Oncology Molecular Pathology Digital and computational pathology Liquid biopsy Omics

AI Models Predict CNS Tumors From Spinal Fluid

At AMP 2025, a South Korean research team outlined a multimodal approach that integrates CSF, DNA, and MRI data to enable early, noninvasive tumor classification.

12/04/2025 News 1 min read

article Full Article Summary Notecard

Share

AI models integrate genetic and imaging data.
Developed at Soonchunhyang University.
Potential for early noninvasive diagnosis.
Models classify CNS tumors with high accuracy.
Reduced need for invasive biopsies.
Presented by Dr. Jieun Kim at AMP 2025.
Dense neural network had MCC of 0.8822.