Analytical, Diagnostic and Therapeutic Techniques and Equipment Commons^™

Id-More Vision: Real-Data Machine-Learning Assessment For A Digital-Twin-Inspired, Xr-Ready Rehabilitation Prototype, Rickey L. Clark

Master's Theses

Rehabilitation assessment often relies on periodic observation, while many XR prototypes show scripted rather than recorded-motion evidence. iD-MORE Vision is an offline pipeline trained on KIMORE and IRDS and linked through JSON packets to a two-mode Unity desktop prototype. Both datasets include controls and rehabilitation participants with neurologic, musculoskeletal, or mobility impairments. This improves relevance but does not clinically validate the system.

Under fixed subject-wise splits, the primary five-seed Random Forest predicted KIMORE clinician scores with MAE 6.087 ± 0.044 cTS and R² 0.568 ± 0.006; the subject-level R² interval crossed zero. The primary IRDS five-run CUDA GRU averaged 0.877 …

Evaluating Chest-Worn Light Logger Adherence: Development And Application Of A Wear/Non-Wear Model, Carlyn Patterson Gentile, Adah Thomas, Ryan Shah, Blanca Marquez De Prado, Nichelle Raj, Christina Szperka, Geoffrey Aguirre

Student Papers, Posters & Projects

Light exposure plays an important role in overall health because it entrains circadian rhythms. Recent technological advances in wearable light loggers allow measurement daily light exposure habits. Using a chest-worn light logger, our goal was to (1) develop methodology for differentiating adherent versus non-adherent use, and (2) define differences in lighting intensity in indoor and outdoor environments, to improve data reliability in future clinical studies using this technology. Four testers used a 10-channel chest worn light logging device under different conditions of wear and non-wear (experiment 1), and another tester made measurements with the light logger across a variety of …

Hyposense: Multimodal Physiologic And Voc Wearable For Early Hypoglycemia Detection, Nidhi R. Patel

Honors Theses

Hypoglycemia remains a significant safety concern for insulin-dependent individuals, especially when blood glucose declines rapidly during exercise, sleep, or periods of impaired hypoglycemia awareness. Current continuous glucose monitors and automated insulin delivery systems have helped to improve diabetes management, but these technologies primarily respond to glucose trends after they have already started and may not actually provide sufficient warning before symptoms occur. This thesis describes the development of HypoSense, a noninvasive wrist worn wearable concept that is designed to support earlier hypoglycemia risk detection through multimodal physiologic and volatile organic compound emission sensing. HypoSense combines volatile organic compounds/gas sensing with …

Spectral Characterization Of Colorectal Cancer Using Excitation-Scanning Hyperspectral Imaging, Natalie Hadad

Honors Theses

Colorectal cancer is one of the most common and deadly cancers worldwide, with a lifetime risk of approximately one in twenty-three for men and one in twenty-five for women. Although early detection significantly improves survival, current diagnostic approaches such as tissue biopsy are invasive, reliant on subjective interpretation, have extended turnaround times, and risk missing flat or depressed lesions that are more difficult to detect than raised polyps. These limitations highlight the need for objective, minimally invasive technologies capable of detecting early tissue changes. This study investigates excitation-scanning hyperspectral imaging (Ex-HSI) as a label-free optical approach for identifying colorectal cancer–associated …

An Investigation Of Data Granularity In Rag Pipelines For Personalized Medicine, Paritosh Pandey

Master's Theses

Generative AI, exemplified by large language models like the OpenAI GPT and Meta LLaMA families, can produce diverse content in response to prompts. This capability offers a promising solution to challenges in precision medicine, which seeks to tailor treatments to individual clinical profiles but often struggles with data collection, cost, and privacy concerns. By generating realistic, privacy-preserving patient data, generative AI has the potential to transform patient-centric healthcare. With such motivation, this research develops a comprehensive Generative AI pipeline emphasizing data granularity for accurate prediction of personalized treatments. The pipeline features a central Large Language Model interacting with a Machine …

