JP3864390B2 - EEG analysis method - Google Patents
EEG analysis method Download PDFInfo
- Publication number
- JP3864390B2 JP3864390B2 JP2000351125A JP2000351125A JP3864390B2 JP 3864390 B2 JP3864390 B2 JP 3864390B2 JP 2000351125 A JP2000351125 A JP 2000351125A JP 2000351125 A JP2000351125 A JP 2000351125A JP 3864390 B2 JP3864390 B2 JP 3864390B2
- Authority
- JP
- Japan
- Prior art keywords
- waveform
- time series
- electroencephalogram
- artifact
- electrocardiogram
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000004458 analytical method Methods 0.000 title claims description 13
- 238000000034 method Methods 0.000 claims description 51
- 238000012935 Averaging Methods 0.000 claims description 13
- 210000004556 brain Anatomy 0.000 claims description 7
- 230000001360 synchronised effect Effects 0.000 claims description 7
- 238000005259 measurement Methods 0.000 claims description 6
- 238000000537 electroencephalography Methods 0.000 description 22
- 238000010586 diagram Methods 0.000 description 12
- 238000010606 normalization Methods 0.000 description 9
- 238000012545 processing Methods 0.000 description 4
- 210000000624 ear auricle Anatomy 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000005684 electric field Effects 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 210000003128 head Anatomy 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 210000004761 scalp Anatomy 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000003925 brain function Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 230000000763 evoking effect Effects 0.000 description 1
- 210000002837 heart atrium Anatomy 0.000 description 1
- 210000001308 heart ventricle Anatomy 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
Images
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Description
【０００１】
【発明の属する技術分野】
この発明は、波形形状把握に基づいた心電図アーチファクト除去処理による脳波解析方法に関する。
【０００２】
【従来の技術】
脳機能の状態を調べる有効かつ簡便な方法として脳波記録方法があるが、一般的な脳波記録方法としては、耳朶に基準電極をおく導出法が用いられる。しかしながら、この方法では、耳朶にしばしば脳電位が混入する（耳朶活性化現象）ため、正確な電位マップ等を求める際の障害となる。そこで、脳電位の影響が及ばない頭部外に基準電極を設置する方法が採用されているが、この方法では、脳波よりも非常に大きい振幅（１０倍以上）を有する、心臓の鼓動によって生ずる電位、すなわち、心電図アーチファクト（脳波測定にとって雑音）が混入するため、この心電図アーチファクトの除去が重要となる。
【０００３】
従来の頭部外基準電極による脳波記録方法としては、人体の胸面と背面とに電極を設置し、２電極間に適当な抵抗を挟んだものを基準電極として用いる方法 「Y. Ishiyama, M. Ebe, I. Honma and Z. Abe: Elimination of EKG artifacts from EEG s recorded with balanced non-cephalic reference electrode method, Electroenceph. clin. Neurophysiol., 53, 662/665, 1982 」があるが、これは抵抗値の変動などで心電図アーチファクトを精度良く除去できない場合があった。また、脳波と同時に記録した心電図をトリガー信号として心電図Ｒ波による頂点同期加算平均によってアーチファクト形状を推定し除去する方法 ［M. Nakamura and H. Shibasaki: Elimination of EKG artifacts from EEG records; A New method of Non-cephalic Referential EEG Recording,Electroenceph. clin. Neurophysiol., 66, 88/92, 1987 ］があるが、個々の心電図アーチファクトの形状変動が無視できないので、除去精度が低いという問題点があった。
【０００４】
【発明が解決しようとする課題】
そこで、本発明は、上記課題に鑑み、高精度に心電図アーチファクトの除去を行うことができる、波形形状把握に基づいた心電図アーチファクト除去処理による脳波解析方法を提供することを目的とする。
【０００５】
【課題を解決するための手段】
