Case Study: Oxford Cancer Biomarkers advances precision oncology with CSI Ltd's AI infrastructure

Oxford Cancer Biomarkers uses CSI Ltd to target the 8% of bowel cancer patients most likely to benefit from adjuvant chemotherapy

Oxford Cancer Biomarkers (OCB), a cancer research facility spun out from the University of Oxford, faced the challenge of developing precision oncology diagnostic technology. Their goal was to use machine learning on histopathology images to identify which bowel cancer patients would benefit from adjuvant chemotherapy, sparing others from unnecessary treatment. This required a high-performance, computationally intensive IT infrastructure, which led them to partner with vendor CSI Ltd.

CSI Ltd implemented a high-performance platform for AI deep learning, provided as an Infrastructure as a Service (IaaS) model from the CSI PowerCloud. The solution featured an IBM Power System AC922 server with POWER CPUs and NVIDIA Tesla GPUs, along with IBM V7000 storage. This allowed OCB to successfully train its large AI models on high-definition images. The results enable clinicians to stratify patients more effectively, improving patient outcomes and creating significant cost-efficiencies throughout the healthcare value chain by streamlining processes and reducing unnecessary treatments.

Oxford Cancer Biomarkers