Machine Learning Algorithm for the Analysis of Cardiac Tissue Cross Sections; A Virtue Ethics Analysis of the Programmers of the Patriot Missile System

My technical work and my STS research are connected through the theme of machine learning, understanding how moral decisions can affect outcomes, and further advancement of engineering. Machine learning synergizes diverse fields such as data science, artificial intelligence, healthcare, manufacturing, and energy to optimize processes, make predictions, and extract insights from data, driving innovation across industries. This concept connects both my technical and my research projects. However, both projects are different in their exploration of machine learning. My research expands on moral decisions when integrating machine learning into target-based systems such as Missile Systems. At the same time, my technical work creates a machine learning pipeline to quantify Fibrosis and Cell Proliferation. Although my projects approach machine learning from various perspectives, they acknowledge that Machine learning is the future that expands engineering.
My technical project explores using machine learning to quantify Fibrosis and Cell Proliferation. Cell proliferation is the process of cells dividing and increasing in number. At the same time, fibrosis is the excessive accumulation of fibrous connective tissue in an organ or tissue, often resulting from chronic inflammation or injury. My capstone group created a machine learning pipeline using software like Ilastik, Cell Profiler, and ImageJ to quantify these metrics from cross-sectioned cell images. The pipeline was built to automate image analysis of hypertrophy, DNA content, and nuclear segmentation. This pipeline will be used continuously in the Saucerman and Wolf lab to further aid in high throughput drug screening of mice involved in vivo experimentation with hypertrophic cardiomyopathy. The ultimate goal of this pipeline is to increase the speed at which researchers can iterate their experiments and reduce potential bias. As a result, this increase will enhance drug testing and streamline the process of data quantification.
My STS research explores the application of machine learning from a narrowed perspective. My research delves into the moral implications of using Machine learning in complex targeting systems and explores the morality of actors when making decisions to construct and use the system. Virtue ethics is a moral philosophy centered on cultivating virtuous character traits, such as honesty and courage, as the foundation for ethical decision-making. My claim is that the programmers of the missile system I analyzed lacked morality due to 3 failures in upholding virtue ethics. My paper expands using the framework based on failures in showing professionalism, safety, and proper documentation of classification and identification. My research aims to draw attention to the importance of character manifesting in engineering endeavors.
Working on these projects in parallel was a transformative experience. My technical work provided insight into machine learning integration, while my STS paper enhanced my appreciation for the role of an engineer's character in making systemic engineering decisions. This interdisciplinary approach allowed me to grasp the intricacies of Machine learning at a higher level, enhancing my decision-making skills when creating a pipeline. Ultimately, it made me a more morally conscious engineer, capable of making ethical decisions with Machine learning and developing a machine learning pipeline.

School of Engineering and Applied Science

Bachelor of Science in Biomedical Engineering

Technical Advisor: Jeffrey Saucerman

STS Advisor: Benjamin J. Laugelli

Technical Team Members: Marcus Elward