The proposed project involves development of a user-interactive software suite for medical image analysis. The core algorithms consist of simultaneous segmentation and registration based on active contour and active surface models. This software is designed to deliver a practical, intuitive, and easy-to-use desktop software tool to doctors, clinicians, and other health-care professionals to extract geometric tissue features (segmentation) from collected image data (including modalities such as MRI, CT, PET, etc.) while simultaneously aligning such geometric structures across separately collected datasets (registration). Included in the software application will be tools to visualize and quantitatively measure the extracted geometric information, thereby providing heavily sought information to aid in diagnosis, treatment and surgical planning, as well as evaluation of trials and studies.
Current clinical practices in medical image analysis are labor intensive and time consuming. The proposed software suite has the potential to be a powerful tool that may reduce the time from current several work days to less than an hour. If successfully developed, this tool could enable higher patient throughput and assist in extracting accurate diagnostic information from medical images. The technology finds application in cardiac function analysis, structural neuro-imaging, planning cardiac resynchronization therapy, and planning radiation therapy.
Our I-Corp team was selected for this NSF award based on a clinically oriented user-interactive 3D image segmentation and registration software tool (based on previously funded NSF research by the PI Yezzi). Our team's initial ideas for commercialization of this technology was to develop a stand-alone desktop software package that could be purchased by clinicians at hospitals, clinics, and medical imaging research labs. As a result of the I-Corp training activities (the kick-off workshop and the weekly online courses and customer discovery process) we learned that the market for this product was already limited and croweded with competing products that we originally were unaware of. As such, throughout the first month of this project, our team underwent a major pivot and found a way to use the same underling computational algorithms for a very different clinical application, namely a computer-vision based balloon spirometery technique for which there were no other similar competing products. In the final months of the project, we developed a prototype of the software component of the spirometer project, and continued to refine the original product. Plans are to commercialize the new technology through a newly created company, Vintinura Imaging Inc. (founded in part by the PI) and to jointly develop the original application of the technology (segmetation/registration) with another company Syntermed which already has a related product in the market and therefore has already captured a significant part of the available market share. We will therefore benefit from their already existing customer base and distribution channels as a result. This will help us, in turn, continue to complete the prototype of the new spirometry application which we plan to commercialize completely on our own.
