Three-dimensional histo-morphometric features from light-sheet microscopy images result in improved discrimination of benign from malignant glands in prostate cancer

摘要

Glandular features play an important role in the evaluation of prostate cancer. There has beensignificant interest in the use of 2D pathomics (feature extraction) approaches for detection, diagnosis,and characterization of prostate cancer on digitized tissue slide images. With the development of 3Dmicroscopy techniques, such as open-top light-sheet (OTLS), there is an opportunity for rapid 3Dimaging of large tissue specimens such as whole biopsies. In this study, we sought to investigatewhether 3D features of gland morphology, namely volume and surface curvature, from OTLS imagesoffer superior discrimination between malignant and benign glands compared to the traditional 2Dgland features, namely area and curvature, alone. In this study, a cohort of 8 de-identified fresh prostatebiopsies comprehensively imaged in 3D via the OTLS platform. A total of 367 glands were segmentedfrom these images, of which 79 were identified as benign and 288 were identified as malignant. Glandswere segmented using a 3D watershed algorithm followed by post-processing steps to filter out falsepositiveregions. The 2D and 3D features were compared quantitatively and qualitatively. Ourexperiments demonstrated that a model using 3D features outperformed one using 2D features indifferentiating benign and malignant glands. In 3D, both features, gland volume (p = 1.45 × 10~(?3)) andsurface curvature (p = 3.2 × 10~(?3)), were found to be informative whereas in 2D, only gland area (p =9 × 10~(?18)) was found to be discriminating (p = 0.79 for 2D curvature). Notable visual and quantitativedifferences between 3D benign/malignant glands encourage the development of additional moresophisticated features in the future.