Quantitative Fluorescence Imaging Using A Paired Agent Approach For Improving Surgical Decision Making, Sanjana Pannem

Dartmouth College Ph.D Dissertations

Surgical resection remains to be the cornerstone of treatment for most solid tumors, where the primary objective is to remove all visible and microscopic disease. However, the extent of tumor resection is inherently limited by the surgeon’s ability to distinguish tumor from normal tissue, intraoperatively. As a result, positive surgical margins (PSM), where residual tumor remains after surgery, are a common occurrence and are associated with increased recurrence, need for adjuvant therapies, dismal patient outcomes, and high healthcare costs.

Fluorescence guided surgery (FGS) has emerged as a promising technique to enhance intraoperative visualization of tumors and may help reduce the …

Robust Non-Invasive Cardiac Index Prediction Via Feature Integration And Data-Augmented Neural Networks, Chih-Hao Chang, Mei-Ling Chan, Yu-Hung Fang, Po-Lin Huang, Tsung-Yi Chen, Tsun-Kuang Chi, I Elizabeth Cha, Tzong-Rong Ger, Kuo-Chen Li, Shih-Lun Chen, Liang-Hung Wang, Jia-Ching Wang, Patricia Angela R. Abu

Department of Information Systems & Computer Science Faculty Publications

Concurrent with the rising consumption of ultra-processed, high-calorie diets and the decline in physical activity, obesity and related cardiovascular conditions among young adults have continued to increase, becoming an important global public health concern. This study integrates non-invasive Internet of Things (IoT) sensing devices, including the TERUMO ES-P2000 blood pressure monitor (Terumo Corp., Tokyo, Japan) and the PhysioFlow PF07 Enduro cardiac hemodynamic analyzer (Manatec Biomedical, Poissy, France), with an artificial neural network (ANN) for cardiac index (CI) prediction. Through appropriate data preprocessing and model training strategies, the generalization ability and stability of the proposed CI prediction model were significantly enhanced. …

Humanity Is Evolving Its Consciousness: The Role Of Archetypal Energies As Guides During An Unfolding Weeding Out And Alignment Process, Carroy U. Ferguson

Psychology Faculty Publication Series

Humanity is evolving its consciousness at individual and collective levels.  Given these seemingly tumultuous times, as of this writing (January 2026), to make such a statement may sound like a strange thing to say.  However, I suggest that if you are alive today and if you are reading these words, these are the very times for which you were born—to assist Humanity as it evolves its consciousness with your unique gifts, whatever they may be.  That is, this period of our individual and collective human being-ness may be characterized as an unfolding period of weeding out and alignment with the …

Implementation And Clinical Utility Of Ultra-Low-Field Portable Magnetic Resonance Imaging For Postprocedural Neurological Evaluation In Ambulatory Neurosurgery: Illustrative Cases, Devan Patel, Vinay Jaikumar, Taysia P. T. Morioka, Laz Rifkin, Kenneth S. Jacoby, Jaims Lim, Anais Andrade, Aimee C. Degaetano, Pui Man Rosalind Lai, Elad I. Levy

EVMS School of Health Professions Faculty Publications

BACKGROUND

Elective endovascular neurosurgical procedures are increasingly performed in ambulatory neurosurgery centers, enabled by advances in catheter technology, safety of conscious sedation, and refined patient selection. Although complication rates are low, rapid evaluation of postprocedural neurological deficits remains critical. Conventional MRI is often impractical in outpatient or procedural settings, whereas ultra-low-field portable MRI (ULF-pMRI) systems (such as Swoop) allow bedside imaging with favorable diagnostic performance.