上記目的を達成するために、本発明の脳波解析方法は、（ａ）記録された原脳波時系列とこの原脳波時系列に混入した心電図アーチファクト波形とから、原脳波時系列を心電図アーチファクト波形の周期毎に分割し、分割した各々の心電図アーチファクト波形の頂点時刻を同期させて加算平均し、この加算平均して求めた波形を電位０となる時刻によって要素波形に分割し、この要素波形毎に振幅及び持続時間を正規化して、原脳波時系列に混入した心電図アーチファクト波形の正規化標準波形を作成し、（ｂ）この心電図アーチファクト正規化標準波形を用いて正規化標準波形の相似波形を作成し、この相似波形と原脳波時系列との相関を計算し、この相関の最も高い相似波形の相似パラメータから波形変動特性を考慮した心電図アーチファクト推定時系列を作成し、（ｃ）原脳波時系列から心電図アーチファクト推定時系列を差し引き、脳波時系列成分のみを抽出することを特徴とする。
【０００６】
上記構成において、前記相似波形の相似パラメータは、好ましくは、前記分割した波形毎の各々の要素波形の、持続時間、前記頂点時刻及び振幅である。また、前記正規化標準波形は、好ましくは、心電図アーチファクト波形の測定毎にこの心電図アーチファクト波形の正規化標準波形を求め、この正規化標準波形を重み付きで既使用中の正規化標準波形に加算平均して更新する。
【０００７】
また、本発明の脳波解析方法は、（ａ）記録された原脳波時系列と脳波記録と同時記録した心電図時系列とを心電図波形周期毎に分割し、個々の心電図波形より０電位を基準として波形を分割し、それぞれの要素波形から振幅に関するパラメータと、持続に関するパラメータを抽出し、パラメータを用いて脳波時系列を正規化し、この脳波時系列を加算平均することによって心電図アーチファクト波形の正規化標準波形を作成し、（ｂ）この心電図アーチファクト波形の正規化標準波形を上記パラメータを用いて逆正規化して、波形変動特性を考慮した心電図アーチファクト推定時系列を作成し、（ｃ）原脳波時系列から心電図アーチファクト推定時系列を差し引き、脳波時系列成分のみを抽出することを特徴とする。
【０００８】
上記構成によれば、心電図アーチファクト波形の正規化標準波形を、原脳波時系列または同時記録した心電図時系列から求めるから、正確な波形形状把握が成された心電図アーチファクト正規化標準波形を得ることができる。
正確な心電図アーチファクト正規化標準波形を用い、要素波形毎に相似係数法または逆正規化法に基づいて原脳波時系列から心電図アーチファクト推定時系列を得るようにすれば、心電図アーチファクト個々の変動特性を含んだ正確な心電図アーチファクト推定時系列を得ることができる。
原脳波時系列から正確な心電図アーチファクト推定時系列を差し引くようにすれば、脳波成分のみを抽出することができる。
さらに、本発明によれば、正確な脳波測定が可能になると共に、これまでの脳波測定方法では実現できなかった遠隔電場誘発脳電位の正確な頭上マップの作成や脳波活動量（脳電位）の絶対値の測定も可能になる。
また、正規化標準波形を、心電図アーチファクト波形の測定毎に更新するので、実時間処理など逐次的に脳波データの処理を行う場合に、より高度な心電図アーチファクト除去が可能となる。
【０００９】
【発明の実施の形態】
以下、本発明の脳波解析方法の実施の形態を、図面に基づいて詳細に説明する。
脳波の記録は、国際１０−２０法に従って設置された頭皮上の電極と、脳電位の影響が及ばない左手に設置した基準電極との電位差を記録することで行った。
【００１０】
心電図アーチファクトの除去の全体手順は、（ａ）記録された原脳波時系列から、この原脳波時系列に混入した心電図アーチファクト波形の正規化標準波形を作成し、（ｂ）この心電図アーチファクト正規化標準波形を用いて、原脳波時系列から波形変動特性を伴った心電図アーチファクト推定時系列を作成し、（ｃ）心電図アーチファクト推定時系列を原脳波時系列から差し引き、脳波時系列成分のみを抽出する、段階から成る。
上記（ａ）、（ｂ）、（ｃ）の手順を以下に順番に説明する。
【００１１】
（ａ）正規化標準波形の作成
正規化標準波形の作成法には２つの方法がある。一つは、処理対象である脳波時系列の情報のみを用いた同期加算平均による方法、もう一つは、同時記録した心電図の情報を用いる方法である。
【００１２】
最初に、同期加算平均による方法を説明する。
図１は、本発明の同期加算平均法による正規化標準波形の作成方法を示す図である。
図１（Ａ）に示すように、頭部外基準電極法により記録した原脳波時系列ｙ（ｔ）は、心電図波形と脳波が重畳しているが、心電図波形が支配的であり、心電図アーチファクト波形の最大振幅時刻、すなわちＲ波頂点時刻を検出する。原脳波時系列ｙ（ｔ）のｋ個目の頂点時刻をｔp （ｋ）とし、その時刻より２００ｍｓさかのぼった時刻から、次の頂点時刻ｔP （ｋ＋１）より２００ｍｓ前までの波形を一つの心電図アーチファクト波形と定義し、図１（Ｂ）に示すように、これらを一つの頂点時刻に同期させてそろえ、次の加算平均式によって加算平均を行うことで脳波の影響を相殺し、図１（Ｃ）に示すように、心電図アーチファクト加算平均波形ｙ* （ｔ）を得る。
【００１３】
【数１】
続いて、得られた心電図アーチファクト加算平均波形の振幅と持続時間を規格化し、心電図アーチファクトの正規化標準波形を作成する。加算平均波形ｙ* （ｔ）を電位が０となる時刻によって分割し、それぞれの波形をＰ波、Ｑ波、Ｒ波、Ｓ波、Ｔ１波、Ｔ２波とする。分割された個々の波形に対して、振幅を１に、各要素波形の頂点時刻を境に前後の持続を１に正規化し、図１（Ｄ）に示すように、正規化標準波形ｘ（τ）を得る。
心電図アーチファクト波形は、心臓の心房、心室それぞれの動きに起因するため、要素波形毎に変動特性も異なる。そのため、心電図アーチファクト波形を電位０となる時刻によって要素波形に分割し、処理を行うことによって、高精度の心電図アーチファクト除去が可能になる。
【００１５】
次に、脳波記録と同時に記録した心電図の情報を用いる正規化標準波形の作成方法を説明する。
図２は、本発明の脳波記録と同時に記録した心電図の情報を用いる正規化標準波形の作成方法を示す図である。
図２（Ａ）に示すように、頭部外基準記録脳波時系列ｙ（ｔ）と同時記録した心電図ｕ（ｔ）より、個々の心電図波形を抜き出す。個々の心電図波形より、図２（Ｂ）に示すように、０電位を基準として波形を分割し、それぞれの要素波形から振幅に関するパラメータ〔Ｈp （ｋ），ＨQ （ｋ），ＨR （ｋ），ＨS （ｋ），ＨT1（ｋ），ＨT2（ｋ））と、持続に関するパラメータ（ＤP f （ｋ）， ＤQ f （ｋ），ＤR f （ｋ），ＤS f （ｋ），ＤT1 f （ｋ），ＤT2 f（ｋ）， ＤP b （ｋ），ＤQ b （ｋ），ＤR b （ｋ），ＤS b （ｋ），ＤT1 b （ｋ）， ＤT2 b （ｋ）〕を抽出する。図２（Ｃ）に示すように、これらのパラメータを用いて脳波時系列ｙ（ｔ）を正規化し、振幅と持続に関する変動を抑えた脳波時系列ｙ**（τ）を得る。この脳波時系列ｙ**（τ）を加算平均することによって、図（Ｄ）に示す正規化標準波形ｘ（τ）を得る。
