SBIR Phase II: An adaptive machine learning-based platform to improve surgical quality and patient outcomes

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase II project will be to help usher in personalized and tailored surgical care within a shifting healthcare context toward value-based care. Hospitals and surgeons are seeking solutions that will enable them to target, as opposed to generalizing, improvements in surgical quality for enhanced patient outcomes and effective use of resources. By proactively identifying surgical risks and matching patients to interventions most appropriate for these risk strata, the proposed technology is designed to support hospitals in meeting their value-based care objectives. The larger vision is to apply this paradigm in all of medicine by leveraging Artificial Intelligence and Machine Learning for prediction, proactive intervention, and outcomes tracking in a closed feedback loop. Demonstrating this in a high-cost, high-risk specialty like surgery provides a path for expanding the technology into other medical specialties and serving a greater domestic and international market. Ultimately, the lessons learned from the wide-spread use of this technology will allow society to derive key kernels of knowledge in applied data science, preventative medicine, and technical scalability of hospital enterprise solutions. This project is an interdisciplinary representation of crucial activities needed to drive the tipping point of medical technology. This Small Business Innovation Research (SBIR) Phase II project builds upon the results of Phase I, which included predictive engine development, scalable data processing pipeline development, and hospital stakeholder engagement activities. Phase II efforts focus on further developing the technology to facilitate its commercial use and integration in clinical settings. Key objectives for the Phase II project are as follows: (1) development of an Application Programming Interface (API) to deliver tailored machine learning models to broad users across varying needs, (2) expansion of a clinical intervention library supported by clinical evidence across multiple surgical specialties, and (3) development of an outcomes dashboard to display postoperative patient outcomes from automated extraction of electronic health records. The result of this project will be a closed-loop clinical and technical infrastructure that is agile to the needs of a diverse range of surgical customers to enable quality improvement across an entire surgical ecosystem. 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.