Iterative Health
What They Do
Iterative Health is pioneering the application of computer vision and machine learning to gastrointestinal medicine. The company’s first product, Skout, automates the detection of pre-cancerous polyps and other lesions during colonoscopies, significantly enhancing the diagnostic performance of the average procedure.
The company’s technology also automates the scoring of IBD disease (Crohn’s and Ulcerative Colitis). Patients today - both clinically and in drug trials - are assessed based on a 4-point MES (Mayo Endoscopy Score), by human video “readers.” This human-mediated diagnostic process is inherently subject to human error and variability. Furthermore, it taxonomizes disease in a coarse manner and is incapable of segmenting patients granularly on the basis of novel disease features like lesion color, the geographic distribution of lesions, or longitudinal change relative to prior procedures. Iterative Health has partnered with multiple pharmaceutical companies working in the IBD space to standardize disease scoring, develop novel biomarkers, and facilitate patient recruitment to clinical trials.
Why We Invested
Iterative Health’s technology has the potential to revolutionize the diagnosis and treatment of gastrointestinal diseases that remain enormous unmet medical needs. We see Skout as a much-needed product that should be implemented in every colonoscopy performed worldwide. Ultimately, we believe the company’s data partnerships, engineering talent, and early clinical collaborations position it to achieve a step-change in patient outcomes and diagnoses, as this computer augmentation becomes a standard component of clinical practice.
IBD drug development is now a major priority at multiple top 10 biopharmas, and yet understanding the underlying disease biology frustrates precision therapeutics development. We predict a “fragmentation” of IBD into far more granular, biomarker-defined sub-types, similar to what has occurred in oncology in the past 2 decades. By defining each patient’s disease more precisely, mechanistically-relevant therapeutics can be developed and deployed. As with oncology, disease diagnosis in IBD will almost certainly integrate multiple parameters: genetic data, image data, phenotypic data, medical records, and otherwise. We believe that Iterative Health has great potential to mediate the collection and processing of this data for clinical and research stakeholders.