A hybrid deep learning approach for gland segmentation in prostate histopathological images
作者：
Highlights：
• Gland architecture plays a crucial role in prostate cancer reporting.
• A new hybrid deep learning method is presented to segment prostate glands.
• A softmax-driven active contour model is proposed to detect the glandular areas.
• No performance degradation is observed when segmenting pathological structures.
• Our approach could be easily extended to other histological structures and tissues.
摘要
•Gland architecture plays a crucial role in prostate cancer reporting.•A new hybrid deep learning method is presented to segment prostate glands.•A softmax-driven active contour model is proposed to detect the glandular areas.•No performance degradation is observed when segmenting pathological structures.•Our approach could be easily extended to other histological structures and tissues.
论文关键词：Glands segmentation,Digital pathology,Computer-aided image analysis,Prostate cancer,Deep learning
论文评审过程：Received 17 September 2020, Revised 8 April 2021, Accepted 10 April 2021, Available online 16 April 2021, Version of Record 16 April 2021.
论文官网地址：https://doi.org/10.1016/j.artmed.2021.102076
