The Malone Center for Engineering in Healthcare recently awarded six grants through its first round of funding from its 2025 Seed Grant Program in collaboration with the School of Nursing and the Department of Surgery, supporting engineering innovations that aim to:

  • Develop an AI-powered, personalized nutrition model for patients with cardiovascular-kidney-metabolic syndrome;
  • Create a real-time reinforcement learning system to optimize sedation management in pediatric intensive care units;
  • Test an AI-enhanced, robot-supported physical activity intervention for sedentary older adults;
  • Develop autonomous robotic surgical capabilities for xenotransplant kidney procurement;
  • Create a transformative, tailored motor function recovery risk calculator; and
  • Integrate genomic biomarkers with mental health screening for early identification of PTSD in vulnerable populations.

With the funding for this round of projects, the center will have awarded a total of 21 seed grants since the program’s inception in 2018. To learn about the past projects, visit the Malone Seed Grant page.

In collaboration with the School of Nursing, the Malone Center will fund the following three projects:

AI-Powered Personalization Engine for Precision Nutrition in Chronic Disease Management
Principle Investigators: Bunmi Ogungbe (Nursing); Ilya Shpitser (Computer Science/Malone)

Precision nutrition holds promise for improving chronic disease management by tailoring dietary interventions to individual patients. However, current approaches lack scalable personalization methods that can adapt to individual barriers, preferences, and clinical profiles. This project will develop and validate an AI-powered personalization engine that analyzes dietary patterns, cultural preferences, and clinical data to generate tailored nutrition recommendations for patients with cardiovascular-kidney-metabolic syndrome.

Using machine learning algorithms and causal inference methods applied to existing trial and institutional electronic health record data, the research team will create personalized models across individual, clinical, cultural, and environmental dimensions. The team will establish a “dietitian-in-the-loop” model that combines AI efficiency with clinical expertise through causal reinforcement learning, enabling continuous algorithm improvement based on expert feedback and further establishing Johns Hopkins as a leader in precision nutrition technology.

Real-Time Dose Optimization for Pediatric ICU Sedation Using Reinforcement Learning
Principal Investigators: Jessica LaRosa (Anesthesiology and Critical Care Medicine); Sapna Kudchadkar (Anesthesiology and Critical Care Medicine); James “Jim” Fackler (Anesthesiology and Critical Care Medicine/Malone); Nicholas Durr (Biomedical Engineering/Malone); Kristen Brown  (Nursing)

Currently, more than 340,000 children receive pediatric critical care annually in the United States, yet only 58% achieve optimal sedation levels, with sedation-associated adverse events occurring in up to 70% of cases. Existing sedation assessments depend on subjective nursing observational scales that are administered infrequently and show significant variability. This project aims to develop a real-time reinforcement learning system that optimizes sedation management in pediatric intensive care units (PICUs) by integrating pharmacokinetic modeling with continuous physiological monitoring data.

The team’s approach combines multi-drug pharmacokinetic models for common PICU sedatives with accelerometry and vital signs data to train distributional reinforcement learning algorithms that automatically predict sedation-agitation scores. This automated decision-support system will provide objective, continuous assessment of patient sedation states, enabling more accurate dosing decisions, reducing nursing workload, and establishing a foundation for AI-driven clinical decision support in pediatric critical care and nursing simulation training.

Leveraging AI and Wearables to Promote Physical Activity Behavior Change in Sedentary Older Adults: A Pilot Trial of a Robot-Supported Fitbit Intervention
Principle Investigators: Junxin Li (Nursing); Chien-Ming Huang (Computer Science/Malone)

Physical activity is critical for healthy aging, yet many older adults face behavioral and functional barriers that limit sustained engagement. Wearable devices like the Fitbit can support self-monitoring, but often have limited impact on long-term physical activity behavior change when used alone. When combined with behavioral strategies—such as goal setting, personalized feedback, and motivational support—wearables are more effective in promoting lasting activity in older adults. Building on the ongoing NIH-funded R01 mPATH trial, which integrates Fitbit use with phone or in-home human physical activity coaching, this pilot study will test the feasibility, acceptability, and preliminary efficacy of “Robot-Fitbit,” a novel, AI-enhanced, robot-supported physical activity intervention for sedentary older adults.

