Google Research 3:00 am on May 23, 2024
In this study, we assessed ML system assistance in evaluating CT scans for lung cancer by presenting radiologists with challenging cases both with and without the model's help. Results indicated a 57% increase in specificity when using the system. Partnerships are being explored to integrate these findings into practical healthcare applications, highlighting potential cost savings, reduced patient anxiety, and improved screening programs efficiency.
- Radiologist-Assisted ML Study: Evaluation of CT scans for lung cancer with and without AI assistance.
- Significant Impact on Specificity: 57% increase in correctly identifying non-cancerous findings when using the system.
- Potential Healthcare Benefits: Fewer unnecessary procedures, lower anxiety for patients and cost reductions in screening programs.
- Partnership Initiatives: Collaborations with DeepHealth and Apollo Radiology International to integrate the system into healthcare services.
- Open Source Contributions: Sharing of research methodologies, insights, and code for broader academic and clinical use.
http://blog.research.google/2024/03/computer-aided-diagnosis-for-lung.html
< Previous Story - Next Story >