In such a case, the entire system of the present invention is more stable because R is larger than R.
[0119]
In the embodiment of the present invention, the assist pressure Pvent is determined based on the substantially real-time spontaneous breathing pressure Pmus by calculating the target pressure Pin according to the flow rate deviation ΔF. Therefore, even if the patient's spontaneous breathing pressure Pmus sequentially changes its waveform pattern, peak value, and generation period, it is possible to give the assist pressure Pvent based on the spontaneous breathing pressure Pmus at that time.
[0120]
As a result, the original purpose of the pressure-assisted ventilator for supplying the assist gas of the assist pressure Pvent, which is amplified in proportion to the spontaneous breathing pressure Pmus, to the patient's respiratory tract can be more reliably realized. In addition, it is possible to suitably realize the progress treatment for detaching the patient from the servo mechanism.
[0121]
FIG. 11 is a block diagram showing the airway pressure calculator 55. For example, the airway pressure calculator 55 has a servo mechanism model that is modeled by simulating the servo mechanism 20. In this case, when the target pressure Pin is given, the airway pressure calculator 55 calculates the discharge pressure that the servo mechanism 20 will discharge, that is, the support pressure Pvent. Specifically, the transfer function Gp (s) set in the airway pressure calculator 55 is a transfer function Gc (s) · e that simulates the transient characteristics of the servo mechanism 20.^{-Τ ・ s}Set to
[0122]
By setting the transfer function Gp (s) of the airway pressure calculator 55 in this manner, the calculated airway pressure ^ Paw that changes with time due to the transient characteristics of the servo mechanism 20 can be calculated.
[0123]
The airway pressure calculator 55 gives the calculated calculated airway pressure ^ Paw to the observer 54. The airway pressure calculator 55 can estimate the airway pressure ^ Paw more accurately by taking into consideration the pipe resistance of the intake pipe, the sampling time of the control device, the pressure propagation delay, and the like.
[0124]
FIG. 12 is a block diagram showing the control amount calculation means 53. The control amount calculator 53 includes a volume calculator 64, a flow rate calculator 65, a volume gain multiplier 66, a flow rate gain multiplier 67, a first adder 68, and a second adder 69.
[0125]
The volume calculator 64 calculates a volume calculation value (^ E · ΔF / s) obtained by multiplying the value obtained by sequentially integrating the flow rate deviation ΔF from the support gas supply start time by the estimated elastance ^ E. The volume gain multiplier 66 has a volume amplification gain β that is predetermined for the volume calculation value._VGAnd give the value to the first adder 68.
[0126]
  The flow rate calculator 65 calculates a flow rate calculation value (ΔF · ^ R) obtained by multiplying the flow rate deviation ΔF calculated by the deviation calculation means 52 by the estimated airway resistance ^ R. The flow rate gain multiplier 67 is a flow rate gain β that is predetermined for the flow rate calculation value._FGAnd give the value to the first adder 68. The first adder 68 isFlow rateAn operation value given from the gain multiplier 67;volumeThe calculated value given from the gain multiplier 66 is added and given to the servo mechanism 20 as the target pressure Pin. As a result, volume amplification gain β_VGAnd flow rate gain β_FGCan be individually set, and convenience can be improved.
[0127]
Also, as shown in FIG. 12, the flow rate calculator 64 and the volume calculator 65 give the calculation results to the second adder 69, respectively. The second adder 69 can add the calculation result of the flow rate calculator 64 and the calculation result of the volume calculator 65 and calculate the spontaneous breathing pressure ^ Pmus. Volume amplification gain β_VGAnd flow rate gain β_FGIs equal to β, the target pressure Pin amplified by (1 + β) times with respect to the calculated spontaneous breathing pressure ^ Pmus can be calculated.
[0128]
Further, the control device 21 may include a display unit 63. The display means 63 can acquire the spontaneous breathing pressure ^ Pmus calculated by the second adder 69 of the spontaneous breathing pressure estimator 61 and display the acquired spontaneous breathing pressure ^ Pmus. This allows a doctor or the like to check the spontaneous breathing pressure Pmus, which is a non-invasive, patient's breathing strength.
