Lesion candidacy is detected making use of the characteristic quantity of form of the shading. It creates the medical image processing device, the distribution histogram every of size of the image part which consists of the form of the inside specification of the particular medical image making use of the data of the medical image, one or concerning the form where multiple differs and the lesion candidacy which is included in the aforementioned medical image on the basis of the result which the characteristic quantity in the distribution histogram every of size (step S4), aforementioned one which was created or concerning the form where multiple differs the characteristic quantity was analyzed, it is detected (step S5) it displays the detection result of candidacy detection means and indicatory means and aforementioned lesion candidacy in aforementioned indicatory means (step S6) control meansIt has.陰影の形状の特徴量を用いて病変候補を検出する。医用画像処理装置は、医用画像のデータを用いて、当該医用画像のうち特定の形状からなる画像部分のサイズ毎の分布ヒストグラムを、一又は複数の異なる形状について作成し(ステップS4)、前記作成した一又は複数の異なる形状についてのサイズ毎の分布ヒストグラムの特徴量を特徴量解析した結果に基づいて、前記医用画像に含まれる病変候補を検出する(ステップS5)候補検出手段と、表示手段と、前記病変候補の検出結果を前記表示手段に表示する(ステップS6)制御手段と、を備える。
