The display control unit 24 displays the brain image in which the infarction region has been discriminated on the display unit 14. FIG. 8 is a diagram illustrating the displayed CT image of the brain. In addition, FIG. 8 illustrates a slice image in one tomographic plane of the CT image Bc1. The infarction region A1 discriminated in the CT image Bc1 is displayed so as to be surrounded by a line as illustrated in FIG. 8. In addition, the display control unit 24 may display the infarction region A1 in any aspect on the display unit 14. For example, the display control unit 24 may display the infarction region A1 so as to be hatched, may give a specific color to the infarction region A1, may apply an arrow to the infarction region A1, or may highlight the infarction region A1 differently from other regions.
Next, a process performed in this embodiment will be described. FIG. 9 is a flowchart illustrating a process performed during learning in this embodiment. First, the image acquisition unit 21 acquires the CT images Bc0 and the MR image Bm0 of the brain of the subject that has developed cerebral infarction (Step ST1). Then, the learning unit 22 trains the discriminator 23, using the infarction region specified in the CT image Bc0 as training data, so as to discriminate an infarction region in the input CT image Bc1 and trains the discriminator 23, using a data set of an image set of the CT image Bc0 and the MR image Bm0 registered with each other as training data, such that the MR estimated image Dm1 is output from the input CT image Bc1 (Step ST2). Then, the process ends.