This collaborative project aims to evaluate the feasibility and acceptability of the Robot-Fitbit intervention; assess its impact on physical activity behavior change (e.g., step count, active minutes, moderate to vigorous physical activity, sedentary time); examine preliminary effects on key health outcomes (e.g., cognitive performance, physical function, sleep and mental health, etc.); and develop a measurement approach for attributing the costs of developing and implementing the Robot-Fitbit intervention. Findings from this pilot will inform future NIH-funded trials aimed at advancing scalable, cost-effective, AI-enhanced interventions to promote physical activity and healthy aging, with strong potential for implementation in resource-limited and underserved settings.

In collaboration with the Department of Surgery, the Malone Center will fund:

Autonomous Robotic Surgery for the Procurement of Xenotransplant Kidneys at Scale
Principle Investigators: Axel Krieger (Mechanical Engineering/Malone); Jeff Jopling (Surgery)

Over 100,000 Americans with renal failure await life-saving kidney transplants, with demand far exceeding supply. Xenotransplantation offers a revolutionary solution through gene-edited pig organs. Once established as a viable therapy for end-stage kidney disease, procuring pig organs at the necessary scale will be an unprecedented challenge due in part to surgical workforce limitations, potentially preventing patients who could benefit from this technology from receiving it.

This project aims to mitigate this problem by developing autonomous robotic surgical capabilities for precise ureter management during xenotransplant kidney procurement. By leveraging foundation models and embodied AI, the research team will create a system capable of autonomously identifying, dissecting, and preserving the ureter during kidney procurement procedures. The team’s approach combines expertise in surgical robotics, AI, and transplantation medicine to demonstrate the feasibility of autonomous execution of this critical step in both ex vivo and in vivo settings. This project represents a crucial first step toward scalable xenotransplantation and addresses an urgent public health need while establishing a foundation for future autonomous surgical applications.

Finally, the Malone Center will fund the following two projects in precision medicine:

AI-Based Approaches to Quantify Muscle Denervation After Motor Nerve Injury and Predict Functional Outcomes After Motor Nerve Surgical Repair
Principle Investigators: Haris Sair (Radiology and Radiological Science/Malone); Sami Tuffaha (Plastic and Reconstructive Surgery)

Peripheral nerve injuries (PNIs) are debilitating consequences of trauma, affecting up to 5% of trauma patients and disproportionately impacting young, working-age individuals, with 80–90% of patients under 50 years old. Underdiagnosis is widespread, with only 30% of PNIs diagnosed within 90 days of injury, and functional recovery of motor nerve injuries depends critically on timely reinnervation, typically within 3–6 months. Most surgical nerve repairs are delayed due to two primary factors: observation periods to assess for spontaneous recovery in injuries-in-continuity and delayed specialist referral—averaging 11 months—frequently resulting in missed therapeutic windows and irreversible muscle atrophy with potential loss of limb function. In patients presenting 9–15 months post-injury, decisions regarding nerve reconstruction versus secondary reconstructive procedures (e.g., tendon or free functional muscle transfers) become increasingly complex.

This project addresses a critical gap in the field by developing and validating an AI-driven platform to extract quantitative radiomic biomarkers of denervated muscle via fluid-sensitive T2-weighted MRI. By integrating imaging biomarkers with clinical variables (patient demographics, surgical details, and peri-operative data), the research team aims to develop an objective, patient-specific machine learning-enhanced tool to predict denervated motor function recovery metrics. Overall, this project seeks to develop a transformative motor functional recovery risk calculator that tailors motor nerve injury interventions to individual patient characteristics to improve long-term limb function and quality of life.

Integrating Multiomic Signatures and Inflammatory Biomarkers to Improve PTSD Diagnostic Precision in Trauma-Exposed Populations
Principle Investigators: Tamar Rodney (Nursing); Alexis Battle (Biomedical Engineering/Computer Science/Malone)

This interdisciplinary project seeks to advance the field of mental health precision care by integrating genomic biomarkers with clinical mental health screening for early identification of post-traumatic stress disorder in vulnerable populations, including youth and veterans. By combining psychiatric nursing and community-based trauma care with machine learning and genomics, the project aims to create a scalable, predictive tool for PTSD risk stratification. This collaboration will foster innovation at the intersection of nursing science and bioinformatics, with implications for policy reform, early intervention, and improved mental health outcomes in underserved communities.

The Malone Center Seed Grant Program aims to assist faculty and research staff with development of innovative, collaborative proposals that will advance the Malone Center mission. New partnerships and research directions are often achieved, opening up opportunities that may otherwise not come to fruition.