Artificial intelligence-based versus manual assessment of prostate cancer in the prostate gland: a method comparison study

Aim To test the feasibility of a fully automated artificial intelligence-based method providing PET measures of prostate cancer (PCa). Methods A convolutional neural network (CNN) was trained for automated measurements in F-18-choline (FCH) PET/CT scans obtained prior to radical prostatectomy (RP) in 45 patients with newly diagnosed PCa. Automated values were obtained for prostate volume, maximal standardized uptake value (SUVmax), mean standardized uptake value of voxels considered abnormal (SUVmean) and volume of abnormal voxels (Vol(abn)). The product SUVmean x Vol(abn) was calculated to reflect total lesion uptake (TLU). Corresponding manual measurements were performed. CNN-estimated data were compared with the weighted surgically removed tissue specimens and manually derived data and related to clinical parameters assuming that 1 g approximate to 1 ml of tissue. Results The mean (range) weight of the prostate specimens was 44 g (20-109), while CNN-estimated volume was 62 ml (31-108) with a mean difference of 13 center dot 5 g or ml (95% CI: 9 center dot 78-17 center dot 32). The two measures were significantly correlated (r = 0 center dot 77, P<0 center dot 001). Mean differences (95% CI) between CNN-based and manually derived PET measures of SUVmax, SUVmean, Vol(abn) (ml) and TLU were 0 center dot 37 (-0 center dot 01 to 0 center dot 75), -0 center dot 08 (-0 center dot 30 to 0 center dot 14), 1 center dot 40 (-2 center dot 26 to 5 center dot 06) and 9 center dot 61 (-3 center dot 95 to 23 center dot 17), respectively. PET findings Vol(abn) and TLU correlated with PSA (P<0 center dot 05), but not with Gleason score or stage. Conclusion Automated CNN segmentation provided in seconds volume and simple PET measures similar to manually derived ones. Further studies on automated CNN segmentation with newer tracers such as radiolabelled prostate-specific membrane antigen are warranted.