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 that simulate blood flow, allowing doctors to visualize and predict the behavior of blood within vessels. Randles' system, which can simulate over 700,000 heartbeats, helps in diagnosing and treating diseases noninvasively by providing insights into blood dynamics, such as wall shear stress and vortices, which are linked to heart disease.

Randles' journey began in high school, where she combined biology and computer programming. After earning a degree from Duke University and working with supercomputers at IBM, she completed her doctorate at Harvard, where she developed a blood circulation model named Harvey. This model has been validated against patient data, demonstrating its accuracy in predicting blood flow dynamics.

Her team is also leveraging machine learning to reduce computational demands, enabling real-time predictions during patient consultations. However, Randles emphasizes the importance of interpretability in machine learning models to ensure that doctors understand the basis for their decisions. She acknowledges the challenges of integrating diverse data sources and the need for a comprehensive understanding of individual patient geometries to improve predictive accuracy. Despite the hurdles, Randles is optimistic about the future of digital twins in healthcare, envisioning a system that could integrate various biological data for proactive medical interventions.