Signal Processing to Improve Signal-to-Noise Ratio in Ultrasound Molecular Imaging; Perverse Incentives: Role of Healthcare Reimbursement Models and Billing Codes in Utilization of Lifestyle Medicine for Chronic Disease Prevention

Chronic diseases, most notably heart disease, stroke, cancer, and diabetes, represent some of the greatest threats to public health as the foremost contributors to premature and overall death. Despite an ongoing epidemic in these non-communicable conditions, there is good cause for optimism: in a medical landscape increasingly pervaded by technological innovation, particularly in the domains of screening and diagnosis, healthcare is poised for a revolution. In addition to breakthroughs improving early detection and management of these illnesses, a growing emphasis on preventive medicine can stop them from taking root in the first place. Both avenues for combatting chronic disease encounter serious obstacles, however. With a rapidly developing technology like ultrasound molecular imaging, for instance, there is a need for improved methods to assist physicians in gathering useful diagnostic and prognostic information. In the latter case, attempts to shift towards preventive medicine reveal barriers related to healthcare finance, culture, and infrastructure. Given these concerns, the technical report here details an attempt to improve signal processing for ultrasound of the heart in hopes of facilitating fast, non-invasive diagnosis of heart disease. The STS research, on the other hand, investigates the influence healthcare reimbursement models and billing codes have on utilization of lifestyle medicine, a type of preventive care which targets root causes of disease. Together, both projects stand to enhance understanding of chronic disease and improve treatment by addressing prevailing obstacles to care.

The basis for the technical report is to improve the signal-to-noise ratio for ultrasound molecular imaging (UMI), a relatively fast, inexpensive, and non-invasive means of diagnosing heart disease at the patient’s bedside. Critically, UMI’s effectiveness depends on the ability to isolate a signal of interest—frequently, small molecules (contrast agents) which bind to diseased heart tissue and allow clinicians to visualize damage in real time. Current methods of assessing damage are largely qualitative, however, and the current gold-standard quantitative method has shortcomings related to post-processing. As such, my capstone project undertook the task of developing a new algorithm for quantifying contrast agent in clinical videos of live tissue. We developed a robust principal component analysis (RCPA) filter and utilized motion-tracking algorithms like dense optical flow to achieve this. Computer simulations of heart blood vessels allowed us to validate the filter’s efficacy for stationary tissue, but given difficulties in accounting for tissue motion, application of the filter to clinical data requires further work.

The aim of the STS research was to understand how healthcare reimbursement models and medical billing codes drive the (under)utilization of lifestyle medicine—a medical specialty uniquely positioned to target chronic diseases by addressing their root in modifiable lifestyle factors. Addressing these factors, including but not limited to nutrition, physical activity, and substance abuse, offers a safe, evidence-based approach for preventing disease onset. Analysis of the literature indicates traditional fee-for-service reimbursement models give rise to perverse incentives by encouraging clinicians to prioritize volume and value of services rendered to a patient. Given its use of relatively low-valued services, practice of lifestyle medicine is effectively discouraged by current billing code schemes. Actor-network analysis makes further light of this issue by situating conventional and lifestyle medicine as two philosophies put at odds with each other by these market-born incentives. Despite continued difficulties in providing lifestyle-based care, billing code reforms and adoption of alternative payment models show promise for helping healthcare providers expand their practice.

Ultimately, the technical project successfully implemented some of its aims but fell short of delivering a durable algorithm that surpasses the current gold-standard method of measuring UMI signal-of-interest. Computer simulations and application to clinical data indicate the algorithm can successfully isolate signal in stationary tissue, but more work is needed for imaging the (moving) heart. The STS research, meanwhile, revealed a plethora of issues in healthcare finance that inhibit use of lifestyle medicine. Fee-for-service models, physician billing code quotas, and lack of adequately valued codes present obstacles to expanding evidence-based preventive care. The research cast light on a myriad of other interesting findings which suggest more complexity to the problem than was initially assumed. While outmoded systems of reimbursement appear to be at the core of healthcare’s slow propensity to adopt lifestyle medicine, other obstacles related to medical culture and philosophy emerged as significant co-factors. Future investigations might help lifestyle medicine grow by considering either what characterizes successful implementation of it on a practical level, or how financial obstacles differ in the U.S. relative to other nations, which may yield insights into how national policy could be reshaped.

School of Engineering and Applied Science

Bachelor of Science in Biomedical Engineering

Technical Advisor: Jonathan Lindner, John Hossack

STS Advisor: Kent Wayland

Technical Team Members: Parker Crotty