Electrical and Computer Engineering Commons^™

Evolving Ai Integration In Complex Medical Decision-Making And Multidisciplinary Transplant Care: A Systematic Review Of Human-Ai Collaboration, Rachel L. Dzieran, Cihan H. Dagli, Robert J. Marley

Engineering Management and Systems Engineering Faculty Research & Creative Works

Purpose of Review: Artificial intelligence (AI) in healthcare has evolved dramatically from early expert systems, which were initially considered replacements for clinical judgment, to today's collaborative frameworks that aim to augment physician decision-making. This evolution is particularly crucial in domains such as transplant surgery, where decisions carry irreversible consequences and require the integration of complex, often ambiguous data. Drawing on peer-reviewed literature from 2019 to 2025, we conducted a systematic review that analyzed key elements distinguishing successful human-AI partnerships from those that fail. Recent Findings: The ideal balance incorporates human expertise into AI systems through weighted integration approaches, rather than …

Ncf Sensor Coated With Cofe1.96la0.04o4/Lafeo3 Heterostructures For Room-Temperature Detection Of Liquefied Petroleum Gas Concentration, Ziqiang Liu, Fujian Tang, Yufang He, Jie Huang

Electrical and Computer Engineering Faculty Research & Creative Works

Liquefied petroleum gas (LPG) is a highly flammable fuel widely used for cooking, heating and transportation, making real-time leak detection essential for preventing fire, explosion, and suffocation hazards. In this study, a room-temperature no-core fiber (NCF) sensor coated with CoFe_1.96La_0.04O₄/LaFeO₃ heterostructures was developed for LPG detection. The heterostructures were characterized using XRD, Raman spectroscopy, SEM, TEM, XPS, PL spectroscopy, and UV–vis spectroscopy. Experimental results showed that La incorporation and heterostructure formation refined the particles size, improved surface accessibility, and modulated the electronic structure and band gap, thereby enhancing LPG adsorption, carrier redistribution, and …

Ph-Compensated Hydrogen Peroxide Quantification Using A Dual-Modal Fiber-Optic Probe, Homayoon Soleimani Dinani, Bohong Zhang, Maryam Karimi, Rex E. Gerald, Shelley D. Minteer, Jie Huang

Electrical and Computer Engineering Faculty Research & Creative Works

We report a dual-modal fiber-optic probe that integrates electrochemical quantification of hydrogen peroxide (H₂O₂) with co-localized fluorescent pH sensing for pH-indexed interpretation of the H₂O₂ response. H₂O₂ is a reactive oxygen species involved in oxidative stress, inflammation, and cellular signaling, and local pH modulates both its production and electrochemical response. Many electrochemical H₂O₂ sensors exhibit pH-dependent sensitivity, creating ambiguity unless pH is measured and used for compensation, which is difficult in small, heterogeneous, or rapidly changing microenvironments. A three-electrode configuration—working (WE), counter (CE), and Ag/AgCl pseudo-reference (pRE) electrodes—is fabricated directly on the cylindrical surface of a 710-µm-diameter optical fiber using …

Graphene Oxide–Driven In Situ Bismuth Reduction Enables Highly Efficient Electroreduction Of Co2 To Formic Acid, Hao Feng, Bohong Zhang, Jie Huang, Xinhua Liang

Electrical and Computer Engineering Faculty Research & Creative Works

To enable large scale efficient electrochemical CO₂ reduction reaction (CO₂RR) to formic acid (HCOOH), it is important to develop catalysts that can be operated in a wide potential window with good stability. Herein, we successfully synthesized Bi₂O₃ catalyst supported on graphene oxide (GO) and graphene (G) and found that Bi₂O₃/GO catalyst had a better overall performance than Bi₂O₃/G. The Bi₂O₃/GO catalyst demonstrated an outstanding CO₂RR performance with a greater than 90% faradaic efficiency (FE) across a wide applied potential window …

Thermo-Mechanical Behaviour Of Metal-Coated Optical Fibers For Distributed High-Temperature Sensing: From Laboratory Characterization To Industrial Case Validation, Koustav Dey, Rony Kumer Saha, Bohong Zhang, Laura Bartlett, Jeffrey D. Smith, Ronald J. O'Malley, Jie Huang

