Adding MRI to a multimodal AI model significantly improves breast cancer detection and risk prediction.
Key Details
- 1NYU Langone researchers evaluated a transformer AI model using mammography, DBT, ultrasound, breast MRI, and clinical variables.
- 2The training set included 1.3 million exams from 274,388 women (2010–2022).
- 3A separate test cohort comprised 1,944 women with 18,201 exams.
- 4Without MRI input, the AUROC for 5-year cancer risk prediction was 0.899; with MRI, it increased to 0.94.
- 5Improved model performance allows better identification of high-risk women for screening and prevention.
Why It Matters
This work demonstrates that integrating MRI with other imaging modalities in AI models can meaningfully enhance risk stratification and detection accuracy for breast cancer. Such advances support precision medicine, enabling more targeted screening and preventive interventions in clinical radiology.
Source
AuntMinnie
Related News
•Radiology Business
Patients Favor AI in Imaging Diagnostics, Hesitate on Triage Use
Survey finds most patients support AI in diagnostic imaging but are reluctant about its use in triage decisions.
•HealthExec
FDA Eases Path for AI in Clinical Decision Support and Healthcare Innovation
FDA publishes new guidance to promote innovation in general wellness and clinical decision support, impacting medical AI including radiology.
•AuntMinnie
Deep Learning AI Outperforms Radiologists in Detecting ENE on CT
A deep learning tool, DeepENE, exceeded radiologist performance in identifying lymph node extranodal extension in head and neck cancers using preoperative CT scans.