PROBLEM TO BE SOLVED: To reduce accuracy deterioration of a heart beat feature amount originating from measurement abnormality.SOLUTION: A state estimation apparatus in the embodiment has: extraction means which extracts electrocardiographic R waves on the basis of a signal outputted by electrocardiographic measurement means measuring a subjects electrocardiograph calculation means which calculates the instantaneous heart rate being an interval between time-serially adjacent two R waves extracted by the extraction means evaluation means which evaluates a measurement state of the instantaneous heart rate calculated by the calculation means, which is the interval between the R waves in accordance with whether or not respective measurement states of the adjacent two R waves extracted by the extraction means are normal classification means which classifies the instantaneous heart rate evaluated by the evaluation means into one of a plurality of kinds of classes in accordance with the kind of heart beat feature amount according to the kind of evaluation and selection means which selects the instantaneous heart rate classified into a class to which a desired kind of heart beat feature amount corresponds among instantaneous heart rates of respective classes classified by the classification means.SELECTED DRAWING: Figure 1COPYRIGHT: (C)2018,JPO&INPIT【課題】計測異常に由来する心拍特徴量の精度低下を低減する。【解決手段】実施形態における状態推定装置は、被験者の心電を計測する心電計測手段から出力される信号に基づいて心電のR波を抽出する抽出手段と、前記抽出手段により抽出した、時系列で隣接する2つのR波の間隔である瞬時心拍を算出する算出手段と、前記抽出手段により抽出した、前記隣接する2つのR波のそれぞれの計測状態が正常であるか否かに応じて、当該R波の間隔である、前記算出手段により算出した瞬時心拍の計測状態を評価する評価手段と、前記評価手段により評価した瞬時心拍を、当該評価の種別に応じて、心拍特徴量の種別に応じた複数種類のクラスのいずれかに分類する分類手段と、前記分類手段により分類された各クラスの瞬時心拍のうち、所望の種別の心拍特徴量が対応するクラスに分類された瞬時心拍を選択する選択手段とを有する。【選択図】図1