[0129]
FIG. 13 is a simulation result of the entire system 14 according to the embodiment of the present invention. FIG. 13 shows a simulation result when the control device 21 has the configuration shown in FIGS. 11 and 12. Further, when the estimated airway resistance ^ R and the estimated elastance ^ E are deviated from the actual airway resistance R and the actual elastance E (^ R ≠ R, ^ E ≠ E), and the servo mechanism 20 When the transfer function includes a first-order lag element and a time delay element (Gc (s) · e^{-Τ ・ s}≠ 1).
[0130]
[Table 1]

[0131]
Table 1 shows the set values of the parameters of the entire system in FIG. In the simulation, the spontaneous breathing pressure Pmus is set in a simulated manner, and when the set spontaneous breathing pressure Pmus is given, the support gas detection flow rate F, the support gas detection volume V, the estimated spontaneous breathing pressure ^ Pmus, and the support The time response of the pressure Pvent is obtained.
[0132]
FIG. 14 is a simulation result of the entire system 5 of the prior art. In FIG. 14, each parameter is set so as to correspond to FIG. That is, flow rate gain K_faIs set to ^ R · α and volume gain K_vaIs set to ^ E · α, and the amplification factor is set to 10 times. Other parameters are the same as those in FIG.
[0133]
As shown in FIG. 14, in the entire system 5 of the prior art, as a result of poor responsiveness, the flow rate F of the assisting gas does not become zero at the inspiration end time T1 when the inhalation period of the patient ends, and FIG. The support gas having a flow rate indicated by F1 in 2) is supplied to the patient's airway. That is, even after the patient's state is switched from the inhalation state to the exhalation state, the support gas is supplied to the patient, which is a so-called asynchronous state. Asynchronous state increases the burden on the patient.
[0134]
Further, as shown in FIGS. 14 (1) and 14 (4), the support pressure Pvent does not have a waveform proportional to the spontaneous breathing pressure Pmus due to the lack of rapid response of the entire system 5. The burden of.
[0135]
On the other hand, since the assist pressure Pvent is set according to the transient characteristic of the servo mechanism 20 in the entire system 14 of the present invention, as shown in FIG. 13 (2), the assist gas at the intake end time T1. The detected flow rate F becomes zero. That is, the entire system 14 of the present invention can be set or adjusted so as to reduce the possibility of an asynchronous state, and the support gas can be supplied only during the inspiration period according to the patient's spontaneous breathing pressure Pmus. .
[0136]
As shown in FIGS. 13 (1) and 13 (5), a support pressure Pvent proportional to the spontaneous breathing pressure Pmus can be applied. In the entire system 14 of the present invention, even when the amplification gain β is increased, the asynchronous state does not occur. Further, even if the estimated airway resistance ^ R and the estimated elastance ^ E have an error with respect to the actual estimated airway resistance R and elastance E, they are not in an asynchronous state. By preventing the asynchronous state in this way, the burden on the patient can be further reduced and artificial respiration can be performed.
[0137]
In this way, the asynchronous state can be canceled and the support pressure Pvent proportional to the spontaneous breathing pressure Pmus can be applied because the control device 21 uses the servo mechanism 20 and the model in which the delay of pressure transmission by the air circuit is predicted inward. This is because the estimated flow rate ^ F is estimated based on the model.
[0138]
FIG. 15 shows a simulation result when the amplification gain β is changed to 19 among the parameters shown in Table 1. As shown in FIG. 15, when the amplification gain β is extremely increased, in the entire system 14 of the present embodiment, the flow rate F of the support gas may become oscillating as shown in FIG. 15 (2). is there. Thus, it is not so preferable that the support pressure Pvent becomes oscillatory.
[0139]
FIG. 16 shows the flow rate gain gain β by the flow rate gain multiplier 67 from the state shown in FIG._FGIs a simulation result when the flow rate calculation value is reduced to 50%. When the flow rate F of the support gas is oscillating, the flow rate amplification gain β_FG, And giving the value to the first adder 68, it is possible to prevent the flow rate of the support gas from becoming oscillating as shown in FIG. As a result, the amplification factor can be amplified without causing the flow rate of the support gas to vibrate.
