An Electrical Circuit Model of Circle of Willis to Predict Stroke



Journal Title

Journal ISSN

Volume Title



Is it possible to predict which individuals are more prone to stroke years before they actually get one? An affirmative answer to this question has big implications for public health. The Circle of Willis (CoW), a ring of arteries that distributes blood flow to the two hemispheres of the brain, is only completely present in ~50% of all people. We hypothesized that the differences in the variations explain the differences in the relative probability of individuals suffering a blockage owing to a stroke. To address this, it is necessary to model these variations and to introduce blockages in these modeled variations and observe how it affects blood flow, temporarily or permanently. This research modeled the CoW using different parameters of the biological structure with electrical components. Biological data from empirical, clinical studies constrained our circuit model. High-Performance-Computing was utilized to match electrical current ratios to blood flow ratios obtained from empirical studies to compute biologically plausible circuit models where the values of the circuit components match the anatomical values. Approximately 215 million potential solution models were computed and compared with clinical data, of which 3,000+ models with flow rates matching clinical data for complete and two incomplete variations of the CoW were found. Then the circuit model was simplified and the changes in blood flow across variations were analyzed. We anticipate our solution models to be the starting point for in-depth analyses of the variations or blockages in the CoW for which there are no clinical data readily available.