Development of a Noninvasive Comprehensive Digital Tool for Modeling the Structure and Function of the Small Intestine
Small intestinal diseases affect patients throughout their entire lifespan. While the clinical burden of small bowel pathology is astonishingly high, there are no current methods to directly, noninvasively quantify small intestine function. Clinically, small intestine function is inferred using surrogate markers, but more direct assessment of the small intestine physical properties are achievable using novel quantitative techniques. Our overall purpose is to develop a validated, comprehensive digital tool that enables non-invasive, subject-specific measurements of small intestine physical properties. In patients with intestinal failure, this tool will allow us to predict patients’ ability to grow and thrive without parenteral nutrition and to predict whether small bowel anatomy is amenable to, and the likely success of, surgical procedures.
Decision Support Tool for Triage of Trauma Patients: Using Machine Learning to Take the Complexity out of Trauma Triage
Approximately 10-15% of all traumatically injured children sustain life-threatening injuries that require a systematic and rapid approach to their treatment. When a trained trauma resuscitation team is present on the arrival of the seriously injured child, mortality can be decreased by 25%-30%. Undertriage is defined as a triage decision that classifies a patient as not requiring a trauma team activation, when in fact they do. Undertriage is a medical problem, which may result in adverse patient outcomes. Overtriage occurs when a trauma team activation is triggered despite the patient not meeting criteria for such an activation. Overtriage is a resource utilization problem. The American College of Surgeons views overtriage and undertriage rates based on trauma team activation criteria as surrogate markers for quality trauma patient care.
We are conducting exploratory studies aimed at demonstrating the feasibility of a non-invasive, population-based predictive model for trauma triage. Our ultimate goal is to develop a prospectively validated clinical decision support tool that can be used in conjunction with clinical judgement from human providers to make the best trauma triage decisions possible.
Using Fractals to Investigate How Tissue Structure and Organization Promote Vasculogenesis in 3D Engineered Tissues
The fractal concept is a mathematical model initially used to characterize the complex geometry of nature, whose irregularity and complexity cannot be readily simulated using traditional linear geometry. We are currently using tissue engineering constructs, such as cellular arrays in hydrogels as an archetype system to demonstrate and illustrate self-organization and emergence, as characterized using fractal mathematics. We are also investigating how the organizational properties of these cellular constructs may affect vasculogenesis in 3D engineered tissues.