【００１６】
（ｂ）心電図アーチファクト推定時系列の作成
次に、心電図アーチファクト推定時系列の作成について説明する。
個々の心電図アーチファクト波形は、その持続時間と振幅にばらつきが大きいため、それぞれの要素波形毎に変動特性を考慮した心電図アーチファクト推定時系列を作成することによって、心電図アーチファクト推定時系列の精度が向上する。アーチファクト推定時系列の作成方法には、相似係数を用いる方法と、心電図情報を用いた逆正規化による方法の二つがある。
【００１７】
最初に、相似係数を用いる方法を説明する。
図３は、本発明の相似係数を用いた心電図アーチファクト推定時系列の作成方法を示す図である。
相似係数を用いた方法は、持続の推定と振幅の推定という二つの段階を踏む。図３に示すように、始めに、正規化標準波形ｘ（τ）に対して相似係数を導入し、相似波形を作成する。相似波形は、分割したＰ波、Ｑ波、Ｒ波、Ｓ波、Ｔ１波、Ｔ２波のそれぞれの要素波形に対して作成する。相似波形の作成は、波形振幅が最も大きく、その変動も最も大きいＲ波部分から行い、続いてその両隣のＱ波とＳ波、Ｐ波とＴ１波、最後にＴ２波の順で行う。Ｒ波部分について説明すると、Ｒ波部分の正規化標準波形に対して、持続時間を変化させるための相似パラメータαR （ｋ）とβR （ｋ）、及びＲ波部分の頂点時刻ｌR （ｋ）を導入し、Ｒ波に対する相似波形ｓR （ｔ）を下記式によって作成する。
【００１８】
【数２】
この相似波形ｓR （ｔ）と原脳波時系列ｙ（ｔ）との間で相似係数を計算し、その時の相似パラメータαR （ｋ），βR （ｋ），ｌR （ｋ）を求める。相似係数の計算は、下記の式によって行う。
【００２０】
【数３】
上記式中、バーｓ及びバーｙを、明細書ではそれぞれｓm ，ｙm と表示する。ｓm ，ｙm はｙ（ｔ）とＳR （ｔ）それぞれの相似係数計算区間における平均値を表す。［数３］は、相似パラメータαR （ｋ），βR （ｋ），ｌR （ｋ）を変化させて脳波時系列との相関を計算し、その値が最も大きいものを求めることを意味している。［数３］を満たすときの相似波形をｓR （ｋ，ｔ）とする。
【００２２】
続いて、得られた相似波形ｓR （ｋ，ｔ）を用いて、振幅の推定を行う。振幅推定のためのパラメータａR （ｋ）を導入し、原脳波時系列ｙ（ｔ）と相似波形ｓR （ｋ，ｔ）の間の誤差が最も小さくなるように、即ち
【数４】
【数５】
【００２３】
同様の手順をＲ波部分の両隣にあるＱ波，Ｓ波に対しても行い、続いてＰ波，Ｔ１波部分、最後にＴ２の波部分の心電図アーチファクト推定波形を求める。Ｐ波，Ｑ波，Ｒ波，Ｓ波，Ｔ１波，Ｔ２波に対してそれぞれ求めた心電図アーチファクト推定波形を全ての時間軸上で足し合わせることで、ｋ個目の心電図アーチファクトに対する推定波形ｘ* （ｋ，ｔ）を得る。同様にして、全てのｋにおいてｘ* （ｋ，ｔ）を得ることによって、心電図アーチファクト推定時系列を作成する。
【００２４】
次に、心電図情報を用いた逆正規化による心電図アーチファクト推定時系列の作成方法を説明する。
図４は、本発明の心電図情報を用いた逆正規化による心電図アーチファクト推定時系列の作成方法を示す図である。
上記に説明した、正規化標準波形の作成で抽出した心電図アーチファクトの振幅に関するパラメータ〔Ｈp （ｋ），ＨQ （ｋ），ＨR （ｋ），ＨS （ｋ）， ＨT1（ｋ），ＨT2（ｋ）〕と、持続に関するパラメータ〔ＤP f （ｋ），ＤQ f （ｋ），ＤR f （ｋ），ＤS f （ｋ），ＤT1 f （ｋ），ＤT2 f （ｋ），ＤP b （ｋ），ＤQ b （ｋ），ＤR b （ｋ），ＤS b （ｋ），ＤT1 b （ｋ），ＤT2 b （ｋ）〕を用いて、図４に示したように正規化標準波形ｘ（τ）の持続と振幅を変化させる逆正規化の手順を踏むことで、ｋ個目の心電図アーチファクトに対する推定波形ｘ* （ｋ，ｔ）を得る。同様にして、全てのｋにおいてｘ* （ｋ，ｔ）を得ることによって、心電図アーチファクト推定時系列を作成する。
【００２５】
（ｃ）差し引き
以上の手順で作成した心電図アーチファクト推定時系列を原脳波時系列ｙ（ｔ）から
【数６】
【００２６】
図５は、本発明の脳波記録処理法によって心電図アーチファクトが除去された脳波データを示す図である。
図５（Ａ）は、波形形状把握を伴った本発明方法による結果で、図５（Ａ）において上から順に原脳波時系列ｙ（ｔ）、心電図アーチファクト推定波形ｘ* （ｋ，ｔ）、処理脳波時系列ｚ（ｔ）を表す。最下段の処理脳波時系列ｚ（ｔ）中からは心電図アーチファクトがほぼ完全に除去されているのが分かる。一方、図５（Ｂ）は、従来の単純同期加算平均による除去法による結果であり、特に心電図Ｒ波の部分に大きな残差が残っている。したがって、波形形状把握を伴った本発明方法によるアーチファクト除去精度が優れていることが分かる。
【００２７】
次に、正規化標準波形の更新方法について説明する。
全ての脳波データを一括して処理する方式においては必要ないが、実時間処理など逐次的に脳波データの処理を行いたい場合には、心電図アーチファクトの波形変動特性を考慮して正規化標準波形の形状を更新することで、より高度な心電図アーチファクト除去が可能となる。
図６は、本発明の正規化標準波形更新の概念を示す図である。
ｋ−１個目の心電図アーチファクト除去に用いた正規化標準波形をｘ（ｋ−１，τ）とし、これにｋ個目に検出された心電図アーチファクト波形の振幅と持続時間を規格化したものを重み付きの加算平均によって足しあわせることで、更新された心電図アーチファクトの正規化標準波形ｘ（ｋ，τ）を得る。
【００２８】
【発明の効果】
上記説明から理解されるように、本発明の脳波解析方法によれば、脳波時系列から心電図アーチファクトを高精度で除去することができ、高精度の脳波測定が可能になる。
また、本発明の脳波解析方法によれば、脳電位の絶対量の測定、遠隔電場誘発脳電位の正確な頭皮上マップの作成など、これまで困難とされてきた脳波解析が可能となる。
また、本発明の脳波解析方法はコンピュータのソフトウエアによって実現されるため、特別な装置を新たに製作する必要がなく、光磁気デスクなどの電子媒体に記録された脳波データさえあれば、脳波解析ができるという利点もある。
【図面の簡単な説明】
【図１】この発明の同期加算平均法による正規化標準波形の作成方法を示す波形図である。
【図２】本発明の脳波記録と同時に記録した心電図の情報を用いる正規化標準波形の作成方法を示す波形図である。
