The rapid advancement of Artificial Intelligence (AI) has significantly transformed healthcare, enhancing traditional methods of diagnosis and treatment. These innovations have enabled quicker disease detection, improved management, and more personalized care. However, many AI tools currently used in clinical settings suffer from algorithmic and data-driven biases, often due to inadequate representation of certain racial, gender, and age groups. These gaps can result in misdiagnoses, health disparities, and inequitable outcomes. Therefore, addressing these biases is essential.

This project investigates the presence and impact of such biases by examining both pre-processing and post-processing stages of AI model development, using a widely adopted real-world healthcare dataset from primary care patients. It uncovers previously overlooked biases and offers practical strategies to reduce disparities related to race, gender, and age. By applying machine learning algorithms and utilizing the Fairlearn toolkit, the study identifies and quantifies the biases, evaluates their effects on predictive performance, and presents methods to mitigate them. The findings provide strong evidence of systemic bias in healthcare AI systems, particularly in how they influence resource distribution and decision-making. As a result, it is imperative to incorporate bias detection and mitigation techniques to ensure that AI technologies in healthcare are fair, dependable, and ethically sound.

Severe burn injuries often trigger musculoskeletal complications that extend far beyond skin damage. Many survivors develop “hidden” bone-related conditions such as fractures, bone mineral density loss, and osteomyelitis, which may not be easily detected through routine clinical assessment. Because these complications can progress silently and lead to long-term skeletal deterioration, early prediction and targeted monitoring are essential for preserving mobility and quality of life. This project presents a multimodal framework that combines clinical predictive modeling with advanced imaging analysis to evaluate bone health complications in burn patients. The clinical component analyzes an AI-generated synthetic dataset containing 1,538 burn patient records using machine learning and principal component analysis (PCA) to identify key factors associated with fractures and osteomyelitis. The imaging component applies deep learning–based feature extraction using ResNet-50, followed by Uniform Manifold Approximation and Projection (UMAP), to explore structural variation across a 72-slice Computed Tomography (CT) scan derived from a larger dataset of over 5,000 CT images of fractured limbs.

Preliminary results indicate that burn severity index, bone damage score, duration of hospital stay, comorbidity score, and treatment material properties are the strongest predictors of fracture risk and osteomyelitis development. Additionally, the CT slice embeddings form distinct anatomical clusters, demonstrating that the model captures meaningful structural variation within bone regions. A bone segmentation step further enhances this analysis by isolating bone structures and reducing soft-tissue variability in imaging features.

Overall, this multimodal approach highlights the complementary value of clinical and imaging data in understanding burn-related skeletal complications and is hoped to advance early detection strategies for these conditions. While the imaging analysis remains exploratory, the findings suggest that deep-learning based CT features may support the future development of integrated models for improved bone health assessment in burn patients.

The human voice, a biomarker of complex movement of communication has been known to change for a person as they age or have changes in health.  Neurological, cardiovascular, and respiratory disorders are disease that can alter the acoustic features of voice. Fluid accumulation in vocal fold, fatigue can lead to change the features of voice such as pitch, formant, jitter, shimmer and higher order features such as MFCCs. 

In this study, we aim to distinguish the feature change between patients with Neurological, cardiovascular, and respiratory disorders and have multiple comorbidities. We have used Bridge2AI voice dataset, which consists of 302 subjects with the disorders. We have analyzed 140 acoustic features to investigate significant difference between diseases and found several features were different.

We have further implemented machine learning algorithms to identify the patients (for both subject dependent and independent cases) from their voice features and achieved F1 score above 0.60. We aim to enhance this study in future to increase the accuracy level.

This capstone project explores the relationship between anxiety and lung cancer diagnosis, addressing a gap in research where psychological factors have often been overlooked. Using survey data from Kaggle and the UCI Machine Learning Repository, I built a logistic regression model to see if people who report experiencing anxiety are more likely to also report a lung cancer diagnosis. After accounting for factors like age, gender, smoking habits, and chronic disease, the results demonstrated that individuals reporting anxiety were over twice as likely to report having lung cancer. Overall, the model demonstrated solid performance, with consistent results observed across key demographic and behavioral subgroups, including smokers and individuals over 60. These findings suggest that mental health, particularly anxiety, may play a larger role in physical health outcomes than we often consider. By bringing psychological factors into the conversation, this research encourages a more holistic approach to how we think about cancer risk.