Thomas Jefferson University researchers developed an AutoML model that distinguishes pituitary macroadenomas from parasellar meningiomas on MRI with over 97% accuracy.
Key Details
- 1AutoML model trained to classify pituitary macroadenomas vs. parasellar meningiomas on preoperative MRI.
- 2Achieved 97.55% overall accuracy, with sensitivities of 97% (macroadenoma) and 98.41% (meningioma), and specificities of 98.96% and 95.53%, respectively.
- 3External validation conducted on 959 additional MRI images.
- 4Model allows different confidence thresholds, aiding both community screening and tertiary centers.
- 5First reported use of AutoML for this specific neuro-oncology imaging task.
- 6The study published in Otolaryngology–Head and Neck Surgery (Dec 2025); presented at AAO-HNSF 2025.
Why It Matters
Accurate, non-invasive differentiation of similar-appearing benign brain tumors using MRI can improve preoperative planning, reduce surgical risks, and ensure patients receive optimal treatment. This demonstrates AutoML's potential to streamline diagnostic AI development and deployment in neuro-radiology.
Source
EurekAlert
Related News
•EurekAlert
AI-Powered OCT Enables Rapid 'Optical Biopsy' for Early Endometrial Cancer Detection
A team at Washington University has developed a catheter-based 3D OCT system with AI to quickly and noninvasively detect early endometrial cancers.
•EurekAlert
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
•EurekAlert
KAIST, MIT, Microsoft Develop Efficient AI Image Upsampling for Robotics
KAIST, MIT, and Microsoft have created 'Upsample Anything,' a training-free AI method to restore high-resolution visual data from compressed images with up to 16x improved GPU memory efficiency.