OBSERVATIONS

Two women in their 60s developed acute neurological deficits at an ambulatory neurosurgery center (ANSC) after diagnostic cerebral angiography in one case and elective internal carotid artery flow diversion in the other. In both …

Mechanical-Medical Convergence In Heart Failure: Artificial Intelligence, Finite-Element Modeling, And 3d Printing For Diagnosis And Prognosis, Quazi Noor E. Sabrina, Quazi Md Zobaer Shah, Quazi Noor E. Sohela, Md Mahabub Hasan Mousum, Md. Moyeen Uddin Chisty, Quazi Md. Akbar Shah

Mechanical & Aerospace Engineering Faculty Publications

Heart failure remains a leading cause of global morbidity and mortality, yet routine clinical indices often miss the regional biomechanical disturbances that drive progression and shape treatment response. This State-of-the-Art review examines how finite-element (FE) modeling, additive manufacturing, and artificial intelligence (AI) are converging to improve the diagnosis, phenotyping, procedural planning, and prognostic assessment of heart failure (HF). Although these technologies have matured in structural heart disease and transcatheter intervention research, their greatest translational potential may lie in HF, where patient-specific ventricular remodeling, myocardial stress–strain heterogeneity, valve-ventricular coupling, and device-tissue interaction are incompletely captured by conventional clinical indices. We synthesize …

Advances And Challenges In Digitally Connected Point-Of-Care Biosensing, Abdellatif Ait Lahcen, Jegan Rajendran, Gymama Slaughter

Center for Bioelectronics Publications

Point-of-care (POC) biosensors are undergoing a paradigm shift from isolated diagnostic tools to digitally connected, intelligent platforms that enable continuous and decentralized healthcare delivery. This review critically examines recent advances in wearable, implantable, and portable biosensors, highlighting how integration with wireless communication, the Internet of Medical Things (IoMT), and artificial intelligence is transforming their functionality and clinical utility. Particular attention is given to innovations such as smartphone-enabled interfaces, cloud-based analytics, and machine learning-assisted analysis, which collectively enhance sensitivity, specificity, and user accessibility across diverse healthcare settings, from personalized home monitoring and bedside diagnostics to deployment in resource-limited regions. The review …

Recent Advances In Triboelectric Nanogenerators For Biomedical And Cardiovascular Monitoring, Amit Sarode, Jegan Rajendran, Gymama Slaughter

Center for Bioelectronics Publications

Triboelectric nanogenerators (TENGs) have emerged as versatile self-powered platforms for wearable and implantable biomedical sensing, offering an alternative to battery-dependent electronic devices. By converting biomechanical energy from physiological motion into electrical signals, TENGs enable simultaneous energy harvesting and active sensing within flexible, lightweight, and biocompatible architectures. This review summarizes recent advances from 2020 to 2025 in triboelectric nanogenerator (TENG)-based cardiovascular monitoring. The discussion focuses on material systems, device configurations, sensing mechanisms, and applications including pulse detection and cuffless blood pressure estimation. Representative studies are compared to highlight emerging trends in wearable and self-powered sensing technologies. However, differences in experimental conditions, …

Modern Potentiostat Architectures For Electrochemical Sensing: Design, Integration, And Future Directions, Reagan Aviha, Gymama Slaughter

Center for Bioelectronics Publications

Potentiostats are essential to electrochemical sensing, enabling precise control of electrode potentials and measurement of current responses. As demand grows for portable, wearable, and point-of-care systems, potentiostat design has evolved from benchtop instruments to compact, low-power, and wirelessly connected platforms. This review provides a comprehensive, system-level perspective on modern potentiostat architectures, covering operational principles, analog front-end design, signal generation and acquisition, communication protocols, and software integration. Unlike prior reviews that treat these aspects independently, this work integrates electrochemical theory with electronic design and data communication frameworks. Key components, including operational amplifiers, transimpedance amplifiers, DAC/ADC subsystems, and microcontroller-based control, are examined …

Consensus And Controversies Of International Guidelines For The Diagnosis, Surveillance, And Management Of Fetal Growth Restriction: An Updated Comparison, Daniele Diane Mascio, Suneet P. Chauhan, Tullio Ghi, Asma Khalil, Juliana G. Martins, Sara Sorrenti, Tamara Stampalija, Fabrizio Zullo, Francesc Figueras

Department of Obstetrics & Gynecology Faculty Publications

OBJECTIVE: To compare areas of consensus and disagreements across contemporary international and national guidelines on the diagnosis, surveillance, and management of fetal growth restriction (FGR).

