Signal Processing to Improve Signal-to-Noise Ratio in Ultrasound Molecular Imaging; Innovation and Inequality: Intellectual Property in the Pharmaceutical Industry

The rapid pace of medical innovation has dramatically improved health outcomes across the globe. Advances in diagnostic imaging, pharmaceuticals, and surgical techniques have enabled longer and healthier lives, transforming once dreadful diseases such as HIV or pediatric lymphoma into manageable conditions with high survival rates. These improvements, however, come with complex ethical, economic, and logistical challenges. Central to this tension is the role of intellectual property in the development and accessibility of medical technologies. While patents are essential for incentivizing research by ensuring return on investment, they can also limit access to life-saving treatments, especially in areas where brand-name drugs are 100-fold the price of generic drugs. At the same time, many technological advances are not radical breakthroughs but iterative improvements. One such example is a novel diagnostic technique using targeted ultrasound contrast agents and computer vision algorithms to identify damaged cardiac tissue after a heart attack. Advancements like these will slowly improve the medical landscape. Together, these topics demonstrate the need to balance innovation with access to medicine for the future.
My STS paper examines the debate between whether or not intellectual property should exists for pharmaceuticals. I use the Co-Production theory to demonstrate the interconnectivity between the economic arguments and the ethical arguments. The pharmaceutical industry is a $1.6 trillion industry with many pharmaceutical drugs requiring millions to billions of dollars to develop. However, because of issues with affordability, innovation incentives, and economic theory, the existence of patents remains a hot-topic. The main arguments supporting the existence of patents cover the high barrier to entry industry landscape, monopoly-incentives, and property rights. Many of these arguments can be explained by utilitarian logic described by Jeremy Bentham or natural rights described by John Locke. Those that argue against the existence of patents point out problems with the status quo such as affordability problems, the nature of ‘me-too’ patents, and the source of funding. These provide contradictions in the natural rights theory while also questioning the fundamental economic theories. With this paper I hope to provide the reader with a holistic understanding of the landscape of pharmaceutical patents. Because of the scale of the industry in the United States, an educated voting population could possible make significant changes in the world of medicine.
Echocardiographic memory molecular imaging is a promising, non-invasive, and cost-effective technique for diagnosing acute coronary syndrome (ACS) at the bedside. This method uses lipid microbubbles innervated with molecules that bind to ishemic vascular tissue that are subsequently imaged through ultrasound. These microbubbles appear as signal in ultrasound accurately locating the diseased tissue. However, only a fraction of the microbubbles adhere to the ischemic tissue and the rest remain in circulation, resulting in difficulty discerning the adhered microbubbles from freely circulating. To overcome these limitations, we developed a novel algorithm aimed at improving signal-to-noise ratio (SNR) and enabling real-time detection of retained microbubbles. We implemented a robust principal component analysis (RPCA) filter combined with a motion-tracking algorithm in both simulated and clinical ultrasound datasets. Using computer simulations, we modeled microbubbles in vasculature and showed SNR improvement from 103.3% to 182.5% after applying the filter. Similar results were achieved with in vivo liver and kidney tissues. Development of a dense optical flow algorithm followed, but encountered obstacles related to accuracy and computational demand. The results compiled reflect favorably on the potential of RPCA for assessing bubble adhesion in stationary tissue, but the difficulty of reliable motion-tracking remains a significant challenge for real-time application in the contracting and relaxing left ventricular myocardium. Continued efforts should explore alternative computer vision techniques and conduct comparative analysis of RPCA with related filters. Successful integration of such a design with clinical software suggests significant potential for improving the speed and precision of ACS diagnosis.
With my STS paper I hope to demonstrate the different sides of a debate of importance to my future field of work, highlighting to other engineers the unique context of the medicine industry where sometimes philosophies clash. However, there is always more to discuss; specific drug contexts, more nuanced patent laws, and international enforcement interactions can provide additional insight for each side. And with my technical project, we successfully demonstrated a particular application of computer vision on an existing diagnostic technique to create an alternative and improved version. With many engineering projects, further improvements are always warranted; runtime-optimizations and user experience are challenges that future engineers could accomplish.

School of Engineering and Applied Science

Bachelor of Science in [Insert Major] [Include only first major on transcripts]

Technical Advisor: Johnathan Lindner, John Hossack

STS Advisor: Kent Wayland

Technical Team Members: Ryan Warnick