To execute clustering with a division number with higher reliability when cluster analysis based on MS imaging data is executed in order to grasp a distribution state of a specific substance, etc. in a biological sample.

A specific portion extraction part 24 extracts a specific portion from a microscopic image acquired by performing coloring or fluorescent label to the specific portion on a sample 4. A cluster analysis part 25 and a division number determination processing part 26 evaluate similarity with spatial distributions of all pixels belonging to a certain cluster when respective pixels are classified into a plurality of clusters based on similarity between spatial distribution of the specific portion and MS imaging data. Since the specific portion is the one containing the same characteristic substance, when the similarity of the spatial distributions is high, it is determined that clustering is appropriate. Then, the appropriate division number in cluster analysis is determined based on correlation of the spatial distributions, and a cluster analysis result under the division number is output from a display part 31.