心音図による小児先天性心疾患の自動検出:深層・手作り特徴の融合 Automated detection of pediatric congenital heart disease from phonocardiograms using deep and handcrafted feature fusion
AI要約 心音図(PCG)を用いて小児先天性心疾患を自動検出する手法を提案。深層学習による特徴と手作り特徴を融合することで、従来手法を上回る検出精度を達成した。
EN This paper proposes an automated method for detecting pediatric congenital heart disease from phonocardiograms by fusing deep learning-based features with handcrafted features, achieving improved detection accuracy over conventional approaches.