A machine-learning MRI radiomics model shows high accuracy for preoperatively subtyping colorectal cancer, potentially enabling improved treatment planning.
Key Details
- 1Study used preoperative multiparametric MRI radiomics from 253 colorectal cancer (CRC) patients.
- 2Model predicted consensus molecular subtype 4 (CMS4) with AUCs of 0.85 (internal) and 0.84 (external).
- 3Performance surpassed established deep learning models (AUCs 0.70–0.75; p < 0.01).
- 4The model was also predictive of recurrent metastasis risk (hazard ratio: 5.96; p < 0.001).
- 5Transcriptomic analyses linked the radiomics signature to aggressive tumor pathways.
- 6Current methods for CMS4 typing are limited by cost and tissue requirements; the radiomics approach offers a noninvasive, cost-effective alternative.
Why It Matters
This research demonstrates that MRI-based radiomics, powered by machine learning, can stratify aggressive CRC subtypes noninvasively, potentially guiding individualized therapy and improving outcomes. It highlights a significant advance in the integration of quantitative imaging and genomics (radiogenomics) for oncology.
Source
AuntMinnie
Related News
•Radiology Business
Private Equity Backs AIRS Medical to Expand MRI AI Globally
TA Associates is investing in AIRS Medical to accelerate its global expansion of AI-powered MRI efficiency solutions.
•AuntMinnie
Deep Learning Models Rival Radiologists for Pancreatic Cancer Detection on CT
Deep-learning models achieved comparable or superior accuracy to experienced radiologists in detecting pancreatic cancer on CT scans, especially for small tumors.
•Radiology Business
Radiology AI Devices at Elevated Risk for FDA Recalls, Study Finds
Radiology AI devices are more likely to face FDA recalls, largely due to deviations from intended use and incomplete clinical data.