【図３】本発明の相似係数を用いた心電図アーチファクト推定時系列の作成方法を示す波形図である。
【図４】本発明の心電図情報を用いた逆正規化による心電図アーチファクト推定時系列の作成方法を示す波形図である。
【図５】本発明の脳波記録処理法によって心電図アーチファクトが除去された脳波データを示す波形図である。
【図６】本発明の正規化標準波形更新の概念を示す波形図である。
【符号の説明】
ｔp （ｋ） 心電図アーチファクトＲ波頂点時刻
ｙ（ｔ） 原脳波時系列
ｙ* （ｔ） 加算平均波形
ｙ* * （ｔ） 正規化した原脳波時系列
ｘ（τ） 正規化標準波形
ｘ* （ｋ，τ） 心電図アーチファクト推定波形
ｕ（ｔ） 心電図時系列
ｓ（ｋ、ｔ） 正規化相似波形
αT1（ｋ）、βT1（ｋ） 持続時間の相似パラメータ
ＨR （ｋ）、ＨT1（ｋ） 振幅の正規化パラメータ
ＤT1 f （ｋ）、ＤT1 b （ｋ） 持続時間の正規化パラメータ
Ｐ，Ｑ，Ｒ、Ｓ、Ｔ１、Ｔ２ 要素波形[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an electroencephalogram analysis method based on an electrocardiogram artifact removal process based on grasping a waveform shape.
[0002]
[Prior art]
There is an electroencephalogram recording method as an effective and simple method for examining the state of brain function. As a general electroencephalography method, a derivation method in which a reference electrode is placed on the earlobe is used. However, in this method, since a brain potential is often mixed in the earlobe (earlobe activation phenomenon), it becomes an obstacle to obtaining an accurate potential map or the like. Therefore, a method is adopted in which a reference electrode is placed outside the head where the influence of the brain potential does not reach. In this method, it is caused by the heartbeat having an amplitude (10 times or more) much larger than that of the electroencephalogram. Since an electric potential, that is, an electrocardiogram artifact (noise for electroencephalogram measurement) is mixed, removal of the electrocardiogram artifact is important.
[0003]
As a conventional method for recording an electroencephalogram using a reference electrode outside the head, a method in which electrodes are placed on the chest and back of a human body and an appropriate resistance is sandwiched between the two electrodes is used as a reference electrode [Y. Ishiyama, M Ebe, I. Honma and Z. Abe: Elimination of EKG artifacts from EEG s recorded with balanced non-cephalic reference electrode method, Electroenceph.clin. Neurophysiol., 53, 662/665, 1982 In some cases, ECG artifacts could not be accurately removed due to fluctuations in values. In addition, the method of estimating and removing artifact shape by vertex-synchronized averaging with ECG R-wave using ECG recorded simultaneously with EEG as trigger signal [M. Nakamura and H. Shibasaki: Elimination of EKG artifacts from EEG records; A New method of Non-cephalic Referential EEG Recording, Electroenceph. Clin. Neurophysiol., 66, 88/92, 1987], but there is a problem that the removal accuracy is low because the shape variation of each electrocardiogram artifact cannot be ignored.
[0004]
[Problems to be solved by the invention]
In view of the above problems, an object of the present invention is to provide an electroencephalogram analysis method by electrocardiogram artifact removal processing based on waveform shape grasping that can remove electrocardiogram artifacts with high accuracy.
