Automated Classification of Pediatric Soft Tissue Sarcoma from Histopathology Images

Jackson Laboratory, Bar Harbor ME

Investigators

Linked publications & trials

Abstract

PROJECT SUMMARY The mission of the Jackson Laboratory Cancer Center (JAXCC) is to discover precise genomic solutions for cancer by making basic discoveries with human impact. Committed to our role as a basic research Cancer Center, we harness our transdisciplinary strengths in complex genetics, functional genomics, and computational oncology to achieve this goal. Organized in a single research program, JAXCC’s members combine a unique capability to model human cancers in mice with innovations in genomic and computational analytical approaches. From RNA biology to chromatin dynamics, from organismal to in silico models of cancer, from panels of genetically diverse strains to humanized mice, from single cell genomics to Patient Derived Xenografts, and from genomic engineering technologies to genomic diagnostics – the JAXCC brings advanced technologies to bear on persistent cancer problems through collaborative partnerships with clinical investigators. In the last decade, we have seen unprecedented growth with expertise in immunology, computational biology, genomic technologies, epigenetics, tumor microenvironments, and clinical genomics. Our Shared Resources (Cancer Model Development, Computational Sciences, Genetic Engineering Technologies, Genome and Single Cell Technologies, and Phenotyping Technologies) provide comprehensive support for this wide range of research. The JAXCC strives to identify origins and consequences of complex structural genomic alterations in cancer.. The JAXCC has a strong foundation in computational oncology, coupled with a robust network of clinical collaborators and proven success in coordinating the aggregation, integration, and analysis of data from diverse sources. Pediatric soft tissue sarcoma is a rare diagnosis, comprised of myriad histological subtypes with widely variable prognosis and clinical behavior. Existing datasets are insufficiently powered to enable to construction of tools with high diagnostic accuracy. The goals of this supplement project are to harness JAXCC’s strengths, as a coordinating center and also in the areas of computational oncology and digital pathology, to augment existing CCDI resources with an expanded collection of digitized whole slide images of pediatric soft tissue sarcomas. We aim to contribute further to the CCDI tool set by building computational classifiers based on the gathered digital histopathology slides to improve diagnostic accuracy in this rare group of cancers.