[0140]
Also, the flow amplification gain β_KGCorresponds to a proportional gain in PI control. Therefore, flow rate gain β_KGBy adjusting, the quick response to the spontaneous breathing pressure Pmus can be improved. Also, volume amplification gain β_VGCorresponds to an integral gain in PI control. Therefore volume amplification gain β_VGThe steady gain of the target pressure Pin can be adjusted by adjusting.
[0141]
Thus, volume amplification gain β_VGAnd flow rate gain β_FGBy adjusting this, it is possible to set the target pressure Pin by improving control characteristics such as quick response and attenuation in combination with the steady gain. As described above, the overall system 14 of the present invention has improved stability, so that the degree of freedom of parameter selection is large, and the amplification gain β or the volume amplification gain β_VGAnd flow rate gain β_FGEven if the change is made, runaway is unlikely to occur and can be adjusted appropriately.
[0142]
FIG. 17 is a block diagram illustrating an example of the ventilator 17. The control device 21 includes a control device main body 33 including a computer, a flow rate detection means 50, an input means 39, a display means 40, and a servo amplifier 48. The control device 21 may further include airway pressure detection means 61.
[0143]
The flow rate detection means 50 converts the flow rate of the gas flowing through the intake pipe 25 of the servo mechanism 20 into an electrical signal and gives the electrical signal to the control device main body 33. The input means 39 includes an estimated airway resistance R ^ and estimated elastance ^ E, an amplification gain β, a time constant Tc of the servo mechanism 20 and a dead time τ from a manager who manages the servomechanism 20 such as a doctor and a nurse. Entered. The input means 39 gives a signal indicating the input information to the control device body 33.
[0144]
The display means 40 is an informing means for informing the patient's airway pressure. Based on the display command signal received from the control device main body 33, the display means 40 displays a waveform indicating a temporal change in the spontaneous breathing pressure Pmus of the patient on the display screen.
[0145]
The servo amplifier 48 gives a signal indicating the target pressure Pin calculated by the control device main body 33 to the pump actuator 31. The pump actuator 31 controls the pump based on a signal indicating the target pressure Pin, and the discharge pressure of the servo mechanism 20 is feedback-controlled.
[0146]
The control device main body 33 includes an interface 101, a calculation unit 102, a temporary storage unit 103, and a storage unit 104. The interface 101 receives a signal from the connected flow rate detection means 50 and gives the signal to the calculation unit 102. The storage unit 104 stores a program to be executed by the control device main body 33, and the calculation unit 102 reads and executes the program stored in the storage unit 104, whereby the flow rate estimation unit 51, the deviation calculation unit 52, the control The quantity calculation means 53 can be realized. As a result, the control device main body 33 can control the servo mechanism 20 described above. The storage unit 104 may be a computer-readable recording medium such as a compact disc.
[0147]
The servomechanism 20 is not particularly limited as long as the control device 21 can control the pressure of the assist gas to be discharged, and the intake duct 25 that guides the assist gas to the patient's airway 15 is formed. For example, a ventilator having a bellows type pump as shown in FIG. 17 may be used. Moreover, the artificial respirator which supplies assistance gas via piping may be used.
[0148]
The transient characteristics of the servo mechanism 20 do not vary greatly compared to the patient's condition, and the characteristics can be obtained in advance. For example, a transfer function of the servo mechanism 20 is obtained in advance, and elements of the transfer function are set in the airway pressure calculator 55. Similarly, the detection delay of the flow rate detecting means is also obtained in advance and given to the control device 21.
[0149]
FIG. 18 is a block diagram showing the entire system 13 according to still another embodiment of the present invention. The entire system 13 illustrated in FIG. 18 has the same configuration as the entire system 14 illustrated in FIG. 2 except that a part of the configuration of the flow rate estimation unit 51 is different. Therefore, the description of the same configuration is omitted, and the reference numerals corresponding to the entire system 14 in FIG.
