About

Amanda Watson is an Assistant Professor in Electrical Engineering and Computer Science at the University of Virginia. Her research focuses on wearable technology for healthcare and athletic performance. She is affiliated with the UVA Link Lab, a multi-disciplinary center dedicated to research and education in Cyber-Physical Systems (CPS). Prior to joining UVA, she was a Postdoctoral Fellow at the PRECISE Center at the University of Pennsylvania, where she was mentored by James Weimer and Insup Lee. She also researched wearable technology as a member of the LENS lab with Gang Zhou. Watson received her Ph.D. in Computer Science from the College of William and Mary in 2020, after earning an M.S. in Computer Science from the same institution in 2016. She completed her B.A. in Computer Science and Mathematics at Drury University in 2014, where she was a four-year varsity volleyball letterman.

Research topics

• Computer Science
• Artificial Intelligence
• Medicine
• Orthodontics
• Physical medicine and rehabilitation
• Engineering
• Embedded system
• Computer vision
• Dentistry
• Acoustics

Selected publications

• HypoxSpike: Ternary Spiking Neural Network for Opioid Overdose Detection

  Proceedings of the AAAI Conference on Artificial Intelligence · 2026-03-14

  articleOpen accessSenior author

  Opioid overdose is a growing global health crisis that claims more than 120,000 lives annually, of which more than half use opioids alone, without access to bystander intervention. Fatal overdose events are marked by motionlessness, respiratory depression, and hypoxemia, yet current wearable systems often rely on a single biomarker, limiting detection speed and accuracy. We present HypoxSpike, a novel ternary spiking neural network designed for real-time, multi-biomarker overdose detection for low-power neuromorphic hardware, optimized for integration into shoulder-based wearables. HypoxSpike combines motion, respiration, and oxygen saturation signals, while accounting for skin tone and body physiology, thus addressing known racial bias in pulse oximetry. Our research leverages an open-source shoulder-worn dataset from 19 patients experiencing sleep apnea, exploiting the shared physiological mechanisms underlying apnea and opioid overdose. This allows a direct comparison of our model with existing overdose detection approaches. HypoxSpike classifies three stages of hypoxemia with an average accuracy of 94%, outperforming state-of-the-art shoulder-based hypoxemia estimation while reducing false positive alert rates by 23.5%. By minimizing false positives, HypoxSpike supports accurate and power-efficient overdose detection, improving trust and usability for high-risk populations often overlooked by conventional systems.

  PublisherOA PDFDOI

• VIBRANT: Early Prediction of Life-Threatening Uterine Atony Using Maternal Heart Rate

  2025-06-24 · 1 citations

  articleOpen access

  Uterine atony accounts for a vast majority of all postpartum hemorrhages (PPH), the leading cause of maternal mortality worldwide. Uterine atony occurs when the uterine muscle (called the myometrium) does not sufficiently contract to arrest parturient bleeding after delivery. While there exist treatments for uterine atony, delays in intervention reduce their effectiveness. To improve time-to-intervention, postpartum hemorrhage risk prediction tools have been integrated into the standard-of-care for obstetrics. Unfortunately, these tools miss almost half of all PPHs. This paper presents VIBRANT as a clinical decision support tool providing early prediction of potentially life-threatening uterine atony. VIBRANT is physiologically inspired and designed to identify signals of myometrial fatigue argued to be present in streaming maternal heart rate data. Evaluations of the system indicate that VIBRANT, at a clinically actionable specificity, identifies over 80% of potentially life-threatening uterine atony missed by current gold-standard risk prediction tools. Moreover, the system provides between 2 and 8 hours advance warning to care teams prior to delivery for parturients at high-risk, providing ample time for preparation and timely intervention. VIBRANT has been licensed to a commercial partner and is in preparation for a clinical trial to support a regulatory submission and pre-market approval.

  PublisherOA PDFDOI

• Precision Rehabilitation After Youth Anterior Cruciate Ligament Reconstruction: Individualized Reinjury Risk Stratification and Modifiable Risk Factor Identification to Guide Late-Phase Rehabilitation

