Radiomics and deep learning sharpen adrenal mass assessment

A narrative review highlights radiomics and deep learning as promising tools for adrenal mass assessment, including differentiating pheochromocytoma/paraganglioma, cortical adenoma, and cortical carcinoma and predicting genotype/prognosis.

• The review covers the full radiomics workflow from feature extraction to model development for adrenal lesions.
• Deep learning, especially convolutional neural networks, may enable automated segmentation and end-to-end diagnostic frameworks.
• Technical limitations and clinical translation barriers remain key challenges.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.