[0150]
The flow rate estimation means 51 further includes a detection delay calculator 60. The detection delay calculator 60 has a detection means model that is modeled by simulating the flow rate detection means 50. In this case, when the estimated flow rate {circumflex over (F)} is given from the observer 54, the detection delay computing unit 60 computes the estimated flow rate {circumflex over (F)} based on the detection delay of the flow rate detection means 50, and gives this computation result to the deviation computation means 52. . The deviation calculator 52 subtracts the detected flow rate detected by the flow rate detector 50 from the estimated flow rate F calculated by the detection delay calculator 60 to calculate a flow rate deviation ΔF. As a result, the flow rate estimating means 51 is supplied to the patient's airway based on the time change from the detection of the support gas flow F supplied to the patient's airway until the detection means 50 outputs the detection result. The flow rate {circumflex over (F)} of the assist gas to be calculated can be calculated with higher accuracy. Therefore, the asynchronous state can be more reliably prevented.
[0151]
  FIG. 19 shows the present invention.is connected withIt is a block diagram which shows the whole system | strain 12 of embodiment. The overall system 12 shown in FIG. 19 has the same configuration as the overall system 14 shown in FIG. 2 except that a part of the configuration of the flow rate estimation means 51 is different. Therefore, the description of the same configuration is omitted, and the reference numerals corresponding to the entire system 14 in FIG.
[0152]
Instead of the flow rate estimating means 51, a pressure detecting means 61 may be provided. The pressure detector 61 detects an airway pressure Paw, which is a pressure in the patient's airway. The pressure detecting means 61 gives the detected airway pressure Paw to the observer 54. Even if it does in this way, the effect mentioned above can be achieved. Further, by detecting the airway pressure Paw, the airway pressure Paw can be acquired without considering the influence of the delay of the servo mechanism 20, and the support gas is supplied to the patient with the support pressure Pvent that accurately corresponds to the spontaneous breathing pressure Pmus. Can be given to the airway.
[0153]
FIG. 20 is a block diagram showing the entire system 11 according to still another embodiment of the present invention. The entire system 11 shown in FIG. 20 is a model obtained by equivalently converting the transfer function Gp (s) set in the airway pressure calculator 55 of the entire system 14 shown in FIG. It is.
[0154]
[Expression 10]

[0155]
Here, γ is a coefficient set such that 0 <γ <1, and the other symbols correspond to the symbols described above. This transfer function shows the case where the actual airway resistance R and the actual lung elastance E are known. When the actual airway resistance R and the actual lung elastance E are unknown, the equation (15) Is substituted by:
[0156]
## EQU11 ##

[0157]
By setting the transfer function of the airway pressure calculator 55 in this way and setting it in the entire system 11 shown in FIG. 20, the delay due to the dead time element of the servo mechanism 20 is reduced by γ · Gc (s) · e.^{-Τ ・ s}Thus, the delay of the dead time of the servo mechanism can be preferably compensated.
[0158]
Thus, the airway pressure calculator 55 only needs to be able to estimate the airway pressure Paw based on the target pressure Pin. Therefore, the transfer function Gp (s) of the airway pressure calculator 55 may be set to other than the transfer function Gc (s) that simulates the characteristics of the servo mechanism 20.
[0159]
FIG. 21 is a block diagram showing the entire system 10 according to still another embodiment of the present invention. The entire system 10 shown in FIG. 21 is the same as the entire system 14 shown in FIG. 2 except that the estimated airway resistance ^ R and estimated elastance ^ E set in the flow rate estimating means 51 are different. It has a configuration. Therefore, the description of the same configuration is omitted, and the reference numerals corresponding to the entire system 14 in FIG.
[0160]
FIG. 22 is a graph for explaining the airway resistance R. When the flow of the assisting gas flowing in the airway becomes a laminar flow, the flow velocity changes linearly in proportion to the airway pressure Paw. However, in reality, the airway repeats branching and is not uniform in thickness, so the support gas flow becomes turbulent. Therefore, the estimated airway resistance ^ R is set in consideration of the turbulent resistance.
[0161]
Specifically, the estimated airway resistance ^ R set in the flow rate estimating means 51 is a first resistance coefficient ^ R set constant regardless of the flow rate of the support gas._TAnd a second resistance coefficient {circumflex over (K)} K based on the flow rate {circumflex over (F)} of the assist gas calculated by the estimated flow rate calculator._TIt is the value which added and. First resistance coefficient ^ R_TAnd the second resistance coefficient ^ K_TIs set to a coefficient corresponding to the patient's airway resistance. Further, the resistance of the entire respiratory organ including not only the lung but also the thorax may be set as the estimated airway resistance ^ R. The airway resistance R may be approximated by an estimated airway resistance ^ R expressed by another approximate expression.