DATA SOURCES: Electronic searches of MEDLINE from database inception up to March 2026 using MeSH terms and keywords related to FGR and guidelines. STUDY ELIGIBILITY CRITERIA: Critical, structured comparison of national or international guidelines on FGR published since 2010. Final inclusion required unanimous agreement from all authors.

STUDY APPRAISAL AND SYNTHESIS METHODS: Pre-specified extraction across domains: definition; prediction/prevention; surveillance tools and frequency; delivery timing and mode; and labor induction methods. Dual data …

Improving Medical Diagnostics With Vision-Language Models: Convex Hull-Based Uncertainty Analysis, Ferhat Ozgur Catak, Murat Kuzlu, Taylor Patrick, Michel Audette

Engineering Technology Faculty Publications

In recent years, vision-language models (VLMs) have been applied to various fields, including healthcare, education, finance, and manufacturing, with remarkable performance. However, concerns remain regarding VLMs' consistency and uncertainty, particularly in critical applications such as healthcare, which demand a high level of trust and reliability. This paper proposes a novel approach to evaluate uncertainty in VLMs' responses using a convex hull approach on a healthcare application for visual question answering (VQA). For any VLM, temperature refers to a sampling parameter used in probabilistic generation, which controls the randomness of the model's output. The LLM-CXR model is selected as the medical …

Injectable Hydrogels For Bone Regeneration: Mechanical Reinforcement Strategies Using Nanoparticles And Nanofibers: Review Paper, Fariba Ganji, Morteza Mirzagoli, Lobat Tayebi

Electrical & Computer Engineering Faculty Publications

Background and Purpose: The growing demand for bone regeneration following severe injuries highlights the importance of scaffolds in bone tissue engineering (BTE). Injectable hydrogels have emerged as promising candidates because their properties closely mimic the native extracellular matrix (ECM). However, their limited mechanical strength and structural instability restrict their practical application. Approach: This review summarizes recent strategies for reinforcing in situ-forming injectable hydrogels to improve their mechanical performance for bone regeneration. Particular emphasis is placed on nanomaterial-based strategies, including the incorporation of nanoparticles and nanofibers, and their ability to enhance the physical properties of polymeric networks. Key Results: Evidence from …

Standardization Of Neuromuscular Reflex Analysis—Role Of Fine-Tuned Vision-Language Model Consortium And Openai Gpt-Oss Reasoning Llm-Enabled Decision Support System, Eranga Bandara, Ross Gore, Sachin Shetty, Ravi Mukkamala, Christopher K. Rhea, Brittany S. Samulski, Amin Hass, Atmaram Yarlagadda, Shaifali Kaushik, Malith De Silva, Andriy Maznychenko, Inna Sokolowska, Kasun De Zoysa

VMASC Publications

Background/Objectives: Accurate assessment of neuromuscular reflexes, such as the Hoffmann reflex (H-reflex), plays a critical role in sports science, rehabilitation, and clinical neurology. Conventional interpretation of H-reflex electromyography (EMG) waveforms is subject to inter-rater variability and interpretive bias, limiting reliability and standardization. This study aims to develop an automated, interpretable, and robust agentic AI–driven framework for H-reflex waveform analysis. Methods: We propose a fine-tuned Vision–Language Model (VLM) consortium combined with a reasoning Large Language Model (LLM)–enabled decision support system for automated H-reflex interpretation. Multiple VLMs were fine-tuned on curated datasets of H-reflex EMG waveform images annotated with expert clinical observations, …