[0005]
[Means for Solving the Problems]
In order to achieve the above object, an electroencephalogram analysis method according to the present invention comprises: (a) an original electroencephalogram time series is converted into an electrocardiogram artifact waveform from a recorded original electroencephalogram time series and an electrocardiogram artifact waveform mixed in the original electroencephalogram time series; It is divided for each period, and the vertex times of each of the divided electrocardiogram artifact waveforms are synchronized and averaged, and the waveform obtained by this averaging is divided into element waveforms according to the time when the potential becomes 0, and for each element waveform Normalize the amplitude and duration to create a normalized standard waveform of the ECG artifact waveform mixed in the original EEG time series. (B) Create a similar waveform of the normalized standard waveform using this ECG artifact normalized standard waveform. and, a correlation between this waveform similar and the original EEG time series calculated, electrocardiogram arch considering waveform variation characteristics from similar parameters of the highest similar waveforms of the correlation Create a Akuto estimated time series, subtract the ECG artifacts estimated time series from (c) raw EEG time series, and extracts only the EEG time series components.
[0006]
In the above configuration, the similarity parameter of the similarity waveform is preferably the duration, the vertex time, and the amplitude of each element waveform for each of the divided waveforms. Preferably, the normalized standard waveform is obtained as a normalized standard waveform of the electrocardiogram artifact waveform for each measurement of the electrocardiogram artifact waveform, and the normalized standard waveform is added to the normalized standard waveform in use with a weight. Update on average.
[0007]
In the electroencephalogram analysis method of the present invention, (a) the recorded original electroencephalogram time series and the electrocardiogram time series simultaneously recorded with the electroencephalogram recording are divided for each electrocardiogram waveform period, and each electrocardiogram waveform is based on 0 potential. Standardize ECG artifact waveform by dividing the waveform, extracting parameters related to amplitude and parameters related to duration from each element waveform, normalizing the EEG time series using the parameters, and averaging the EEG time series A waveform is created, (b) the normalized standard waveform of the ECG artifact waveform is denormalized using the above parameters, and an ECG artifact estimation time series in consideration of the waveform variation characteristics is created, and (c) the original brain wave time series It is characterized by subtracting an electrocardiogram artifact estimation time series from the above and extracting only an electroencephalogram time series component .
[0008]
According to the above configuration, since the normalized standard waveform of the electrocardiogram artifact waveform is obtained from the original electroencephalogram time series or the simultaneously recorded electrocardiogram time series, it is possible to obtain an electrocardiogram artifact normalized standard waveform with an accurate waveform shape grasp. it can.
By using an accurate ECG artifact normalized standard waveform and obtaining an ECG artifact estimated time series from the original EEG time series based on the similarity coefficient method or inverse normalization method for each element waveform, the fluctuation characteristics of each ECG artifact can be obtained. An accurate electrocardiogram artifact estimation time series can be obtained.
If an accurate electrocardiogram artifact estimation time series is subtracted from the original electroencephalogram time series, only the electroencephalogram component can be extracted.
Further, according to the present invention, accurate electroencephalogram measurement becomes possible, and an accurate overhead map of a remote electric field evoked electroencephalogram that could not be realized by the conventional electroencephalogram measurement methods and the electroencephalogram activity (electroencephalogram) The absolute value can also be measured.
In addition, since the normalized standard waveform is updated every time the electrocardiogram artifact waveform is measured, more advanced electrocardiogram artifact removal is possible when the electroencephalogram data is sequentially processed such as real-time processing.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of an electroencephalogram analysis method of the present invention will be described below in detail with reference to the drawings.
The electroencephalogram was recorded by recording the potential difference between the electrode on the scalp placed according to the International 10-20 method and the reference electrode placed on the left hand where the influence of brain potential does not reach.
[0010]
The entire procedure for removing the electrocardiogram artifact is as follows: (a) a normalized standard waveform of the electrocardiogram artifact waveform mixed in the original electroencephalogram time series is created from the recorded original electroencephalogram time series, and (b) the electrocardiogram artifact normalization standard. Using the waveform, create an ECG artifact estimation time series with waveform fluctuation characteristics from the original EEG time series, (c) subtract the ECG artifact estimation time series from the original EEG time series, and extract only the EEG time series components. It consists of stages.
The procedures (a), (b), and (c) will be described in order below.
[0011]
(A) Creation of normalized standard waveform There are two methods for creating a normalized standard waveform. One is a method using synchronous addition averaging using only the information of the electroencephalogram time series to be processed, and the other is a method using the information of the electrocardiogram recorded simultaneously.
[0012]
First, a method based on synchronous addition averaging will be described.
FIG. 1 is a diagram showing a method for creating a normalized standard waveform by the synchronous addition averaging method of the present invention.
As shown in FIG. 1 (A), the electroencephalogram time series y (t) recorded by the out-of-head reference electrode method superimposes an electrocardiogram waveform and an electroencephalogram, but the electrocardiogram waveform is dominant, and an electrocardiogram artifact. The maximum amplitude time of the waveform, that is, the R wave peak time is detected. The k-th apex time of the original EEG time series y (t) is t p (k), and the waveform from the time 200 ms before that time to 200 ms before the next apex time t P (k + 1) As shown in FIG. 1 (B), they are defined as an electrocardiogram artifact waveform, and these are arranged in synchronization with one vertex time, and the effect of the electroencephalogram is canceled by performing addition averaging using the following addition averaging formula. As shown in (C), an electrocardiogram artifact addition average waveform y * (t) is obtained.
[0013]
[Expression 1]
Subsequently, the amplitude and duration of the obtained electrocardiogram artifact addition average waveform are normalized, and a normalized standard waveform of the electrocardiogram artifact is created. The added average waveform y * (t) is divided according to the time when the potential becomes 0, and the respective waveforms are defined as P wave, Q wave, R wave, S wave, T1 wave, and T2 wave. For each of the divided waveforms, the amplitude is normalized to 1 and the duration before and after the vertex time of each element waveform is normalized to 1. As shown in FIG. 1D, the normalized standard waveform x (τ )
Since the electrocardiogram artifact waveform is caused by the movement of the heart atria and ventricles, the fluctuation characteristics are different for each element waveform. Therefore, the electrocardiogram artifact waveform can be removed with high accuracy by dividing the electrocardiogram artifact waveform into element waveforms according to the time when the potential becomes 0 and performing processing.