Electrical and Computer Engineering Faculty Research & Creative Works

Reliable distributed temperature sensing in high-temperature environments remains a significant challenge due to the thermal and mechanical limitations of conventional optical fibers. In particular, polymer-coated fibers degrade above ∼300 °C due to coating failure, mechanical fragility and hydrogen ingress. Metal-coated optical fibers offer a robust alternative for harsh environments such as Electric Arc Furnaces (EAFs), aerospace engines, nuclear systems, and oil and gas wells, owing to their superior mechanical strength and hermetic sealing. In this work, a first comprehensive experimental investigation of the thermo-mechanical behavior of metal-coated optical fibers for distributed high temperature sensing is presented over a wide temperature …

Linear Relationships Of Gibbs Free Energy For Rare Earth Element Oxide, Hydroxide, Chloride, Fluoride, Carbonate, Ferrite, And Zirconate Minerals And Crystalline Solids, Ruiguang Pan, Alexander P. Gysi, Artaches A. Migdisov, Nicole C. Hurtig, Chen Zhu

Electrical and Computer Engineering Faculty Research & Creative Works

Rare Earth Elements (REE) are classified as critical minerals and materials essential for the transition from fossil fuels to renewable and clean energy. Accurate thermodynamic properties of REE minerals and other crystalline solids are crucial for geochemical modeling of REE speciation, solubility, and transport in ore deposits and for extraction, chemical processing, and recycling. However, the standard Gibbs free energy of formation (∆Gof,REEX) for these solids vary by 10s kJ mol−1 across different sources from literature. We applied the Sverjensky linear free energy relationship (LFER) to evaluate internal consistency for isostructural REE solid groups and predicted ∆Gof values for solid …

Multi-Parametric Nonlinear Programming For Lossy Lmp Sensitivity Analysis Using Outer Progressive Cuts, Yuhan Huang, Tao Ding, Chenggang Mu, Rui Bo, Pengwei Du

Electrical and Computer Engineering Faculty Research & Creative Works

Accurate sensitivity analysis of locational marginal price (LMP) is crucial for risk hedging and market management. This paper proposes multiparametric nonlinear programming for sensitivity analysis of LMP with nonlinear network loss, also known as lossy LMP. Global analytical solutions for lossy LMP and the corresponding critical regions are derived. An outer progressive cut algorithm is developed to constrain the network loss error within a specific range for the whole parametric domain. Case studies show the improved accuracy over methods with lossless or fixed loss factors, and a two-order-of-magnitude online speedup over Monte Carlo methods while maintaining lower total computational time.

Distinct Regulatory Dna Methylation Signatures Across Multiple Sclerosis, Neuromyelitis Optica, And Neurological Post-Acute Sequelae Of Covid-19, Syed Ilyas Munzir, Daniel B. Hier, Michael D. Carrithers

Electrical and Computer Engineering Faculty Research & Creative Works

Background/Objectives: Our prior epigenome-wide association study (EWAS) on multiple sclerosis (MS) identified myeloid-associated methylation signatures and an association with enhancer regions. Here we compared differential DNA methylation across three central nervous system inflammatory disorders: MS, neuromyelitis optica (NMO), and neurologic post-acute sequelae of COVID-19 (neuro-PASC). Methods: Whole-blood DNA was profiled on Infinium MethylationEPIC arrays. Analyses included EWAS at the CpG level, differentially methylation region (DMR) analysis, and gene regulatory-element enrichment using Locus Overlap Analysis (LOLA). Limma linear models were adjusted for race, EPIC array version, age, sex, disease-modifying treatment class, and blood cell composition. Results: All three diseases were associated …

An Instance Segmentation Based Global Mppt Method For Pv Systems Under Partial Shading Condition, Ehab Ur Rahman, Yousef Mahmoud, Rui Bo

Electrical and Computer Engineering Faculty Research & Creative Works

Partial shading in photovoltaic (PV) systems creates multiple Local Maximum Power Points (LMPPs) on the power-voltage (P-V) curve, causing conventional Maximum Power Point Tracking (MPPT) algorithms to converge prematurely to suboptimal operating points and resulting in significant energy losses. This work proposes a novel vision-based framework that integrates computer vision with Global Maximum Power Point Tracking (GMPPT) to maximize energy harvesting in partially shaded PV systems. The proposed method captures real-time imagery of PV panels to identify cell-level shading patterns, which are subsequently mapped to module electrical characteristics to enable direct analytical estimation of the GMPP voltage. An instance segmentation …

Converting Co2 Into Carbon Nanotubes Via Sequential Co2 Methanation And Methane Pyrolysis, Kaiying Wang, Hao Feng, Ahmad Helaley, Xiaoqing He, Bohong Zhang, Jie Huang, Xinhua Liang

