At CVIB, we are creating computer vision that understands the real world and can be trusted in clinical practice
Scientists at CVIB are developing a superhuman machine intelligence that can see and understand medical images as we do. It goes beyond the unintelligible data models from current machine learning using a unique Cognitive AI approach of semantically embedded neural networks. This new machine intelligence can be taught concepts, can learn from data, and can evolve to reach and then exceed human level vision. It can tell us what it knows, and explain its decisions. This vision system is original, it can adapt to new problems and create news solutions without an engineer or data scientist. The resulting vision system is superhuman in its quantitative decision making, consistency, scalability, and unlimited ability to absorb more data.
Computer vision enables precision medicine, where new therapies are evaluated and the right treatment matched to the right patient
Quantitative image analysis provides a vital decision support tool. We are committed to partnering with clinical scientists and providing imaging services for precision medicine. Our computer vision infrastructure provides a unique capability for large scale image analysis and quantitative feature extraction for biomarker development, validation, and clinical translation.
The goal of these biomarkers are to evaluate new therapies and inform patients and their physicians on the best treatment option tailored specifically to them.
We measure success not based on lab benchmarks, but on benefits actually delivered to patients
We believe that the success of AI should be measured in improvements in patient outcomes and broad impact in routine health care. Technology does not benefit patients unless it is translated and validated in clinical practice. Machine learning algorithms must be hardened to handle the variability of real world imaging and integrated carefully into physician workflows at the point of care, biostatisticians, and healthcare IT. CVIB welcomes projects led by collaborators with goals defined and measured in terms if patient benefits and outcomes.