  Orthopaedic Journal of Sports Medicine · 2025-04-01 · 1 citations

  articleOpen access

  Background: After anterior cruciate ligament (ACL) reconstruction, adolescent athletes have a high risk of second ACL injuries, and revision ACL reconstruction is associated with increased medical costs, reduced activity levels, chronic knee pain, and higher rates of knee osteoarthritis, making the prevention of a reinjury a priority. While athlete clearance protocols and algorithms exist, the current methods of identifying the reinjury risk have limited predictive accuracy and are largely based on nonmodifiable risk factors, which limit their clinical application. Purpose: The goal of this study was to develop an ACL reinjury risk prediction (ACL-RRP) model capable of accurately classifying an individual patient's risk, identifying modifiable risk factors, and ranking these factors in the order of importance and ability to be modified. Study Design: Cohort study (Diagnosis); Level of evidence, 2. Methods: A clinician-informed approach was utilized to develop the prediction model and an interpretable output system. The primary outcome variable was the likelihood of sustaining a repeat ACL injury. The data were split into training (80% [n = 628]) and holdout (20% [n = 158]) datasets to train and subsequently validate the model. The accuracy of classification was identified by the sensitivity, specificity, positive/negative predictive values, and odds ratio. Results: The final model included 33 predictor variables, 23 of which are modifiable. The model adjusted the weight of the risk classification and risk factors (predictor variables) on a case-by-case basis. The model demonstrated a sensitivity of 94% and a specificity of 76%. Patients classified as being high risk had 4.5 times the risk of repeat ACL injuries compared with those classified as being low risk. Conclusion: This clinician-informed ACL-RRP model demonstrated a high degree of accuracy when classifying patients as having a high or low risk of repeat ACL injuries and generated patient-specific modifiable risk factors to guide ongoing rehabilitation or patient education to achieve the goals of reducing the ACL reinjury risk.

  PublisherDOI

• OpenSpectro: An Open-Source Spectroscopic Profiling Platform

  2025-07-14

  articleSenior author

  Spectroscopic analysis is essential for identifying optical molecular signatures, which are distinct patterns observed across different wavelengths. Understanding these signatures provides critical insights for designing wearable health-monitoring devices. In particular, constructing three-dimensional (3D) spectroscopic graphs of molecular spectra enables the optimization of multi-wavelength photoplethysmography (PPG) sensors, improving their accuracy and performance. However, no prior work has systematically mapped spectroscopic signatures to optimize wavelength combinations, slowing advancements in multi-wavelength PPG sensor deployment. To address this gap, we introduce OpenSpectro, an open-source spectroscopic profiling platform for visualizing and sharing molecular spectral data, particularly human physiological biomarkers. OpenSpectro features a preliminary spectroscopic database containing 17 biomarkers and a spectral attention optimization model that identifies customized wavelength attention weights for each biomarker.

  PublisherDOI

• A Multi-Wavelength Optical Sensing Framework for Calibration-Free Wearable Blood Pressure Monitoring

  2025-03-12

  articleSenior author

  Blood pressure (BP) is a key indicator of cardiovascular health, with hypertension leading to significant morbidity and mortality worldwide. Continuous monitoring of BP is essential for early detection of cardiovascular disease, however current tools are either cumbersome, unreliable, or not suited for long-term use. Traditional cuff-based BP measurement, while reliable, is impractical for continuous monitoring. Recent advances using photoplethysmography (PPG) waveforms offer an alternative, but they face challenges such as limited interpretability, high computational complexity, and susceptibility to motion artifacts. In this paper, we introduce a novel multi-wavelength optical sensing framework designed for calibration-free wearable blood pressure monitoring. Our system utilizes a broad spectrum of wavelengths and interpretable features, combined with machine learning, to estimate systolic (SBP), diastolic (DBP), and mean arterial pressure (MAP). The framework was tested in a proof-of-concept study involving 9 subjects across varied postures and BP levels, demonstrating accuracy comparable to standard cuff-based techniques. This approach eliminates the need for continuous calibration and provides a scalable, interpretable solution for real-time, wearable BP monitoring.

  PublisherDOI

• DermaGlow: Objective Quantification of Melanin, Erythema and Skin-tone Using Wearable Optical Spectroscopy

  Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · 2025-06-09 · 2 citations

  articleOpen accessSenior author

  Accurate characterization of the skin is essential for optimizing diagnostic and therapeutic dermatological tools, as well as technologies like pulse oximetry that rely on skin perfusion. Traditionally, optical spectroscopy has been used for skin assessments through devices like commercial colorimeters, which are high-cost instruments that, while precise, only provide single measurements rather than continuous data. Additionally, medical wearable devices that use this technology often show variable accuracy based on skin tone. The limitations of existing devices demonstrate the need for a solution that can provide low-cost, accurate, and continuous skin monitoring across varying skin tones in a wearable form-factor. This paper introduces DermaGlow, a novel wearable optical spectroscopy framework designed for low-cost, non-invasive monitoring of melanin, erythema, and skin tone. DermaGlow utilizes an off-the-shelf multi-spectrum wearable device available in various configurations to enable real-time, personalized assessments across diverse skin conditions and skin tones. We assess the performance of the DermaGlow algorithm against a state-of-the-art colorimeter in a comprehensive user study involving a diverse group of 77 subjects, demonstrating a normalized mean absolute error (NMAE) of 5.33% (melanin) and 4.18% (erythema), and ΔE values less than 2.5 for CIE LAB measurements. Furthermore, we present an algorithm that utilizes DermaGlow outputs to correct for pulse oximeter inaccuracies typically found in those with darker skin pigmentation, resulting in an up to 75% decrease in mean absolute error (MAE) in hypoxic readings across skin tones relative to arterial blood measurements. Our findings highlight DermaGlow's potential for short and long-term skin monitoring and as a significant enhancement to existing wearable devices, particularly in improving the accuracy of pulse oximeter readings across different skin tones.