Electrical and Computer Engineering Faculty Research & Creative Works

This study presents a sequential thermochemical catalytic process for converting CO₂ into carbon nanotubes (CNTs), achieving nearly 99% overall CO₂ conversion and solid carbon yields of up to 49% in a single pass. The overall transformation is divided into two thermodynamically favorable steps: (1) low-temperature CO₂ methanation to CH₄ over a sol-gel-prepared Ni87Ce_6.5Zr_6.5-SG catalyst, which achieved >95% CO₂ conversion at 250 °C in a single pass, and (2) subsequent methane pyrolysis to CNTs and H₂ over a Ni@Al₂O₃-IE core-shell catalyst, which showed improved stability and carbon …

Enhanced Uav Surveillance With Rf-Based Drone Identification Using Transfer Learning, Prajoy Podder, Maciej Jan Zawodniok, Sanjay Madria

Electrical and Computer Engineering Faculty Research & Creative Works

With the development of technology and the decrease in costs, drones are now becoming easily accessible to the public. As the accessibility of this technology continues to grow, the concerns of security and surveillance increase, and to ensure a sense of security, the need to have reliable drone detection and identification systems is more urgent than ever. Besides, many civilian applications have been found for drones, which play a huge role in modern security and warfare. Unauthorized drones can be very dangerous regarding security issues, as they can be used for spying, smuggling, or even attacks against critical infrastructure. We …

Evaluation Of The Effect Of Vibration On Signal Reflection In Coaxial Cable Connectors For Vibration Sensing In Aircraft Structures And Systems, Saidanvar Esanjonovich Valiev, Anthony C. Okafor, Jeremiah J. Rittenhouse, Daniel S. Stutts, Jie Huang

Mechanical and Aerospace Engineering Faculty Research & Creative Works

This paper investigates the effects of vibration on signal reflection (S11) in aerospace data transmission line (ADTL) and commercial data transmission line (DTL) connectors for their alternative use as vibration sensors. The impact of vibration on the S11 signal was investigated on five ADTL and four DTL connectors at six vibration frequencies (20 Hz, 40 Hz, 80 Hz, 160 Hz, 320 Hz, and 640 Hz) and four vibration accelerations (0.5G, 1 G, 2 G, and 4G). The experiment was conducted using a split-plot design with a cable type assigned as the main-plot factor, with vibration frequency and acceleration as subplot …

Pso-Style Social Influence In An Ant Colony Algorithm For Continuous-Domain Optimization, Ashraf M. Abdelbar, Donald C. Wunsch

Electrical and Computer Engineering Faculty Research & Creative Works

ACOR is a well-established Ant Colony Optimization (ACO) algorithm for continuous-domain optimization. In this paper, we propose an extension (which we call ACOR∗) in which several fundamental modifications are made to ACOR's solution construction process, including the incorporation of a social influence mechanism borrowed from Particle Swarm Optimization (PSO). Our modifications to the ACOR algorithm are intended to promote search diversity and combat premature convergence. We experimentally evaluate our proposal in the context of training feedforward neural networks for classification using 65 widely used datasets from the University of California Irvine (UCI) repository, as well as the optimization of several …

Multiplexed Fabry-Pérot High-Temperature Sensing Based On Dispersive Microwave-Photonic Frequency-Time Domain Analysis, Ruimin Jie, Chen Zhu, Bohong Zhang, Koustav Dey, Jie Huang

Electrical and Computer Engineering Faculty Research & Creative Works

We propose and experimentally demonstrate a multiplexed high-temperature Fabry-Pérot (FP) fiber sensing system interrogated by dispersive microwave-photonic frequency-time domain analysis (DM-FTDA). In the proposed architecture, incoherent broadband probing light is modulated by radio-frequency (RF) signals and then reflected by a parallel network of hollow-core photonic crystal fiber FP (HCPCF-FP) sensors. A chirped fiber Bragg grating provides strong dispersion to map the composite FP spectral response into a well-defined microwave transfer function. Unlike conventional optical Fourier-domain multiplexing that requires deliberate cavity-length allocation, the proposed approach achieves multiplexing via delay-dominated discrimination. Distinct delay fibers are assigned to each sensor branch, and an …

Advances In The Design Of Bio-Organic Resistive Switching Memory, Muhammad Awais, Yi Sheng Wong, Feng Zhao, Kuan Yew Cheong