Recent Advances In Bioceramics, Fundamental Properties And Future Perspective In Biomedical Applications ‒ A Review, Ayesha Younas, Muhammad Umar Aslam Khan, Mohd Faizal Binte Abdullah, Lobat Tayebi, Shuanghu Wang, Abdalla Abdal-Hay, Yichi Xu

Electrical & Computer Engineering Faculty Publications

Bioceramics are important biomaterials in biomedical engineering because of their biocompatibility, bioactivity, osteoconductivity, and structural resemblance to actual bone tissue. In recent years, materials science and nanotechnology have enabled the use of bioceramics in bone regeneration, dental restoration, tissue engineering, drug delivery systems, and implantable medical devices. This comprehensive review covers advances in bioceramics, including calcium phosphates, hydroxyapatite (HAp), tricalcium phosphate, bioactive glasses, zirconia, alumina, and multifunctional ceramic nanocomposites. Priority is given to techniques such as additive manufacturing, 3D printing, sol-gel processing, electrospinning, and nanostructuring to improve mechanical strength, porosity, bioactivity, and cell interactions. Recent advances include ion doping, surface …

C. Difficile Detection Method For First Responder Glove Application, Alli N. Senedak

Williams Honors College, Honors Research Projects

Clostridioides Difficile (C. diff) is a Anaerobic Gram-positive bacillus that is capable of spore formation, making it difficult to control its spread and duration in clinical environments. This phenomenon can provide danger to first responders, healthcare workers, and patients. The goal of this project is to create a biosensor capable of detecting C. diff in a clinical setting that can be applied to a glove apparatus. The project will involve the use of C. diff aptamers activated on the surface of an electrode. Once surface activation has been verified via surface analysis, the electrodes will be exposed to C. diff …

Wearable Biosensors For Continuous Monitoring Of Chronic Kidney Disease: Materials, Biofluids, And Digital Health Integration, Anupamaa Sivasubramanian, Shankara Narayanan, Gymama Slaughter

Center for Bioelectronics Publications

Chronic kidney disease (CKD) is a progressive and irreversible disorder affecting over 850 million individuals globally and is associated with significant morbidity, mortality, and healthcare burden. Conventional diagnostic approaches rely on intermittent laboratory measurements, including serum creatinine, estimated glomerular filtration rate (eGFR), and urinary albumin, which provide limited temporal resolution and fail to capture dynamic physiological changes. Recent advances in wearable biosensing technologies offer new opportunities for continuous, non-invasive monitoring of biochemical and physiological markers relevant to renal function. This review provides a comprehensive analysis of wearable biosensors for CKD monitoring, focusing on sensing mechanisms (electrochemical, optical, and field-effect transistor), …

Modulation Of Prussian Blue Redox Signaling By Molecular Imprinting For Reagent-Free Electrochemical Detection Of Emtricitabine, Abdellatif Ait Lahcen, Gymama Slaughter

Center for Bioelectronics Publications

Reagent-free electrochemical sensors offer significant benefits for rapid, affordable point-of-care drug testing. In this study, we introduce a novel, reagent-free electrochemical sensor based on a molecularly imprinted polymer (MIP) specifically designed for the selective detection of Emtricitabine (FTC), a common antiretroviral used in HIV therapy. The sensor uses laser-induced graphene (LIG) electrodes, renowned for their high conductivity and porosity, ideal for electrochemical sensing. To enable reagent-free operation, the LIG surface was electrochemically coated with Prussian Blue, serving as a redox-active layer. Next, an MIP-PPy film was electropolymerized onto the Prussian Blue surface in the presence of FTC as a template, …

Nanosecond Pulsed Electric Fields For Extracellular Vesicle Engineering: From Electro-Exocytosis To Cargo Modulation, Art Neal, Teresa Graham, Mohamadmahdi Samandari, Arash Ghorbannia, Anca Dobrian, Stephen J. Beebe, Ruben M. L. Colunga-Biancatelli

