An automated approach to detect human ovarian tissues using type P63 counter stained histopathology digitized color images

摘要

The human ovary contains a fixed number of reproductive tissues at birth. This number decreases with age which leads to complications in conceiving. As part of the medical consultation process and to know the actual condition of a female ovary, manual microscopic routine examination is carried out by pathology experts. Laboratory expert manual microscopic analysis is time consuming and prone to errors. To minimize labor cost and time associated with manual analysis an ultrasound technique is commonly used. This ultrasound technique can only identify larger and more mature tissues rather than small ovarian tissues. To analyze small ovarian reproductive tissues accurately a fully automated approach is proposed in this paper to assist pathology experts. The proposed method processes digitized color histopathology images acquired from biopsy slide and identifies ovarian reproductive tissues automatically. The study's experimental results indicate excellent performance in terms of accuracy.