Electrical and Computer Engineering Faculty Research & Creative Works

Bio-organic materials have garnered significant attention as sustainable candidates for non-volatile resistive switching memory (RSM) because of their specialized chemical, structural, and environmental advantages. This review presents a design-centered perspective on bio-organic RSM by outlining the key device components required for effective device engineering, including electrode materials, memristive thin films, intermediate layers, substrates, and electrical measurement strategies. Each component is discussed in detail with respect to the material properties and operational parameters that influence overall device performance, such as functional groups, interfacial interactions, and processing conditions. The review further analyses the critical roles of electrode pairing, interfacial chemistry, additive incorporation, …

Rioi: A Microwave-Photonic Rf-Interferometric Interrogation Technique For Enhanced Fiber Optic Sensing, Chen Zhu, Ruimin Jie, Chenxi Huang, Jie Huang

Electrical and Computer Engineering Faculty Research & Creative Works

Fiber optic interferometric (FOI) sensors are widely recognized for their high sensitivity, design flexibility, and multiplexing capabilities, making them ideal for applications ranging from structural health monitoring to biomedical diagnostics. However, conventional optical-domain interrogation techniques are often limited by the performance constraints of spectrometers. In this work, we present a radiofrequency (RF)-interferometric optical interrogation (RIOI) method for FOI sensors. This approach leverages microwave photonic (MWP) processing to encode the optical interference phase into an RF signal, which is then combined with a reference RF signal to produce a microwave-domain interferogram. By tracking spectral shifts in the RF domain, RIOI achieves …

Robustness Of Fuzzy Artmap To Adversarial Attacks And Progressive Adversarial Training For Streaming Learning, Shane Cairns

Miners Solving for Tomorrow Research Conference

Incremental learners deployed on streaming data must remain robust to evolving adversarial perturbations, yet most adversarial-robustness studies assume offline multi-epoch training with repeated access to historical data. We investigate adversarial robustness in Fuzzy ARTMAP, a prototype-based Adaptive Resonance Theory model that supports single-pass learning without replay. We propose WB-Softmax, a differentiable relaxation that aggregates category-level activations into class-level scores for gradient-based attacks. WB-Softmax PGD achieves 89–100% attack success on vanilla models, exceeding transfer and query-based baselines. We then study adversarial training under true streaming constraints by comparing offline versus online adversarial example generation and standard versus selective updates. Offline adversarial …

Triple Active Bridge Implementation And Control, Nehemiah Milton

Miners Solving for Tomorrow Research Conference

The Triple Active Bridge (TAB) is a three-port power converter that allows power flow in both directions with galvanic isolation. This has a wide range of applications, like high-frequency DC-DC conversion, electric vehicles, microgrids and renewable energy systems. To control the TAB during operation, two phase shift parameters between the bridges are adjusted. In this presentation, I will showcase the software and simulation optimizations which extend previous work by reproducing hardware results which align with simulations. I will also go over the knowledge I’ve gained and the skills acquired through this project.

Analysis Of Autonomous Vehicle Survivability To 5g Communication Errors, Sydney Clark

Miners Solving for Tomorrow Research Conference

Autonomous vehicles rely on low-latency, high-reliability data exchange for real-time perception and control. Disruptions such as packet loss, latency variation, protocol-level errors, and malicious interference can pose significant safety risks to both passengers and surrounding environments. This project aims to evaluate, quantify, and predict the survivability of autonomous vehicle systems to communication errors, with focus on 5G network environments. The impact of these communication impairments on vehicle stability and control will be investigated through high-fidelity cyber-physical simulation of the vehicle and its surrounding environment. Experiments designed to capture varying network conditions will be used to assess a broad range of …

A Wearable Mxene-Based Sweat Sensor For Real-Time Monitoring Of Inflammatory Biomarkers, Ariel Pilger

Miners Solving for Tomorrow Research Conference

Many conventional biosensing approaches rely on invasive sampling or bulky benchtop instrumentation, limiting their use in continuous and portable applications. This project focuses on the development of wearable sweat-based biosensors that enable non-invasive, continuous, and portable monitoring of physical, chemical, and biological markers. The system will be designed to target markers present in sweat and transduce the biochemical interactions into measurable electrical signals. These signals will be processed through integrated electronics to produce clear, interpretable outputs for users and medical professionals. Supporting circuitry including filters, amplifiers, and an independent power supply will be implemented as necessary to ensure signal accuracy, …