Bioelectrics Publications

Extracellular vesicles (EVs), including small extracellular vesicles (sEVs) and medium/large extracellular vesicles (MVs), have emerged as promising therapeutic vectors and diagnostic biomarkers across various branches of biomedicine. However, the clinical translation of EV-based technologies remains constrained by persistent challenges in manufacturing: insufficient yield from primary cell sources, limited control over cargo composition, and the absence of scalable, standardized production platforms. Nanosecond pulsed electric fields (nsPEF) represent an emerging biophysical approach that can address several of these limitations. Unlike conventional electroporation, which targets the plasma membrane using microsecond-to-millisecond pulses, nsPEF delivers ultrashort (1–300 ns), high-amplitude (10–300 kV/cm) pulses that penetrate intracellularly …

Beyond The Square Pulse: Waveform Shape, Eeg Correlates, And The Pursuit Of Natural Sensation In Tens, Jason Whitson

Honors Undergraduate Theses

This study investigates the relationship between electrical stimulus characteristics of shape and charge on evoked sensations and electroencephalogram (EEG) data during multi-waveform transcutaneous electrical nerve stimulation (TENS) of the median nerve. Neuromodulation methods have traditionally had little control over the location and quality of their associated evoked sensation (e.g., electric, vibration, touch). This experiment utilized five unique stimulus waveforms during TENS stimulation. EEG data were collected concurrently to provide an introductory objective measure of the neural responses underlying these sensory changes. Eleven participants completed three tasks (thresholding, super-threshold stimulation, two-alternative forced choice) using a two-electrode TENS approach. Stimulus waveforms were …

A Chicken Model For Non-Invasive Diagnosis Of Infant Developmental Dysplasia Of The Hip (Ddh) Using Acoustic Transmission Measurements, Aliza Shah

Honors Undergraduate Theses

Developmental dysplasia of the hip (DDH) is a condition characterized by abnormal development and stability of the hip joint. In a healthy infant, the femoral head (ball) fits securely within the acetabulum (socket of the pelvis), allowing for proper joint function. In DDH, the relationship is disrupted, ranging from mild acetabular dysplasia to partial (subluxation) or complete dislocation of the femoral head. This condition represents a range of structural abnormalities that develop during early growth. The integrity of the hip joint is essential for physical mobility, and disruption of this structure can lead to significant functional impairment. DDH may result …

A Portable Potentiostat Integrated With A Pt/Zno/Lig Electrode For Non-Enzymatic Glucose Detection, Reagan Aviha, Gymama Slaughter

Center for Bioelectronics Publications

Continuous glucose monitoring is critical for effective diabetes management; however, conventional benchtop potentiostats are bulky, costly, and unsuitable for decentralized point-of-care (PoC) applications. To address these limitations, this work presents a miniaturized, low-cost electrochemical sensing platform integrating a non-enzymatic glucose sensor with a portable potentiostat. The sensing electrode is based on laser-induced graphene modified with zinc oxide and platinum nanostructures via electrodeposition to enable sensitive glucose detection under physiological conditions. A custom-designed portable potentiostat was developed to control electrode potentials and perform electrochemical measurements, and its performance was experimentally validated against a commercial Metrohm system. Glucose detection was evaluated using …

Recent Insight And Perspective Of Marine-Inspired Biopolymers For Wound Healing Applications - A Review, Muhammad Umar Aslam Khan, Saiqa Yousaf, Abdalla Abdal-Hay, Mohd Faizal Bin Abdullah, Sahar Madani, Muhammad Shahzad Zafar, Goran M. Stojanović, Lobat Tayebi

