A deep learning model leveraging SAM improved segmentation speed and matched radiologist performance in classifying ovarian lesions on MRI.
Key Details
- 1Researchers from Johns Hopkins developed an MRI-based, end-to-end DL pipeline incorporating Meta’s Segment Anything Model (SAM) and DenseNet-121.
- 2Integrated model reduced segmentation time by 4 minutes per lesion compared to manual segmentation.
- 3SAM achieved a Dice coefficient of 0.86 to 0.88 for lesion segmentation.
- 4DL model achieved an AUC of 0.85 internally and 0.79 externally, on par with radiologists’ AUC of 0.84 (p > 0.05).
- 5Training data included 534 lesions (internal) and 87 lesions (external) from the US and Taiwan.
Why It Matters
This study demonstrates that foundation AI models can efficiently streamlines lesion segmentation and classification with accuracy rivaling radiologists, suggesting a path toward more automated and collaborative clinical workflows in radiology.
Source
AuntMinnie
Related News
•Radiology Business
AI-Enhanced Structured Reporting Can Speed Up Radiology Turnaround Times
A study found that combining structured reporting templates with AI assistance significantly reduces radiology report turnaround times.
•Radiology Business
Medicare May Deny Coverage for AI-Based Brain MRI Tools
A Medicare contractor has proposed denying coverage for AI tools used in brain MRI analysis, citing insufficient evidence and data limitations.
•AuntMinnie
Task-Specific NLP Outperforms General LLMs for Lung Nodule Detection in Chest CT Reports
A radiology-trained NLP model significantly outperformed several general-purpose large language models in extracting incidental lung nodule data from chest CT reports.