Working with a KIST-led research team, SolverX will use data from blood experiments to predict how circulating tumor cells move through a blood purification device.
Predicting where circulating tumor cells travel within a blood purification device is a complex challenge. As blood flows through narrow channels, red blood cells can push other cells toward the channel walls, a phenomenon known as margination. Because cells differ in size and flexibility, it is difficult to predict from simulations alone how many tumor cells will pass through the region monitored by the device.
PURIFY, part of the Korean ARPA-H initiative, is exploring blood purification for cancer management. Led by Dr. Jaeho Shin of the Korea Institute of Science and Technology (KIST), the research team is developing a device that uses blood flow and light to detect circulating tumor cells. SolverX has joined the project as a research partner, focusing on modeling blood flow and cell movement within the device.
SolverX developed AI Enhancer to incorporate experimental measurements into its physics models for manufacturing. As part of PURIFY, AI Enhancer will compare predictions from a model trained on simulation data with the cell distributions observed in blood experiments. For example, if experiments show that more cells pass near a channel wall than the model predicts, the technology will use that difference to help refine the AI model.
The goal is to use findings from one prototype to inform the design of the next. If the model can predict how changes in channel shape or flow rate affect the paths of tumor cells, the team could compare different designs before building them and determine which one to test first.