Click here to

Session: Multi-Disciplinary General ePoster Viewing [Return to Session]

Breast Cancer Patient-Specific Reaction-Diffusion From Spatial Spectral Analysis of Immunohistochemistry Slides

S Pasetto*, M Zahid, R Diaz, M Rosa, R Gatenby, H Enderling, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL


PO-GePV-M-355 (Sunday, 7/10/2022)   [Eastern Time (GMT-4)]

ePoster Forums

Purpose: To quantify tumor progression in a reaction-diffusion framework from a single-routinely collected histopathological slice and quantitatively connect radiation to immune response.

Methods: We use spatial-spectral analysis of 20 patients accrued to a phase II study of preoperative SABR with 9.5 x 3 Gy for early-stage breast cancer whose tissues were stained with multiplex immunofluorescence. We employ the reaction-diffusion framework as interpretative theoretical counterpart and the Bayesian framework to solve the inference problem for patient-specific tumor growth and diffusion rate.

Results: Comparing the data-deduced two-point correlation function and the corresponding spatial power spectral distribution with the theoretical predicted by the reaction-diffusion equation, we found that a single histopathological slice suffices to characterize the reaction-diffusion equation dynamics through its power spectral density.

Conclusion: We developed a mathematical approach to infer, from histopathology slices of patient biopsy resection areas taken at the time of the diagnosis, the characteristics of infiltration and growth of the patient's disease. This novel technique aims to anticipate a finely tuned patient treatment with the help of an informative mathematical technique.

Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by NIH/NCI grant U01CA244100, the Florida Breast Cancer Foundation, and by the JAYNE KOSKINAS TED GIOVANIS FOUNDATION FOR HEALTH AND POLICY, a private foundation committed to critical funding of cancer research.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email