  PublisherDOI

• Characterization and Feasibility of Wearable Spectroscopic Tracking of Nutrition Biomarkers

  IEEE Pervasive Computing · 2025-07-01

  articleSenior author

  Disorders related to nutrition and metabolism remain a critical public health challenge globally, affecting a significant portion of the population. While traditional calorie tracking methods help monitor dietary intake, they do not capture how the body synthesizes and metabolizes nutrients, and are often inaccurate, cumbersome, and inconvenient. Wearable biosensing technologies present a promising avenue for more precise, noninvasive monitoring of nutritional intake, providing individuals with real-time feedback to improve their dietary habits. In this study, we explore the feasibility of using an off-the-shelf wearable device to track nutrients in the blood. Utilizing a multiwavelength optical sensor covering a range of wavelengths in the visible spectrum, we compare the device’s ability to measure spectral absorption profiles of various nutritional molecules relative to a spectrophotometer. Through advanced signal processing and statistical techniques, we are able to successfully recover molecular signatures from the wearable device relative to the spectrophotometer. Our results demonstrate the potential of wearable biosensing technology to track nutritional intake, offering a new tool for effectively managing nutritional disorders through real-time monitoring.

  PublisherDOI

• Development and Evaluation of a Metric-based Clinical Simulation Procedure for Assessing Ostomy Care in Nursing Practice

  Clinical Simulation in Nursing · 2024-08-01 · 2 citations

  articleOpen access

  Background: There is a need to understand the clinical decision-making and work practices within ostomy nursing care to support expanding nursing training. Objective: To develop and evaluate a new metric-based simulation for assessing ostomy nursing care using a human factors approach. Sample: This pilot study involved eleven stakeholders in the needs assessment, six nurse participants performing simulated ostomy care, and three independent observers assessing procedure reliability. Method: We conducted a needs assessment of ostomy nursing care and training, developed an enhanced metric-based simulation for ostomy appliance change procedures, and statistically evaluated its reliability for measuring the simulated tasks. Results: The enhanced metric-based simulation captured different tasks within four task categories: product selection; stoma and peristomal skin care; baseplate sizing and adhesion; and infection control strategies. The video review procedure was reliable for assessing continuous (average ICC≥0.96) and categorical (average κ>0.96) variables. Conclusion: The new metric-based simulation was suitable for characterizing a broad range of clinical decision-making and work practices in ostomy nursing care.

  PublisherOA PDFDOI

• Using Decision Tree Classifier to Increase Screening Test Sensitivity for the Prediction of ACL Retear

  2024-09-22

  articleSenior author

  Screening tests are often used in medicine to assess whether a patient is at a high risk of contracting a disease. Recent literature has proposed prediction algorithms for Anterior Cruciate Ligament (ACL) retears that aim to achieve high accuracy. However, these models fail to reach an adequate sensitivity to function as effective screening tests. In such cases, model sensitivity is sacrificed for heightened specificity. Misclassifying patients who will eventually go on to retear their ACL as low-risk patients prevents them from obtaining necessary therapeutic support and is not appropriate for a clinical setting. In this study, we implement a Decision Tree Classifier as a screening test to evaluate a patient's risk of retearing their ACL six months after surgery, before the patient is released to activity. By incorporating a machine learning-based screening technique, we hope to minimize false negatives and create a tool that can readily be adopted in clinical practice.

  PublisherDOI

• Raproto: An Open-Source Platform for Rapid Prototyping with Wearable Devices

  2024-10-15

  articleSenior author

  Advances in wearable technology have enabled ubiquitous use of wearable devices in remote patient monitoring, particularly in clinical trials. Because of the reliance on highquality data in these endeavors, the first and often the most time-consuming step is to build a data collection system. While many systems have been developed to address this, they are often highly specific and customized to the task at hand, and are often not generalized enough to support other tasks. To remedy this, we developed Raproto, an open-source easy-to-use rapid prototyping platform that does not require the time, effort, and expertise needed for custom development. The Raproto platform consists of three components, the wearable device(s), communication protocol, and remote storage. These components support the collection, transmission, storage, analysis, and visualization of large-scale data with applications from smaller-scale research studies to large clinical trials. To reduce the burden of device and application development, we created multipurpose and customizable smartwatch applications on both the Android and Tizen operating systems. We evaluate our platform in a lab setting as well as in two real-world case studies. Overall, we find that we can collect data using our application for over 24 hours on a single charge and there is little to no data loss, thus making it an ideal tool to preface customized device development for real-world impact and commercialization.

  PublisherDOI

Frequent coauthors

• Gang Zhou

  Xinjiang University

  18 shared

• Woosub Jung

  Towson University

  11 shared

• Minglong Sun

  11 shared

• James Weimer

  9 shared

• Insup Lee

  9 shared

• Kenneth Koltermann

  William & Mary

  6 shared

• Shuangquan Wang

  Salisbury University

  5 shared

• Claire Kendell

  University of Pennsylvania

  4 shared

Labs

• Link LabPI