[0162]
FIG. 23 is a graph for explaining lung compliance. The estimated elastance ^ E set in the flow rate estimating means 51 is a value based on the assist gas volume ^ V calculated by the assist gas volume calculator, and is the reciprocal of the lung compliance C. The compliance C increases nonlinearly with the increase in the volume V of the assist gas during the patient's inspiration period, and has a saturation characteristic and a hysteresis characteristic.
[0163]
The alveolar pressure calculator 59 obtains in advance information indicating the relationship between the compliance C and the volume of the support gas, so that the alveolar pressure Palv can be accurately calculated even when the case where the compliance is nonlinear is taken into consideration. It can be calculated.
[0164]
Thus, since the observer has a non-linear respiratory tube model, the estimated flow rate {circumflex over (F)} can be estimated more accurately. As a result, the spontaneous breathing pressure Pmus can be accurately estimated, and the support pressure Pvent can be determined according to the estimated spontaneous breathing pressure Pmus.
[0165]
The above-described embodiment of the present invention is an example of the present invention, and the configuration can be changed within the scope of the invention. For example, the above-described block diagram is merely an example of the present invention, and may be equivalently converted if the same effect can be obtained. When the computer to which the detected flow rate F is given calculates the target pressure Pin, each means may be realized by software.
[0166]
  Also estimated airway resistance^R and estimated elastance^E may be appropriately set by a doctor, but airway resistance R and elastance E measured in advance by a measuring instrument may be used. Further, airway resistance ^ R and elastance ^ E obtained by the estimation method disclosed in Patent Document 2 may be used.
[0170]
【The invention's effect】
  As aboveClaim1According to the described invention, the control amount calculation means determines the target pressure Pin based on the flow rate deviation ΔF that is a value representing the spontaneous breathing pressure Pmus. Therefore, it is possible to reliably realize the original purpose of the pressure-assisted ventilator that supplies the assist gas having the assist pressure Pvent proportionally amplified to the spontaneous breathing pressure Pmus to the patient's airway.
[0171]
  Furthermore, the margin for the stability limit of the entire system can be increased as compared with the prior art in which a gain is directly applied to the detected flow rate F of the assist gas to create a pressure information value. As a result, a system in which runaway is unlikely to occur can be configured. Therefore, even if the actual whole system fluctuates due to disturbances, time delays, or changes in the patient's respiratory system, it is difficult to achieve a positive feedback configuration and prevent runaway can do. By preventing the runaway, a servo mechanism control method that further reduces the burden on the patient can be realized.
  Further, by calculating the pressure ^ Paw in the patient's airway by the airway pressure calculator, the flow rate {circumflex over (F)} F of the support gas can be obtained with high accuracy. For example, the airway pressure calculator can prevent the asynchronous state in artificial respiration by calculating the pressure in the airway in consideration of the detection delay of the detected flow rate, the calculation delay of the control device, and the like. The respiratory organ model includes a comparator, an estimated flow rate calculator, and a support gas volume calculator, and an estimated patient airway resistance ^ R and elastance ^ E are set. Thus, the patient's respiratory organ model can be simulated and modeled.
[0172]
  And claims2According to the described invention, by calculating the flow rate F of the support gas according to the time characteristics of the servo mechanism, the flow rate F of the support gas can be calculated with higher accuracy. Further, it is possible to prevent the expiration period during which the patient sucks the support gas from the supply period during which the servo mechanism supplies the support gas to the patient's airway. A so-called asynchronous state can be prevented. This can further reduce the burden on the patient's breathing motion.
[0173]
  And claims3According to the described invention, the flow rate F of the support gas is calculated according to the time characteristic of the detection means. For example, in the flow rate estimation step, the flow rate {circumflex over (F)} F of the support gas is calculated based on the transfer function of the detection means considering the dead time element. As a result, the flow rate F of the support gas can be calculated with higher accuracy. Further, an asynchronous state caused by the detection delay of the detection means can be prevented. Accordingly, the burden on the patient's breathing motion can be further reduced.