[0015]
Next, a method for creating a normalized standard waveform using the electrocardiogram information recorded simultaneously with the electroencephalogram recording will be described.
FIG. 2 is a diagram showing a method for creating a normalized standard waveform using the electrocardiogram information recorded simultaneously with the electroencephalogram recording of the present invention.
As shown in FIG. 2A, individual electrocardiogram waveforms are extracted from the electrocardiogram u (t) recorded simultaneously with the extra-head reference recording brain wave time series y (t). As shown in FIG. 2B, the waveform is divided from the individual electrocardiogram waveforms on the basis of the zero potential, and parameters related to the amplitude [H p (k), H Q (k), H R ( k), H S (k), H T1 (k), H T2 (k)) and parameters relating to persistence (D P f (k), D Q f (k), D R f (k), D S f (k), D T1 f (k), D T2 f (k), D P b (k), D Q b (k), D R b (k), D S b (k), D T1 b (K), D T2 b (k)] As shown in Fig. 2 (C), these parameters were used to normalize the electroencephalogram time series y (t) to suppress variations in amplitude and duration. An electroencephalogram time series y ** (τ) is obtained, and by normalizing the electroencephalogram time series y ** (τ), a normalized standard waveform x (τ) shown in FIG.
[0016]
(B) Creation of electrocardiogram artifact estimation time series Next, creation of an electrocardiogram artifact estimation time series will be described.
Since each ECG artifact waveform has large variations in duration and amplitude, the accuracy of the ECG artifact estimation time series is improved by creating an ECG artifact estimation time series that considers the fluctuation characteristics for each element waveform. . There are two methods for creating an artifact estimation time series: a method using similarity coefficients and a method using inverse normalization using electrocardiogram information.
[0017]
First, a method using a similarity coefficient will be described.
FIG. 3 is a diagram showing a method for creating an electrocardiogram artifact estimation time series using the similarity coefficient of the present invention.
The method using the similarity coefficient takes two steps: estimation of duration and estimation of amplitude. As shown in FIG. 3, first, a similarity coefficient is introduced into the normalized standard waveform x (τ) to create a similar waveform. Similar waveforms are created for the respective element waveforms of the divided P wave, Q wave, R wave, S wave, T1 wave, and T2 wave. The similar waveform is created from the R wave portion having the largest waveform amplitude and the largest fluctuation, followed by the adjacent Q wave and S wave, P wave and T1 wave, and finally the T2 wave. The R wave portion will be described. Similar parameters α R (k) and β R (k) for changing the duration of the normalized standard waveform of the R wave portion, and the apex time l R (R wave portion) k) is introduced, and a similar waveform s R (t) for the R wave is created by the following equation.
[0018]
[Expression 2]
A similarity coefficient is calculated between the similar waveform s R (t) and the original electroencephalogram time series y (t), and the similarity parameters α R (k), β R (k), l R (k) at that time are obtained. . The similarity coefficient is calculated according to the following formula.
[0020]
[Equation 3]
In the above formula, the bar s and bars y, respectively in the specification s m, indicated as y m. s m, y m represents the average value in y (t) and S R (t) each similarity factor calculation interval. [Equation 3] means that the correlation parameter α R (k), β R (k), l R (k) is changed to calculate the correlation with the electroencephalogram time series, and the one having the largest value is obtained. is doing. Let s R (k, t) be a similar waveform when [Equation 3] is satisfied.
[0022]
Subsequently, the amplitude is estimated using the obtained similar waveform s R (k, t). The parameter a R (k) for amplitude estimation is introduced, and the error between the original EEG time series y (t) and the similar waveform s R (k, t) is minimized, that is,
[0023]
The same procedure is performed for the Q wave and S wave adjacent to the R wave part, and then the electrocardiogram artifact estimation waveform of the P wave, T1 wave part, and finally the T2 wave part is obtained. The estimated waveform x * for the kth ECG artifact is obtained by adding the estimated ECG artifact waveforms obtained for the P wave, Q wave, R wave, S wave, T1 wave, and T2 wave on all the time axes . (K, t) is obtained. Similarly, an electrocardiogram artifact estimation time series is created by obtaining x * (k, t) for all k.
[0024]
Next, a method for creating an ECG artifact estimation time series by denormalization using ECG information will be described.
FIG. 4 is a diagram showing a method for creating an electrocardiogram artifact estimation time series by inverse normalization using electrocardiogram information according to the present invention.
Parameters related to the amplitude of the electrocardiogram artifacts extracted by generating the normalized standard waveform described above [H p (k), H Q (k), H R (k), H S (k), H T1 (k) , H T2 (k)] and parameters [D P f (k), D Q f (k), D R f (k), D S f (k), D T1 f (k), D T2 f (k), D P b (k), D Q b (k), D R b (k), D S b (k), D T1 b (k), D T2 b (k)] 4, the estimated waveform x * (k, t) for the kth electrocardiogram artifact is obtained by performing the denormalization procedure for changing the duration and amplitude of the normalized standard waveform x (τ) as shown in FIG. obtain. Similarly, an electrocardiogram artifact estimation time series is created by obtaining x * (k, t) for all k.
[0025]
(C) The ECG artifact estimation time series created by the above procedure is subtracted from the original EEG time series y (t).
[0026]
FIG. 5 is a diagram showing electroencephalogram data from which ECG artifacts are removed by the electroencephalogram recording processing method of the present invention.