Corrigendum To “Advanced Techniques In Quartz Wafer Precision Processing: Stealth Dicing Based On Filament-Induced Laser Machining” [Opt. Laser Technol. 171 (2024) 110474] (Optics And Laser Technology (2024) 171, (S0030399223013671), (10.1016/J.Optlastec.2023.110474)), Yun Wang, Yutang Dai, Farhan Mumtaz, Kaiyan Luo

Electrical and Computer Engineering Faculty Research & Creative Works

The authors regret, that the affiliation for author Yun Wang was incomplete. To accurately reflect both the author's academic affiliation and the research platform where the work was conducted. The correct affiliation for Yun Wang is updated as above. The authors would like to apologize for any inconvenience caused.

Microwave Dielectric Properties Of Geopolymer Precursor Powders, Linh T. Duong, Abu Naser Rashid Reza, Kristen M. Donnell, Christopher R. Shearer

Electrical and Computer Engineering Faculty Research & Creative Works

Geopolymers are sustainable structural materials with properties similar to ordinary Portland cement concrete. To better understand the fundamental reaction mechanisms of geopolymers, a microwave materials characterization approach is used. In this research work, the dielectric properties (permittivity and loss factor) of common precursor powders used to make geopolymers (GPPs)—fly ash, lime, metakaolin, silica fume, slag, and zeolite—are measured over the S- and X-band frequency ranges (i.e., 2.0–4.0 GHz and 8.2–12.4 GHz, respectively). The physical characteristics, elemental composition, mineralogical properties, and phase characterization of the GPPs are then correlated to dielectric properties. Permittivity of GPP, classified as pozzolanic or latent hydraulic, …

Applying Two-Stage Risk-Based Market Structures For Energy Hub-Based Plug-In Electric Vehicles Using Information Decision Gap Theory And A Hybrid Recurrent Convolutional Network, A. Heidari, R. C. Bansal, R. Bo

Electrical and Computer Engineering Faculty Research & Creative Works

This paper investigates the optimal operation of an energy hub engaged in both day-ahead and real-time trading. A two-stage optimization framework Information Gap Decision Theory (IGDT) for day-ahead bidding and stochastic programming with Monte Carlo scenarios for real-time recourse is applied. Risk-neutral, risk-averse, and risk-taking strategies are considered to capture different risk preferences. The hub integrates combined heat and power, renewable energy, plug-in electric vehicles, and vehicle-to-grid and grid-to-vehicle technologies. Price and load forecasts are generated using a hybrid recurrent convolutional network (HRCN). Results highlight the trade-off between risk management and economic performance: costs are 16.5 % higher in the …

Machine Learning In Peak Demand Forecasting: Foundations, Trends, And Insights, Shuang Dai, Fanlin Meng, Hongsheng Dai, Qian Wang, Xizhong Chen, Wenlei Bai, Peizhi Shi, Richard Allmendinger, Yuchen Zhang, Jian Liu

Electrical and Computer Engineering Faculty Research & Creative Works

Peak demand forecasting involves predicting the maximum electricity demand within a specific period, which plays a key role in maintaining the efficiency and stability of power systems. The rapid evolution of power systems, driven by advanced metering infrastructure, local energy applications such as electric vehicles, and the increasing adoption of intermittent renewable energy, has introduced greater randomness and reduced predictability in peak demand. Given the pressing need to address more diverse implementation requirements across different contexts, accurate and reliable peak demand forecasting has become increasingly important. To the best of our knowledge, this study is the first to provide a …

Robust And High-Efficiency Demodulation Of Ultra-Weak Fbg Arrays In Ofdr-Based Distributed Sensing, Zhaopeng Zhang, Yuxuan Cao, Xu Liu, Dingcheng Wang, Bo Liu, Chen Zhu

Electrical and Computer Engineering Faculty Research & Creative Works

A robust and high-efficiency demodulation scheme for optical frequency domain reflectometry (OFDR) based ultra-weak fiber Bragg grating (UWFBG) array detection system, originating from the Buneman frequency estimation (BFE) algorithm, is proposed and experimentally demonstrated. Due to the current limitations and imperfections of FBG inscription technology, the quasi-continuous inscription approach, along with its less-than-ideal outcomes, gives rise to problems of grating spectrum splitting and spectral distortion during the grating demodulation process. This renders the traditional approach of directly applying the BFE algorithm for grating demodulation ineffective, despite its significant enhancement of demodulation efficiency. To address this issue, we propose utilizing the …