[0174]
  And claims4According to the described invention, the flow rate deviation ΔF is set to the flow rate gain K as described above._FGAnd volume gain K_{V G} TheBy applying, the assist pressure Pvent proportional to the magnitude of the patient's spontaneous breathing pressure Pmus can be applied. As a result, the support gas can be supplied to the patient at the support pressure Pvent corresponding to the time change of the spontaneous breathing pressure, and the burden on the patient can be reduced.
[0175]
Each flow gain K_FGAnd volume gain K_VGBy adjusting this, it is possible to improve the quick response to the spontaneous breathing pressure Pmus and to prevent the support pressure Pvent from becoming oscillating.
[0177]
  And claims5According to the described invention, the estimated airway resistance R is set to a value obtained by adding the first resistance coefficient and the second resistance coefficient. The estimated elastance ^ E is set based on the assist gas volume ^ V calculated by the assist gas volume calculator. As a result, a model of the patient's respiratory system can be accurately simulated and modeled, and the support gas can be supplied at the support pressure Pvent that accurately corresponds to the patient's condition. As a result, the burden on the patient can be further reduced.
[0178]
  And claims6According to the described invention, the estimated airway resistance ^ R and the estimated elastance ^ E are set to be changeable during the control, thereby improving convenience. For example, by changing the estimated airway resistance ^ R and the estimated elastance ^ E according to the patient's condition, the type of servomechanism, etc., the patient's respiratory system model can be changed in the actual airway resistance and lung elastance. Can be followed.
[0179]
For example, when the change in the flow rate of the support gas becomes oscillating, the gain corresponding to the estimated airway resistance R can be reduced to prevent the change in the flow rate of the support gas from becoming oscillating. This can further reduce the burden on the patient.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a ventilator 17 and a patient 18 according to an embodiment of the present invention.
FIG. 2 is a block diagram specifically showing the entire system 14 according to the embodiment of the present invention.
FIG. 3 is a graph showing the relationship between the ventilation volume Vmus during spontaneous breathing and the ventilation volume Vast during assist breathing in the entire system 14 of the present invention.
4 is a block diagram showing the entire system 14 of the present invention of FIG.
5 is a block diagram showing equivalent conversion of the entire system 5 of the prior art shown in FIG. 24. FIG.
FIG. 6 is a graph for explaining the stability of a model.
FIG. 7 shows a Nyquist diagram for the entire system 14 of the present invention when a <1.
FIG. 8 shows a Nyquist diagram for the entire prior art system 5 when a <1.
FIG. 9 shows a Nyquist diagram for the entire system 14 of the present invention when a> 1.
FIG. 10 shows a Nyquist diagram of the entire system 5 of the prior art when a> 1.
11 is a block diagram showing an airway pressure calculator 55. FIG.
12 is a block diagram showing a control amount calculation means 53. FIG.
FIG. 13 is a simulation result of the entire system 14 according to the embodiment of the present invention.
FIG. 14 is a simulation result of the entire system 4 of the prior art.
15 is a simulation result when the amplification gain β is changed to 19 among the parameters shown in Table 1. FIG.
16 shows a flow rate amplification gain β from the state shown in FIG._FGIs a simulation result when the flow rate calculation value is reduced to 50%.
FIG. 17 is a block diagram showing an example of a ventilator 17;
FIG. 18 is a block diagram showing an entire system 13 according to still another embodiment of the present invention.
FIG. 19 shows the present invention.is connected withIt is a block diagram which shows the whole system | strain 12 of embodiment.
FIG. 20 is a block diagram showing an entire system 11 according to still another embodiment of the present invention.
FIG. 21 is a block diagram showing an entire system 10 according to still another embodiment of the present invention.
FIG. 22 is a graph for explaining airway resistance R;
FIG. 23 is a graph for explaining lung compliance;
FIG. 24 is a block diagram showing the entire system 5 including a prior art ventilator 1 and a patient 2.
FIG. 25 is a graph showing the relationship between the ventilation volume Vmus during spontaneous breathing and the ventilation volume Vast during assist breathing in the entire system 5 of the prior art.
