This Small Business Innovation Research Phase I project establishes the feasibility of Computer Aided Prognosis of Debilitating Disease. Because disease arises through a complex interaction of multiple molecular signals and pathways often confounding the eventual effect, tools and approaches are needed to identify key pathways that reflect the underlying pathological processes. While both functional imaging modalities have recently emerged, the computational tools that would allow for accurate analysis of these imaging modalities in order to allow for prediction of therapy discovery, development, disease stratification, and personalized medicine are sorely lacking. Previous approaches rely on identifying one or a relatively small number of distinguishing features hypothesized to be precursor to an acute event. CAP seeks to build on this by providing functional characterization that extends the static diagnostic categorization to prognosticate the likely future progression. The research objective is to develop an integrated segmentation, registration, and classification toolkit for prognosis prediction of vascular disease from dynamic time-series imaging data. The goal of Phase I research is a software endpoint. We demonstrate probable clinical utility by the successful extraction of values that meet or exceed the manually produced preliminary studies when assessed on the available animal and human data sets.
The broader impact/commercial potential of this project is to develop methods for identifying prognostic imaging signatures of disease aggressiveness and predicting potential patient outcome so as to improve it. The task of distinguishing which subtypes of vascular lesions will have favorable outcome as opposed to unfavorable outcome requires sophisticated image analysis and quantification and feature characterization algorithms to accentuate the subtle imaging differences between these related pathologies. Scientific and technological understanding of how dynamic aspects of disease progression may be discerned from higher-order processing which optimizes information content from imaging assays. This technology represents a cost effective, safe and capable plaque assessment tool, so that patients could be treated more effectively, sooner, and more appropriately. This project creates an end-user capable prototype that may also be extended in preparing first a 510(k) and subsequently a PMA application as a prognostic for individual patient management. vascuVis will support the market initially by selling software licenses and later by developing a pay per use business model. Commercially, over 20,000 MRI units installed worldwide could benefit from this product. At $75K average pricing, the opportunity is as high as $1.5B. The total accessible market for this product could be as high as $1.5B.
This Small Business Innovation Research Phase I project optimized a computerized image analysis toolkit, vascuCAPTM, in the context of longitudinal, time series imaging data and evaluate it in the specific context of identifying atherosclerotic plaques on DCE-MRI that are "vulnerable" to disruption and thrombosis and which generally do not produce clinical symptoms and rate-limiting stenosis, in advance of a fatal or debilitating event. vascuCAPTM analyzed quantitatively and objectively the verifiable biophysical hallmarks that, taken in composite, present a biomarker predictive of atherosclerotic plaque instability. Physicians face a complex and heterogeneous series of clinical manifestations of disease. Because disease arises through a complex interaction of multiple molecular signals and pathways often confounding the eventual effect, tools and approaches are needed to identify key pathways that reflect the underlying pathological processes. While both functional imaging modalities have recently emerged for characterization of these disease processes and to obtain a better mechanistic understanding of the underlying biologic processes to distinguish more aggressive from less aggressive disease phenotypes, the computational tools that would allow for accurate analysis of these imaging modalities in order to allow for prediction of therapy discovery, development, disease stratification, and personalized medicine are sorely lacking. Computer-aided prognosis (CAP) of disease is a new and exciting complement to the field of computer-aided diagnosis (CAD) and involves developing and applying computerized image analysis and multi-modal data fusion algorithms to digitized patient data (e.g. imaging, tissue, genomic) for helping physicians predict disease outcome and patient survival. Since CAP approaches distinguish between different subtypes of a particular disease (as opposed to CAD schemes trying to distinguish diseased from benign processes), there is a need for more sophisticated image analysis, computer vision, and machine learning methods to identify subtle disease signatures that can separate unstable from stable disease. While CAD tools have been developed for tumor detection and diagnosis in the context of lung, breast, brain, and prostate cancers, there has been little work in developing corresponding CAP methods for identifying prognostic imaging signatures of disease aggressiveness and subsequently predicting patient outcome. The task of distinguishing which pathology will have favorable outcome as opposed to poor outcome requires more sophisticated image analysis approaches than required to address detection and diagnosis. The chosen application in this completed project is a use case which has the potential to radically increase the power of applications to support clinicians in pursuit of personalized medicine.
