University of Pittsburgh researchers are developing a novel artificial intelligence (AI) assisted non-invasive diagnostic tool to diagnose the severity of deep vein thrombosis (DVT) and predict those most at risk of developing post thrombotic syndrome (PTS).
Description
DVT, where a blood clot forms in the deeper veins of the body, occurs mostly in the legs, and affects 1 in 1000 adults annually. The mortality rate in the first month can be as high as 30%. For those who survive, as many as half will develop PTS within two years, a condition where the most severe form can lead to venous ulceration, significantly impacting on quality of life, and with major economic and social impact. Identifying those most at risk of developing severe PTS would allow clinicians to implement early preventative treatment such as compressive socks, reducing the risk of severe PTS and improving outcomes for DVT and PTS patients.
Applications
• Deep vein thrombosis
• Post thrombotic syndrome
Advantages
To date, no non-invasive methods exist to diagnose and characterize the severity of DVT. Additionally, it is not possible to identify which patients with DVT may develop PTS, with patient characteristics (age, sex, etc.,) having no clear impact on risk. Identifying those most at risk of developing PTS which is a result of damage to the veins following DVT resulting in venous reflux, would allow clinicians to instigate early treatment.
This approach aims to understand the fundamental changes within veins regarding volume and pressure to provide an insight into venous compliance. Venous compliance is vital for returning deoxygenated blood to the heart and in damaged veins suboptimal compliance can lead to the debilitating and fatal impacts of DVT and PTS. Through biomechanical stress analysis it should be possible to understand the transition of DVT to PTS. This novel approach will use non-invasive analysis to measure mechanical changes in the vein wall after DVT and use AI to predict a patient’s risk of developing PTS.
Invention Readiness
Currently at the concept phase, work is ongoing to biomechanically assess the veins of patients recently diagnosed with DVT and PTS to provide insight into the relationship between venous compliance and the risk of developing PTS. A piezoelectric vibration/sound detection device is being developed to allow for non-invasive analysis of veins in the form of a novel pressure cuff. All mechanical and clinical information will be used to develop a machine learning model based on the data collected from patients to produce an AI tool capable of classifying the severity of DVT and PTS.
IP Status
https://patents.google.com/patent/WO2024091670A1Related Publication(s)
Chung, T. K., Liang, N. L., & Vorp, D. A. (2022). Artificial intelligence framework to predict wall stress in abdominal aortic aneurysm. Applications in Engineering Science, 10, 100104. https://doi.org/10.1016/j.apples.2022.100104
Chung, T. K., Gueldner, P. H., Kickliter, T. M., Liang, N. L., & Vorp, D. A. (2022). An Objective and Repeatable Sac Isolation Technique for Comparing Biomechanical Metrics in Abdominal Aortic Aneurysms. Bioengineering, 9(11), 601. https://doi.org/10.3390/bioengineering9110601
