| AI-POCUS Artificial intelligence-guided screening of under-recognized cardiomyopathies at the point of care | |
|---|---|
| Lead | Rohan Khera, MD, MS |
| Team | Evan Oikonomou, MD, DPhil |
| Description |
Handheld and portable devices are increasingly used across health settings and in communities to get ultrasound videos of the heart. These videos are captured while clinicians look for some key abnormalities they have been trained to detect, but they are unable to use these for advanced cardiac diagnosis. We have developed a series of AI tools that can take these videos acquired by almost anyone, and automatically diagnose a range of heart muscle and valve disorders without any additional asks from the clinician. |
| Target Impact | Our AI technology allows early diagnosis of treatable cardiovascular disease by operators with minimal experience through a portable device, now even attachable to a smartphone. This has significant implications for both healthcare systems where dozens of portable ultrasound studies are performed on a daily basis, as well as global settings where AI may enable scalable screening of heart disease through simplified protocols. |
| Project Lead's Vision | Many heart muscle and valve disorders are now treatable but are often diagnosed too late because of the need for specialized testing. This specialized testing is only done once someone has symptoms, losing out on all the treatment options that could have been deployed early in the disease course. Our technology addresses this need by bringing the ability of advanced automated cardiac diagnostics to simpler and accessible data captured on handheld and portable devices that can be used in early stages of the disease. |
Center for Health Care Innovation
