Researchers have developed a clinician-guided AI segmentation pipeline for lung cancer CT scans, aiming to improve prognostic accuracy and clinical trust. The system, guided by the Knowledge-to-Action framework, utilized five deep learning models, with VNet achieving the best performance in segmentation accuracy, radiomic stability, and predictive accuracy using semi-supervised learning. Radiologists found the AI-generated masks valuable for refinement, indicating a promising path for physician-centered AI translation in oncology. AI
IMPACT This research demonstrates a feasible path for integrating AI into clinical workflows for improved lung cancer prognosis, potentially increasing adoption of AI tools in medical imaging.
RANK_REASON This is a research paper detailing a novel AI model and its evaluation. [lever_c_demoted from research: ic=1 ai=1.0]
- 3D Attention U-Net
- Knowledge to action framework for home health monitoring
- Mohammad Salmanpour
- Pysera
- ReconNet
- ResUNet
- SAM-Med3D
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →