Polygon Health Analytics partnered with the National Cancer Institute (NCI) under a Phase I SBIR contract to develop MAGIC—a cloud-based platform that integrates genomics, imaging, and clinical data. MAGIC aims to make petabyte-scale cancer datasets more accessible and actionable for researchers and clinicians.

Despite the availability of rich datasets from initiatives like TCGA, TCIA, and CPTAC, cancer researchers face major hurdles in integrating and analyzing diverse data types—especially unstructured clinical narratives and scanned pathology reports. This fragmentation limits the potential of precision oncology.

• NLP for Pathology Reports: We applied OCR and NLP to extract standardized pathology indicators from scanned PDFs in the TCGA repository.
• Cloud-Based Platform: Built a prototype integrated with the Cancer Research Data Commons (CRDC), featuring secure access, intuitive interfaces, and multimodal analytics.
• Real-World Validation: Demonstrated MAGIC's utility through two studies on squamous cell carcinoma (SCC), including biomarker discovery and survival prediction using deep learning.

• Enabled integrated analysis of clinical, imaging, and genomic data.
• Validated platform usability with 25+ participants from academia and industry.
• Public demo available at https://magic.polygonhealthanalytics.com