- AI Meets Healthcare
- Posts
- 🤝 AI enhances early detection of pancreatic cysts #3
🤝 AI enhances early detection of pancreatic cysts #3
AI Meets Healthcare Newsletter: Bridging AI and Medicine — Daily News & Insights
Today’s Insights
How AI could automate genomics research
Deep learning enhances gait analysis for spinal deformity detection
AI enhances early detection of pancreatic cysts
World's first fully robotic double lung transplant
Media Pick: Can AI catch what doctors miss? — TED talk
📰 NEWS
Credit: UC San Diego Health Sciences/Erik Jepsen
How AI could automate genomics research: Researchers at UC San Diego School of Medicine have shown that GPT-4 can help automate functional genomics - the study of what genes do and how they interact. After testing five different AI models, GPT-4 proved best, being 73% accurate in identifying gene functions and avoiding mistakes by refusing to name 87% of random gene sets. The team, led by Dr. Trey Ideker, created a web tool to help scientists use AI in their gene research, potentially saving hours of manual analysis time. While more research is needed, this study shows the potential of AI in advancing genomics and precision medicine.
Credit: IEEE Access (2024). DOI: 10.1109/ACCESS.2024.3479165
Deep learning enhances gait analysis for spinal deformity detection: Researchers at University of Tsukuba in Japan have developed a new deep learning method to diagnose adult spinal deformity (ASD) by analyzing walking patterns in videos. The method focuses on studying the rhythm and symmetry of body movements during walking, achieving a 71.43% accuracy rate - better than conventional methods at 66.30%. Testing was done on 81 patients, and the technology could enable quick, real-time diagnosis of ASD in clinical settings in the future.
🧬 MORE IN AI & HEALTHCARE
AI enhances early detection of pancreatic cysts
Credit: The Ohio State University Comprehensive Cancer Center.
Researchers at Ohio State University are developing AI to improve pancreatic cancer detection through a "virtual biopsy" procedure that examines pancreatic cysts using mini probe-based laser optical cameras. Led by Dr. Somashekar Krishna, the team is developing an AI algorithm that can rapidly analyze endoscopic ultrasound videos to identify cancer risk factors in seconds rather than hours, potentially preventing unnecessary surgeries since about 50% of removed precancerous cysts likely wouldn't have progressed to cancer.
For patients identified with early-stage or precancerous conditions, Dr. Krishna has developed a minimally invasive alternative to the traditional Whipple procedure, using targeted heat through radiofrequency ablation to eliminate abnormal tissue while preserving the pancreas
World's first fully robotic double lung transplant
The world's first fully robotic double lung transplant took place at NYU Langone Health Center. The procedure was a minimally invasive surgery and aimed to reduce hospital stays as well as speed up recovery periods.
🎬 MEDIA PICK
Can AI catch what doctors miss? — TED talk by physician-scientist Eric Topol.
Thanks for reading! Want to stay ahead of AI's impact on healthcare? Subscribe to get daily curated insights delivered to your inbox. Don't forget to share with others who'd find this valuable.
Until next time,
Alisa