Segmentation of the effective area of images of renal biopsy samples

Abstract

Diagnosis and monitoring of kidney diseases and transplants is supported by microscopic analysis of needle-core biopsy samples. The current methods of analysis allow for inconsistencies, bias, and inaccuracies. We propose image processing methods for automatic segmentation of the effective biopsy area (cortex and medulla) from digital images of renal biopsy samples. The methods include opening-by-reconstruction, a morphological closing operation, and morphological erosion. The results are compared to 100 randomly selected images manually marked by an experienced renal pathologist. Comparative measures indicate that the automatically detected region of interest closely matches the ground truth; the mean distance to the closest point was 5.46 ± 3.92 µm (6 ± 4.31 pixels) and the true-positive fraction was 98.25 ± 1.77%.