I-Corps: Metabolomics and Machine Learning Platform for Diagnostics

The broader impact/commercial potential of this I-Corps project is in developing a simple blood test that can recognize diseases like cancer, chronic heart and lung disorders, and diabetes. To diagnose these diseases faster, the proposed technology uses a specific panel of metabolites in the blood. Metabolites change as cells change their function because of disease and could serve as a specific pattern, "fingerprint" that reports the disease.

These patterns can be recorded by mass spectrometry (MS) and recognized by artificial intelligence (AI). Unbiased identification of the patented "fingerprints" by MS and analysis using AI may produce a diagnosis. The proposed technology may not only reduce time-to-diagnosis from years to days but provide less-invasive tests for customers, replacing techniques such as biopsies, colonoscopies, and heart catheterizations.

This method will not depend on the professional qualification of health care workers and can be used at primary care facilities, hospitals, or urgent care sites.

This I-Corps project seeks to detect cancers and chronic lung and heart diseases. Initially, cancers and lung or heart diseases develop asymptomatically and so the patient does not seek professional help. Even after the onset of nonspecific symptoms, such as tiredness and fatigue, conditions often remain unrecognized for greater than two years.

There is a significant need for novel diagnostic tools to shorten the time-to-diagnosis and initiate therapy at earlier stages of the disease. Cancers and lung and heart diseases are known to significantly alter the metabolic profile. The proposed technology shows that alterations in metabolites occur when diseases are mild and no symptoms are evident.

Thus, profiling of circulating metabolites could become an efficient tool for tracing diseases at early and developed stages. The proposed patented technology detects specific changes in circulating metabolites that relate to discrete pulmonary hypertension from heart disease, colorectal cancer, and other chronic lung conditions. A mass spectrometry (MS) based platform is used for differential diagnosis to distinguish from diseases with similar, non-specific symptoms.

Three preliminary metabolic panels have been developed. Each panel comprises of ten to twelve metabolites allowing separation of a specific group of diseases with current precision greater than 75-98%.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.