ロコモティブシンドローム定量評価のための腓腹筋の3次元画像解析

摘要

近年，運動器の障害により要介護リスクが高い状態を表すロコモティブシンドロームが提唱されている.そのため，簡便かつ定量的に筋肉を評価する方法が求められるが，その方法は確立されていない.本研究では，より定量的な評価を行うために腓腹筋の3次元超音波画像から羽状角を自動測定する方法を提案する.まず，上下の腱膜を曲面近似することにより決定し，それらの上下膜に挟まれた領域を筋繊維領域として抽出した.つぎに，筋繊維領域から筋繊維を抽出し，3次元フーリエ変換を行いてパワースペクトルを算出した.また，パワースぺクトルの周波数成分を大きい成分のみに限定し，その領域の慣性主軸の回転角を筋繊維の角度として求めた.最後に，下腱膜の曲面から接線の平均角度を算出し，筋繊維の角度と接線の平均角度を足し合わせることで羽状角を求めた.本手法で測定した羽状角と2次元超音波画像から測定した羽状角との比較実験を行った結果，0.79の相関係数を得た.%In recent years, A word to describe the state of requiring care and high-risk condition by musculoskeletal disorders as "Locomotive syndrome" has been proposed. Therefore, the methods for quantitative and easy evaluation of muscle have been required, but the method has not been established. In this study, we propose a method for automatic measurement of the pannation angle obtained by 3D-ultrasound images for the more quantitative evaluation of muscle. First, we determined the top and bottom aponeurosis by approximation of curvilinear surface and extracted the area between those top and bottom aponeurosis as the muscle-fiber area. Second, after we extracted muscle fiber from the muscle-fiber area, power spectrum was calculated by 3D-Fourier transform. In addition, we limited the frequency range of the power spectrum only to the large frequencies and considered the angle of rotation of principal axis of moment in the limited area as the angle of the muscle fiber. Finally, we calculated the mean angle of a line tangent to the curvilinear surface of the bottom aponeurosis and determined pannation angle by adding the mean angle of the tangential line and the angle of the muscle fiber. As a result, the correlation coefficient between the angles measured by this method and by the 2D-ultrasound images was 0.79.