FIG. 5 (A) shows the result of the method of the present invention with grasping the waveform shape. In FIG. 5 (A), the electroencephalogram time series y (t), the electrocardiogram artifact estimated waveform x * (k, t), The processed electroencephalogram time series z (t) is represented. It can be seen that the electrocardiogram artifacts are almost completely removed from the lowermost processed electroencephalogram time series z (t). On the other hand, FIG. 5B shows the result obtained by the conventional removal method based on simple synchronous addition averaging, and a large residual remains particularly in the electrocardiogram R wave portion. Therefore, it can be seen that the artifact removal accuracy by the method of the present invention accompanied by the waveform shape grasping is excellent.
[0027]
Next, a method for updating the normalized standard waveform will be described.
It is not necessary in the method that processes all EEG data at once, but if you want to process EEG data sequentially, such as real-time processing, the normalized standard waveform is taken into account by considering the waveform variation characteristics of ECG artifacts. By updating the shape, more advanced ECG artifact removal is possible.
FIG. 6 is a diagram showing the concept of normalization standard waveform update according to the present invention.
The normalized standard waveform used for the removal of the k-1th ECG artifact is x (k-1, τ), and the normalized amplitude and duration of the kth ECG artifact waveform are normalized to this. The normalized standard waveform x (k, τ) of the updated electrocardiogram artifact is obtained by adding the weighted averages.
[0028]
【The invention's effect】
As understood from the above description, according to the electroencephalogram analysis method of the present invention, electrocardiogram artifacts can be removed from the electroencephalogram time series with high accuracy, and electroencephalogram measurement with high accuracy becomes possible.
Further, according to the electroencephalogram analysis method of the present invention, it is possible to perform electroencephalogram analysis that has been difficult until now, such as measurement of the absolute amount of electroencephalogram and creation of an accurate scalp map of the electroencephalogram induced by remote electric field.
Further, since the electroencephalogram analysis method of the present invention is realized by computer software, there is no need to newly create a special device, and electroencephalogram analysis is possible as long as there is electroencephalogram data recorded on an electronic medium such as a magneto-optical desk. There is also an advantage of being able to.
[Brief description of the drawings]
FIG. 1 is a waveform diagram showing a method for creating a normalized standard waveform by the synchronous addition averaging method of the present invention.
FIG. 2 is a waveform diagram showing a method for creating a normalized standard waveform using information of an electrocardiogram recorded simultaneously with the electroencephalogram recording of the present invention.
FIG. 3 is a waveform diagram showing a method for creating an electrocardiogram artifact estimation time series using similarity coefficients according to the present invention.
FIG. 4 is a waveform diagram showing a method for creating an electrocardiogram artifact estimation time series by inverse normalization using electrocardiogram information according to the present invention.
FIG. 5 is a waveform diagram showing electroencephalogram data from which electrocardiogram artifacts have been removed by the electroencephalogram recording processing method of the present invention.
FIG. 6 is a waveform diagram showing the concept of normalization standard waveform update according to the present invention.
[Explanation of symbols]
t p (k) ECG artifact R wave peak time y (t) Progenitor EEG time series y * (t) Summed average waveform y * * (t) Normalized EEG time series x (τ) Normalized standard waveform x * (K, τ) ECG artifact estimated waveform u (t) ECG time series s (k, t) Normalized similarity waveform α T1 (k), β T1 (k) Duration similarity parameter H R (k), H T1 (K) Amplitude normalization parameter D T1 f (k), D T1 b (k) Duration normalization parameter P, Q, R, S, T1, T2 Element waveform
Claims (4)
（ｂ）この心電図アーチファクト正規化標準波形を用いて、該正規化標準波形の相似波形を作成し、この相似波形と上記原脳波時系列との相関を計算し、この相関の最も高い相似波形の相似パラメータから、波形変動特性を考慮した心電図アーチファクト推定時系列を作成し、
（ｃ）上記原脳波時系列から上記心電図アーチファクト推定時系列を差し引き、脳波時系列成分のみを抽出することを特徴とする、脳波解析方法。 (A) From the recorded original electroencephalogram time series and the electrocardiogram artifact waveform mixed in the original electroencephalogram time series, the original electroencephalogram time series is divided for each period of the electrocardiogram artifact waveform, and each of the divided electrocardiogram artifact waveforms is The average time is synchronized with the vertex time, and the waveform obtained by the averaging is divided into element waveforms according to the time when the potential is 0, and the amplitude and duration are normalized for each element waveform, and the original brain wave time series is obtained. Create a normalized standard waveform of the ECG artifact waveform mixed in
(B) Using this electrocardiogram artifact normalized standard waveform, create a similar waveform of the normalized standard waveform, calculate the correlation between the similar waveform and the original brain wave time series, and calculate the correlation waveform having the highest correlation. From the similar parameters, create an ECG artifact estimation time series considering the waveform fluctuation characteristics,
(C) A method for analyzing an electroencephalogram, wherein the electrocardiogram artifact estimation time series is subtracted from the original electroencephalogram time series to extract only electroencephalogram time series components.
（ｂ）上記心電図アーチファクト波形の正規化標準波形を上記パラメータを用いて逆正規化して、波形変動特性を考慮した心電図アーチファクト推定時系列を作成し、
（ｃ）上記原脳波時系列から上記心電図アーチファクト推定時系列を差し引き、脳波時系列成分のみを抽出することを特徴とする、脳波記録の解析方法。 (A) The recorded original electroencephalogram time series and the electrocardiogram time series recorded simultaneously with the electroencephalogram recording are divided for each electrocardiogram waveform period, and the waveform is divided from each electrocardiogram waveform with reference to 0 potential, and each element waveform A parameter related to amplitude and a parameter related to duration are extracted from the above, and the EEG time series is normalized using the parameters, and a normalized standard waveform of the electrocardiogram artifact waveform is created by averaging the EEG time series.
