Statistical image analysis of an ensemble of digital images of histological samples is performed as an auxiliary investigation a result of the recently proposed method of articular cartilage repair utilizing growth plate chondrocytes in a skeleton animal model. A fixed-shift model of maximal likelihood estimates of image histograms applied for monochromatic (grayscale) images or their RGB components confirms the statistically significant effect of the previously proposed medical treatment. The type of staining used to prepare images of histological samples is related to the visibility of the effectiveness of medical treatment. Hellinger distance of escort distributions for maximal likelihood estimates of image histograms of medically treated and control samples is investigated to identify grayscale (or RGB) intensities responsible for statistically significant difference of the estimates. A difference of Shannon entropy quantifying informational content of the histograms allows one to identify staining and image colors which are most suitable to visualize cluster formation typical for articular cartilage repair processes.