Electrical & Computer Engineering Faculty Publications

There is an increasing necessity for advanced, sustainable, and biocompatible materials for wound healing as therapeutic and diagnostic products. Marine environments, characterized by high biodiversity, offer an underutilized source of natural resources with enormous potential for creating novel materials for dressings. This review highlights the revolutionary nature of polymeric biomaterials of marine origin, with a focus on polysaccharides, like alginate, chitosan, and carrageenan; proteins, such as collagen and gelatin. These biopolymers are outstanding in their physicochemical properties, such as biodegradability, bioactivity, and modifiable mechanical strength, which enable their use in wound-healing systems. Besides, these biomaterials may be easily chemically and …

Pin-Plane Electrical Discharge Driven By A Mosfet Dc Current Source, Myles Perry, Sidmar Holoman, Daniel Wozniak, Shirshak Kumar Dhali

Electrical & Computer Engineering Faculty Publications

The generation of atmospheric pressure nonequilibrium plasma using electrical discharges is an active area of research due to its significance in a wide spectrum of applications including medicine, combustion, and manufacturing. In our attempt to create a helium plasma jet in a pin-plane discharge with a constant current source, we observed self-pulsating behavior. We present the results of the electrical, optical, and spectroscopic measurements carried out to characterize the discharge. The duration of the discharge is a few tens of nanoseconds, and the repetition rate is in the few tens of kHz. The effect of the gap distance and gas …

Privacy-Preserving Federated Learning With Optimized Ensemble Weighting And Knowledge Distillation For Covid-19 Detection From Non-Iid Medical Imaging Data, Richard Annan, Hong Qin, Robert Newman, Madhuri Siddula, Letu Qingge

Computer Science Faculty Publications

Medical imaging enables rapid and accurate diagnosis of COVID-19, with CT scans proving especially effective. However, data privacy concerns limit collaborative model development across hospitals. To address this issue, we introduce a novel federated learning framework. It is referred to as Independent Knowledge Distillation with post-Ensemble Federated Learning (IKDEFL). Differential Privacy (DP) is integrated into the framework to improve privacy guarantees. Three DP mechanisms are evaluated. These include Fixed Gaussian, Gaussian Adaptive, and Tree Adaptive. The evaluation has been conducted on heterogeneous and Non-Independent and Identically Distributed (Non-IID) datasets. These datasets reflect real-world hospital scenarios. Results show that IKDEFL significantly …

A Comparative Analysis Of Explainable Ai (Xai) Techniques For Transparent And Reliable Image Classification, Sovon Chakraborty, Shakib Mahmud Dipto, Kevin R. Pilkiewicz, Michael L. Mayo, Pratip Rana

Computer Science Faculty Publications

Evaluating the trustworthiness of black-box machine learning models remains a significant methodological challenge. Their lack of transparency and interpretability limits applicability, because stakeholders often seek transparency before trusting the results of black-box machine learning models. Explainable AI (XAI) methods provide for human-understandable justifications and informed decision-making of these black-box architectures. Therefore, it is imperative to select the proper XAI model tailored to specific tasks. In this research, we focus on examining four XAI techniques: PEEK, LRP, GRAD-CAM, and LIME to understand how they perform against each other for image classification tasks. We evaluate the performance, robustness, generalizability, noise stability, and …

Adaptive Self-Attention For Enhanced Segmentation Of Adult Gliomas In Multi-Modal Mri, Evan P. Savaria, Jiangwen Sun

Computer Science Faculty Publications

Every year there are an estimated 80,000–90,000 new glioma cases, highlighting the need for reliable imaging-based decision support. Although deep learning has improved tumor sub-region segmentation, many state-of-the-art models fail to fully capture complementary information across T1, T1Gd, T2, and FLAIR MRI modalities and often operate as “black boxes,” limiting physician trust when precise delineation is critical for surgical planning, radiation targeting, and treatment monitoring. To address these limitations, we propose AIMS, an Adaptive Integrated Multi-Modal Segmentation framework that maintains modality-specific feature streams and employs adaptive self-attention within a hierarchical CNN-Transformer architecture to prioritize and fuse multi-modal MRI features. We …