Safety Aware Continual Reinforcement Learning-Based Output Tracking Control Of Nonlinear Continuous-Time Systems, Irfan Ganie, Sarangapani Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

An output feedback (OF)-based control scheme utilizing both a scalable multilayer neural network (MNN) observer and actor–critic MNN via integral reinforcement learning (IRL)/adaptive dynamics programming (ADP) approach for a class of nonlinear systems with output constraints is introduced. The proposed observer, critic, and actor MNN weight updates are derived using a singular value decomposition (SVD) of MNN activation function gradient along with output error, Bellman and control input errors, respectively. Next, the approach incorporates continual learning (CL), utilizing a penalty function in the weight update laws for both actor–critic MNNs to consolidate knowledge from previous tasks and enhance learning in …

Robust Broadband Characterization Of Flexible Absorbers, Sheet Material, And Liquids Using A Coaxial Structure, Joseph Christopher Stecher

Masters Theses

Modern high-frequency measurement systems require reliable calibration and sample positioning to ensure measurement fidelity. This thesis presents three studies addressing practical limitations in broadband material parameter extraction and instrumentation.

The first study introduces a modified Nicolson–Ross–Weir (NRW) technique for flexible, compression-sensitive materials from 100 MHz to 18 GHz. Rigid 3D-printed spacers ensure precise sample positioning, and a T-matrix–based de-embedding procedure removes spacer effects. Validation using microstrip measurements and full-wave simulation confirms accurate permittivity extraction across compression levels.

The second study extends NRW to sheet materials enabling accurate material characterization. Independent validation using toroidal inductors with leakage correction and parallel-plate capacitors …

Validating A Low-Cost Radio Frequency Characterization Framework For Development-Grade Software-Defined Radios In Short-Duration Cubesat Missions, Thomas Wayne Francois

Masters Theses

Software-defined radios (SDRs) and CubeSat platforms have reduced the cost and complexity of space-based communication systems, enabling broader participation in satellite missions. While low-cost radio hardware is increasingly accessible, the ability to characterize and validate its performance remains constrained by the high cost and limited access to traditional RF test equipment. This disparity creates a challenge for small satellite development teams, which must characterize communication-system technical performance with limited access to laboratory-grade instrumentation.

This thesis presents a low-cost RF characterization framework for assessing key radio-frequency performance metrics using readily available hardware and measurement techniques. The approach integrates frequency translation, SDR-based …

Distributed Vibration Sensing For Identification Of Loose Connectors In Coaxial Data Transmission Lines, Saidanvar Esanjonovich Valiev, Anthony C. Okafor, Jeremiah J. Rittenhouse, Jie Huang, Daniel S. Stutts

Mechanical and Aerospace Engineering Faculty Research & Creative Works

This study investigates the effect of vibration-induced loose connections on signal reflection (S11) for loose connection identification in aerospace coaxial cables using distributed sensing approach, which is effective in filtering the noise and identifying minor discontinuities. In this approach, a sliding gated window is applied to S11 signal, a fast Fourier transform is performed over the gated windows, cross-correlation is computed between the baseline and vibration-affected signals, and the standard deviation is mapped along the cable length. Sinewave signals from 9 kHz to 5 GHz were swept through cables with vibrating connectors under three conditions: fully tightened, loosened by 180°, …

Evaluation Of Multiple Generative Large Language Models On Neurology Board-Style Questions, Mohammad Almomani, Vijaya Valaparla, James Weatherhead, Xiang Fang, Alok Dabi, Chih Ying Li, Peter Mccaffrey, Dan Hier, Jorge Mario Rodríguez-Fernández

Electrical and Computer Engineering Faculty Research & Creative Works

Objective: To compare the performance of eight large language models (LLMs) with neurology residents on board-style multiple-choice questions across seven subspecialties and two cognitive levels. Methods: In a cross-sectional benchmarking study, we evaluated Bard, Claude, Gemini v1, Gemini 2.5, ChatGPT-3.5, ChatGPT-4, ChatGPT-4o, and ChatGPT-5 using 107 text-only items spanning movement disorders, vascular neurology, neuroanatomy, neuroimmunology, epilepsy, neuromuscular disease, and neuro-infectious disease. Items were labeled as lower- or higher-order per Bloom's taxonomy by two neurologists. Models answered each item in a fresh session and reported confidence and Bloom classification. Residents completed the same set under exam-like conditions. Outcomes included overall and …