Researchers developed a hybrid AI model that significantly improves early lung cancer detection from CT scans.
Key Details
- 1Lung cancer causes approximately 1.8 million deaths annually, largely due to late diagnosis.
- 2The AI system uses a hybrid approach (CNN + transformer) to analyze both fine details and broader context in CT scans simultaneously.
- 3Tested on a dataset of healthy and cancerous cases, the model achieved over 96% accuracy, outperforming previous methods.
- 4The approach aims to support clinicians by improving detection rates and reducing false alarms and time per patient.
- 5Researchers note the need for larger, more diverse datasets for further validation, and clinical trials are planned.
- 6The methodology may also benefit other imaging domains, including brain and breast cancer diagnostics.
Why It Matters
This development could increase early detection rates, improve patient outcomes, and provide radiologists with a more reliable diagnostic tool. Such advances are crucial for adoption and validation of AI in routine clinical workflows.
Source
EurekAlert
Related News
•EurekAlert
AI Pathology Tool SÉMIL Improves Stage II Bowel Cancer Risk Assessment
A La Trobe University-developed AI tool accurately predicts relapse risk in stage II bowel cancer using digital pathology images and descriptions.
•EurekAlert
AI-Guided Handheld Cardiac Ultrasound Reduces Referrals and Costs in Spain
AI-guided handheld cardiac ultrasound enables primary care physicians to detect heart failure, reducing specialist referrals and saving costs.
•EurekAlert
AI Tool Predicts Which Rectal Cancer Patients Benefit from Intensive Therapy
UCL researchers developed an AI that analyzes biopsy slides to identify rectal cancer patients who benefit from adding irinotecan to standard chemoradiotherapy.