Related Experiment Video
Updated: Aug 8, 2026
07:34
Enhancing Prostate Tumor Biobanking Reliability with Improved Sampling Technique and Histological Characterization
Published on: November 17, 2023
Deep learning-based histologic classifiers enable molecular subtyping of metastatic prostate cancer
Zhijun Chen1, Erolcan Sayar2, Daniela Guevara3
1National Cancer Institute, NIH, Bethesda, United States of America.
JCI Insight
|August 6, 2026
Summary
A new deep learning model, NEURAL-PC, accurately diagnoses neuroendocrine prostate cancer (NEPC) from H&E images. This tool also predicts prognosis, aiding advanced prostate cancer subclassification.
Area of Science:
- Oncology
- Computational Pathology
- Artificial Intelligence in Medicine
Background:
- Metastatic prostate cancer is heterogeneous, often developing resistance to therapy.
- Neuroendocrine prostate cancer (NEPC) presents a challenging phenotype with poor prognosis.
- Accurate NEPC diagnosis is difficult due to its complexity.
Purpose of the Study:
- To develop a deep learning model for diagnosing NEPC using H&E stained tissue images.
- To assess the model's ability to provide prognostic information for advanced prostate cancer.
Main Methods:
- Developed NEURAL-PC, a deep learning model using interpretable cellular features.
- Employed a multiple instance learning (MIL) framework for classification from H&E images.
- Validated the model on independent external datasets.
Main Results:
- NEURAL-PC achieved an AUROC of 0.921 for NEPC classification in external validation.
- The model demonstrated strong prognostic value and generalizability across diverse datasets.
- NEURAL-PC successfully subclassified advanced prostate cancer.
Conclusions:
- NEURAL-PC offers a reliable deep learning tool for diagnosing and prognosing advanced prostate cancer.
- The hybrid approach integrating features across domains shows promise for pathology AI.
- NEURAL-PC represents a validated diagnostic and prognostic model for prostate cancer.
