With 'Digital Twins,' the Doctor Will See You Now

Amanda Randles, a computer scientist, is developing digital twins of patients' circulatory systems to enhance medical forecasting. Her work involves creating detailed computer models of blood flow, which can simulate and predict the dynamics of blood within arteries over extended periods. Randles' system utilizes 3D imaging to analyze blood vessels, allowing doctors to observe blood behavior and identify potential issues linked to heart disease. Her model, named Harvey, has advanced significantly, now capable of simulating over 700,000 heartbeats, compared to just 30 a decade ago. This capability enables doctors to visualize critical factors like wall shear stress and blood flow vortices, which can influence treatment decisions.

Randles' background includes a degree in physics and computer science from Duke University and a doctorate from Harvard, where she initially developed her blood circulation model. She emphasizes the importance of visual data in making accurate medical predictions, despite the computational challenges involved. Recently, her team has integrated machine learning to streamline the simulation process, reducing training time from 24 hours to about 10 minutes. However, Randles acknowledges the potential drawbacks of machine learning, such as the risk of bias in data and the need for interpretability in clinical settings. Looking ahead, she envisions a future where comprehensive data integration could lead to proactive healthcare interventions, although she notes that achieving this will require overcoming significant technical challenges.