Research Interests

My research focus is on advancing our understanding of human heath and performance through interpreting complex physiological and biomechanical signals measured using wearable sensors. My research journey is rooted in a deep commitment to practical applications that impact real-world health and performance. My current work focuses on understanding the interplay of sleep, circadian rhythms, and the menstrual cycle, and their collective impact on neuromuscular performance in women.

Past Research Highlights

Remote Monitoring of Activity Patterns & Sleep

Changes in activity patterns and sleep have been used as a biomarker in many mental health, age-related, and neurodegenerative diseases. This work has focused on building out and assessing wearable sensor-based tools for remotely and continuously monitoring activity patterns and sleep using low-cost devices. In these projects, I developed algorithms and analytical frameworks for assessing sleep and understanding the data requirements for analysis of activity patterns. This work supports the remote monitoring of disease dynamics over time that affects both daytime and nighttime behaviors and physiology.

  1. Weed, L., Lok, R., Chawra, D., Zeitzer, J. The Impact of Missing Data and Imputation Methods on Calculation of 24-Hour Activity Patterns. Clocks & Sleep 4(4):497-507, 2022.

  2. Telfer, B., Williamson, JR., Weed, L., Bursey, M., Frazee, R., Galer, M., Moore, C., Buller, MJ., Friedl, K. Estimating Sedentary Breathing Rate from Chest-Worn Accelerometry from Free- Living Data. IEEE EMBC/CMBES 2020, Montreal, Canada.

  3. Weed, L., Gurchiek, R., Tulipani, L., Meyer, B., Allen, D., Ursiny, A., Solomon, A., McGinnis, R. Sleep Detection and Disturbance Characterization from Chest Accelerometer for Multiple Sclerosis. BMES 2020, San Diego, CA. Oral Presentation

Advancements in Pedestrian Tracking

Recent advances in the size, weight, and power of wearable sensors enable continuous and remote monitoring of inertial signals for long durations under real-world conditions. I developed novel methods of performing zero-velocity updates for Kalman filter-based sensor fusion at both foot and shank locations enabling accurate estimation of step trajectory and spatiotemporal gait parameters, typically extracted in an in-lab environment, during multiple tasks and real-world environments including gait, sprinting, obstacle negotiation, and stair climbing. These algorithms have been used to understand the impacts of diseases such as Multiple Sclerosis, risk of injury during military performance, and for biomechanical assessment of equipment for Emergency Medical Services personnel resulting in a first-author publication and multiple conference presentations.

  1. Weed, L., Little, C., Kasser, S., McGinnis, R. A Preliminary Investigation of the Effects of Obstacle Negotiation and Turning on Gait Variability in Adults with Multiple Sclerosis. Sensors 21(17):5806, 2021.

  2. Weed, L., Fellin, RE., Davis, SK., Buller, MJ. Pedestrian Movement Tracking Using Adaptive Zero-Velocity Updates from Shank IMU. ASB/ISB 2019, Calgary, Canada. Oral Presentation.

  3. Fellin, RE., Weed, L., Davis, SK., Welles, AP., Seay, JF., Buller, MJ. Characterizing Marching Gait Parameters in the Field During Load Carriage Using a Shank-Borne Sensor. ASB/ISB 2019, Calgary, Canada. Oral Presentation.

  4. Weed, L., Petrillo, C., Adamowicz, L., McGinnis, RS. Effect of EMS Loading Configuration on Stair Ascent and Descent Biomechanics Using a Kalman Filter and Wearable Inertial Sensors. BMES 2018, Atlanta, Georgia. Oral Presentation.

Exertional Heatstroke Prediction

Exertional heat injury is a major, preventable physiological episode that can result in lasting damage and even death. During sports and military training in hot and humid climates, there is elevated risk for heat stroke. I developed algorithms relying on sensor fusion of biomechanical and physiological signals supporting heat stroke prediction minutes before a physiological episode enabling the delivery of just-in-time care to prevent severe consequence of heat injury. This system and method have been patented and is now used in military training settings to actively prevent heat injuries.

  1. Palmer, J., Telfer, B., Williamson, J., Weed, L., Buller, M., Fellin, R., Seay, J., System and Method for Predicting Exertional Heat Stroke with a Worn Sensor. US Patent No US20210338173A1. Filed 24 June 2020. Published 04 November 2021.

  2. Buller, M., Fellin, R., Atkinson, E., Beidleman, B., Marcello, M., Driver, K., Mesite, T., Seay, J., Weed, L., Telfer, B., King, C., Frazee, R., Moore, C., Bursey, M., Galer, M., Williamson, J. Gait instability and estimated core temperature predict exertional heat stroke. Br J Sports Med 56(8):446-451, 2022.

Prediction of Circadian Rhythms from Wrist-Worn Wearable Sensors

In the field of circadian rhythms, until recently, the central clock has been studied in very expensive, laborious protocols that require a highly controlled laboratory environment. My work has focused on adapting models previously developed from in-laboratory experiments for use in real-word environments using wearable sensors. This work supports observing phenomena of the central clock in ambulatory conditions not previously studied. Additionally, these models enable the prediction of the circadian rhythm under unusual lighting conditions, such as shift work and jetlag, which have been shown to heave detrimental impacts in acute settings on both physical and mental performance and under chronic conditions, on long term health outcomes.

  1. Weed, L., Zeitzer, JM. Predicting Circadian Rhythms from Wrist-Worn Wearable Sensors. SRBR 2022, Amelia Island, FL. Poster Presentation.

  2. Weed, L., Zeitzer, JM. Advancements in Predicting Circadian Rhythms from Wrist-Worn Wearable Sensors. MHSRS 2022, Kissimmee, FL. Poster Presentation.

Monitoring Rehabilitation Progress

Quantitatively monitoring rehabilitation progress is challenging in both in-patient and out-patient settings. In this work, I developed and evaluated wearable sensor and iPhone app-based assessment tool for quick, easy, and convenient clinical assessment. This included developing a step detection algorithm customized for stroke patients with slow, irregular gait that is robust to “step-like” events that occur during rehab sessions and corresponding computer application that is now used in an in-patient stoke rehabilitation clinic. Additionally, I validated the GaitAnalysisPro iPhone application for used in a 10m walk test using wearable sensors for use in assessing patients after hip replacement surgery. Each of these are currently used as qualitative assessment tools in rehabilitation.

  1. Weed, L., Robinson, J., Goodwin, L., McGinnis, RS. Open-Source Wearable Sensor Based Method Feasible for Tracking Steps in Patients Recovering from Stroke. BMES 2018, Atlanta, Georgia. Poster Presentation.

  2. Robinson, JR., Goodwin, LB., Fulk, GD., Borland, R., McGinnis, RS., Weed, L. Implementing High Intensity Gait Training to Improve Recovery Following Stroke Using Knowledge Translation. APTA CSM 2020, Denver, CO. Poster Presentation.

  3. Weed, L., McGinnis, RS. Validation of GaitAnalysisPro App for 10m Walk Test. BMES 2017, Phoenix, Arizona. Oral Presentation.

For a comprehensive list of my publications, see my ResearchGate or Google Scholar.