[Explanation of symbols]
14 The whole system
17 Ventilator
20 Gas supply servo mechanism
21 Control device
50 Flow rate detection means
51 Flow rate estimation means
52 Deviation calculation means
53 Control amount calculation means
54 Observer
Pmus spontaneous breathing pressure
Pin target pressure
Pvent support pressure
F Actual support gas flow rate
^ F Estimated flow rate of support gas
ΔF Flow rate deviation

Claims (6)

Based on a target pressure Pin corresponding to a patient's spontaneous breathing pressure Pmus, a pressure-assisted ventilator that supplies a support gas containing oxygen to the patient's airway at a support pressure Pvent corresponding to the patient's spontaneous breathing pressure Pmus . A control device for controlling a gas supply servomechanism,
Flow rate detection means for detecting a flow rate F of support gas supplied to the patient's airway;
A flow rate estimation means for estimating a flow rate F of the support gas to be supplied to the patient's airway when the support gas is supplied at the support pressure Pvent when the respiratory organ model is formed by modeling the respiratory organ of the patient;
Deviation calculating means for calculating a flow deviation ΔF between the detected flow rate F and the estimated flow rate ^ F;
Control amount calculation means for calculating the target pressure Pin based on the flow rate deviation ΔF and giving the target pressure Pin to the servo mechanism;
The flow rate estimation means further includes an airway pressure calculator that calculates a pressure ^ Paw in the patient's airway based on the target pressure Pin calculated in the control amount calculation step,
The respiratory organ model includes a comparator that subtracts the alveolar pressure ^ Palv generated by the elastic restoring force of the lung from the pressure ^ Paw in the airway calculated by the airway pressure calculator when the spontaneous breathing pressure Pmus does not exist;
An estimated flow rate calculator for estimating the flow rate of support gas to be supplied to the patient's airway by dividing the subtracted value subtracted by the comparator by a preset estimated airway resistance ＲR;
Support for calculating the volume of support gas ^ V to be supplied to the patient's airway from the support gas supply start time by sequentially integrating the support gas flow rate ^ F calculated by the estimated flow rate calculator from the support gas supply start time A gas volume calculator,
An alveolar pressure calculator that calculates the alveolar pressure ^ Palv by multiplying the calculated volume of assisting gas ^ V by a preset estimated lung elastance ^ E, and supplies it to the subtractor. A control device for the gas supply servomechanism of the ventilator . The flow rate estimating means has a servo mechanism model that models the servo mechanism, and has a time characteristic from when the target pressure Pin is given to the servo mechanism until the servo mechanism supplies the support gas at the support pressure Pvent. 2. The control apparatus for a gas supply servomechanism of a ventilator according to claim 1, wherein the flow rate F of the support gas to be supplied to the patient's airway is estimated based on the flow rate. The flow rate estimation unit has a detection unit model obtained by modeling the flow rate detection unit, and a time characteristic from when the flow rate F of the support gas supplied to the patient's airway is detected until the detection unit outputs a detection result. 3. A control apparatus for a gas supply servomechanism of a ventilator according to claim 1, wherein the flow rate F of the support gas to be supplied to the patient's airway is estimated based on The control amount calculation means includes: a first calculation value obtained by multiplying the flow gain K _FG is set in advance in the flow rate difference [Delta] F, the second multiplied by the volume gain K _VG is set in advance in the integrated value of the flow rate difference [Delta] F Calculate the calculated value and
The control device for the gas supply servomechanism of the ventilator according to any one of claims 1 to 3, wherein the target pressure Pin is calculated by adding the first calculation value and the second calculation value . The estimated airway resistance {circumflex over (R)} is obtained by adding a first resistance coefficient set constant regardless of the flow rate of the support gas and a second resistance coefficient based on the flow rate ^ F of the support gas calculated by the estimated flow rate calculator. Value
The ventilator gas supply according to any one of claims 1 to 4 , wherein the estimated elastance ^ E is a value based on a support gas volume ^ V calculated by the support gas volume calculator. Servo mechanism control device. Change means for changing at least one of the estimated airway resistance ^ R and estimated elastance ^ E based on either the flow rate F of support gas supplied to the patient's airway or an input value input from the outside. controller of the ventilator of the gas supply servomechanism according to any one of claims 1 to 5, characterized in that it comprises.

Images