The swallowing diagnosis apparatus 100 detects the respiratory phase before and after the period in which it is estimated that the swallowing estimation unit 116a has caused the swallowing from the respiratory pressure data output from the pressure sensor 114, and performs the swallowing based on the detected respiratory phase. The body sound data output from the sound sensor 131a in a predetermined period including a period in which it is estimated that swallowing has occurred by the first swallowing determination unit 116b that determines whether there is a risk of aspiration and the swallowing estimation unit 116a. A second swallowing determination unit 116c that extracts and acquires a feature amount from the extracted body sound data, executes machine learning processing on the acquired feature amount, and determines whether or not there is a possibility of dysphagia in the swallowing And a display control unit 116d for causing the display unit 111 to display a determination result by each swallowing determination unit.嚥下診断装置１００は、嚥下推定部１１６ａにより嚥下が生じたと推定された期間の前後の呼吸相を、圧力センサ１１４から出力される呼吸圧データから検出し、検出した呼吸相に基づいて当該嚥下に誤嚥のリスクがあるか否かを判定する第１嚥下判定部１１６ｂと、嚥下推定部１１６ａにより嚥下が生じたと推定された期間を含む所定の期間において音センサ１３１ａから出力される生体音データを抽出し、抽出した生体音データから特徴量を取得し、取得した前記特徴量に機械学習処理を実行して当該嚥下に嚥下障害の可能性があるか否かを判定する第２嚥下判定部１１６ｃと、各嚥下判定部による判定結果を表示部１１１に表示させる表示制御部１１６ｄと、を備える。
