StrikeBC

CIR Team: Philipp Seeböck, Georg Langs, Pamela Zolda

Project Summary

The project combines multimodal data from imaging, blood and tumor samples, and laboratory models to better understand why bladder cancer progresses differently in each patient. Using advanced computational, data analysis and machine learning methods, the consortium aims to support more precise therapy decisions: identifying patients who may safely avoid radical cystectomy after successful initial treatment, and tailoring therapies for those with metastatic, treatment-resistant cancer.

The CIR/MANO team will contribute machine learning expertise to StrikeBC. Our work centers on developing ML models for automated tumor segmentation, treatment response assessment via detection of post-treatment changes, and early prediction of therapy response. We use explainable models to explore the data, learn from model behavior, and identify novel imaging biomarkers. While imaging is our primary focus, we aim to explore integration with multimodal clinical and molecular data to strengthen patient-specific predictions.

By closely integrating clinical expertise, machine learning expertise, and patient perspectives, StrikeBC seeks to translate research findings into tangible improvements in survival and quality of life.