(B) Normalizing the normalized ECG artifact waveform using the above parameters, and denormalizing to create an ECG artifact estimation time series in consideration of the waveform variation characteristics;
(C) A method for analyzing an electroencephalogram recording, wherein the electrocardiogram artifact estimation time series is subtracted from the original electroencephalogram time series to extract only an electroencephalogram time series component .
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2000351125A JP3864390B2 (en) | 2000-11-17 | 2000-11-17 | EEG analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2000351125A JP3864390B2 (en) | 2000-11-17 | 2000-11-17 | EEG analysis method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JP2002153435A JP2002153435A (en) | 2002-05-28 |
| JP3864390B2 true JP3864390B2 (en) | 2006-12-27 |
Family
ID=18824239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2000351125A Expired - Fee Related JP3864390B2 (en) | 2000-11-17 | 2000-11-17 | EEG analysis method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP3864390B2 (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5061131B2 (en) * | 2009-01-29 | 2012-10-31 | 富士フイルムＲｉファーマ株式会社 | Organ surface image display apparatus and method |
| JP5419157B2 (en) * | 2009-11-30 | 2014-02-19 | 株式会社心電技術研究所 | EEG measurement apparatus and EEG measurement method |
| DE102014205503A1 (en) * | 2014-03-25 | 2015-10-01 | Hamm Ag | Method for correcting a measured value profile by eliminating periodically occurring measuring artifacts, in particular in the case of a soil compactor |
| EP3381354A1 (en) | 2017-03-31 | 2018-10-03 | Koninklijke Philips N.V. | Methods and system for processing an emg signal |
| JP7024151B2 (en) * | 2017-10-20 | 2022-02-24 | Ｃｏｎｎｅｃｔ株式会社 | EEG measurement system, EEG measurement method, program, and non-temporary recording medium |
| CN112183331B (en) * | 2020-09-28 | 2024-04-16 | 南京伟思医疗科技股份有限公司 | System and method for identifying electrocardiographic artifact of neonatal brain electrical signal |
-
2000
- 2000-11-17 JP JP2000351125A patent/JP3864390B2/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| JP2002153435A (en) | 2002-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fariha et al. | Analysis of Pan-Tompkins algorithm performance with noisy ECG signals | |
| Rieta et al. | Atrial activity extraction for atrial fibrillation analysis using blind source separation | |
| US8055333B2 (en) | Device and method for detecting cardiac impairments | |
| AU2006308467B2 (en) | Method and system for high-resolution extraction of quasi-periodic signals | |
| Arunachalam et al. | Real-time estimation of the ECG-derived respiration (EDR) signal using a new algorithm for baseline wander noise removal | |
| WO2007021645A2 (en) | Method and system for limiting interference in electroencephalographic signals | |
| US5511554A (en) | Real-time artifact removal from waveforms using a dynamic filter having a fixed delay | |
| Butt et al. | Denoising practices for electrocardiographic (ECG) signals: a survey | |
| Xie et al. | Heart rate estimation from ballistocardiography based on hilbert transform and phase vocoder | |
| WO2009110051A1 (en) | On-line vital signal processing method effective during fetal movement and its apparatus | |
| Barbosa et al. | Reduction of electromyographic noise in the signal-averaged electrocardiogram by spectral decomposition | |
| JP3864390B2 (en) | EEG analysis method | |
| Colak | Preprocessing effects in time–frequency distributions and spectral analysis of heart rate variability | |
| Sebastian et al. | Wavelet based denoising for suppression of respiratory and motion artifacts in impedance cardiography | |
| Lemay et al. | Spatiotemporal QRST cancellation method using separate QRS and T-waves templates | |
| Burke et al. | Wavelet based analysis and characterization of the ECG signal | |
| CN115828056A (en) | Electrocardiogram characteristic signal extraction method and terminal | |
| Sahoo et al. | Adaptive thresholding based EMD for delineation of QRS complex in ECG signal analysis | |
| Tun et al. | Analysis of computer aided identification system for ECG characteristic points | |
| JPH10262941A (en) | Electroencephalogram processor | |
| Burke et al. | ECG analysis using the Mexican-Hat wavelet | |
| Vega-Martínez et al. | Wavelet packet based algorithm for QRS region detection and R/S wave identification | |
| CN112826451A (en) | Anesthesia depth and sleep depth assessment method and device | |
| Govreen-Segal et al. | Real-time PC-based system for dynamic beat-to-beat QT-RR analysis | |
| Gautam et al. | Analysis of HRV signal for disease diagnosis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A711 | Notification of change in applicant |
Free format text: JAPANESE INTERMEDIATE CODE: A712 Effective date: 20031031 |
|
| RD03 | Notification of appointment of power of attorney |
Free format text: JAPANESE INTERMEDIATE CODE: A7423 Effective date: 20040129 |
|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20040419 |
|
| A977 | Report on retrieval |
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20060609 |
|
| A131 | Notification of reasons for refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20060620 |
|
| A521 | Request for written amendment filed |
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20060818 |
|
| TRDD | Decision of grant or rejection written | ||
| A01 | Written decision to grant a patent or to grant a registration (utility model) |
Free format text: JAPANESE INTERMEDIATE CODE: A01 Effective date: 20060919 |
|
| A61 | First payment of annual fees (during grant procedure) |
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20060922 |
|
| R150 | Certificate of patent or registration of utility model |
Free format text: JAPANESE INTERMEDIATE CODE: R150 |
|
| LAPS | Cancellation because of no payment of annual fees |