Other Biomedical Engineering and Bioengineering Commons^™

Adversarial Robustness In Biomedical Time-Series Models, Rohan Tiwari

Bioengineering Theses

This study investigates adversarial vulnerabilities in deep learning models for biomedical time-series classification across two clinically important modalities: electrocardiography (ECG) and electroencephalography (EEG). Using the MIT-BIH Arrhythmia and CHB-MIT seizure datasets, I evaluate time-domain attacks (FGSM, PGD), Fourier-domain constrained attacks, and learned spectral perturbations designed to reveal modality-specific sensitivity patterns. Across both tasks, a consistent trend emerges low-frequency components (0–5 Hz) constitute a dominant axis of adversarial vulnerability, with perturbations in this range producing the steepest degradation in classification performance. In ECG models, protecting the physiologically relevant QRS band (5–20 Hz) significantly improves robustness, whereas EEG models remain highly sensitive …

Human System Composite Performance Measures: General Systems Performance Theory Insights, Lawrence Disalvi

Bioengineering Dissertations - Archive

Composite Performance Measures (CPMs) of human performance are widely employed in applications that range from mundane to life-changing, yet they are frequently presented without explanation or reference to theoretical underpinnings. The conceptual soundness of a widely used CPM approach, based on additive combination of constituent components, has been called into question by Kondraske. He argues the relevance of General Systems Performance Theory (GSPT) and performance capacity envelope concepts when applied to multidimensional CPM measures.

To obtain evidence-based answers to assertions regarding CPM types and rationale employed in medical and non-medical contexts, a Systematic Descriptive Review was conducted. Separately, the GSPT-based …

Algorithm Development For Analysis Of Optical Time-Series Data Generated By Biomodal Optical-Electrical Nanopore Sensor, Yu-Shiuan Huang

Bioengineering Theses - Archive

Nanopore biosensors have played an essential role in the scientific investigation of DNA, RNA, proteins, and other bio-analytes. Traditional nanopore biosensors detect variations in electrical current conductance caused by the analyte blocking the current passing transiently through a nano-sized aperture, i.e., the nanopore. This electrical data provides information about the amount of current blockage when the analyte is inside the nanopore and the time it takes for analytes to travel through the nanopore (translocation time). The translocation time of analytes in standard nanopore measurements is typically very short, usually tens of microseconds, limiting the accuracy of measuring the analytes’ characteristics. …

An In-Vitro Model For Visualizing Intraocular Lens And Lens Capsule Interaction In Posterior Capsule Opacification, Amjad Chatila

Bioengineering Dissertations - Archive

Posterior Capsule Opacification (PCO) is a common complication post cataract surgery. It is believed that Intraocular Lens (IOL) physical and chemical properties would affect their PCO potential. The tight binding between IOL and posterior lens capsule (PLC) is believed to have a negative impact on PCO formation, as suggested in the “No space, no cell, no PCO” hypothesis. To test the hypothesis, we have recently developed an in vitro 3D system to simulate IOL and PLC physical interaction. To expand the function and capability of this system, one objective of my work was to develop the next generation adhesion force …