Research Focus

The Laboratory of Computational Medicine at Johns Hopkins University is dedicated to advancing the understanding and management of critical illnesses through a multidisciplinary approach that integrates precision medicine, computational modeling, and data science.

Precision Intensive Care Medicine

We aim to tailor intensive care treatments to the unique biological, physiological, and clinical profiles of individual patients. This involves the application of deep phenotyping, nosographic and nosological advancements, and selective treatment strategies. Key projects in this area include:

  1. Development of models for individualized treatment effects (e.g., oxygen targets in ventilated patients)​
  2. Quantitative approaches to neurological prognostication post-cardiac arrest​
  3. Early detection and prediction of critical conditions such as sepsis and stroke​
  4. Developing AI-driven systems for patient phenotyping with acute neurological syndromes​
  5. Explainable AI models for surgical risk prediction using preoperative ECG data

Biological and Physiological Modeling

We explore the mechanisms underlying health and disease by analyzing networks, signaling pathways, and physiological state transitions. Our work emphasizes the integration of high-dimensional data to improve model efficiency and efficacy. Notable projects include:

  1. Predictive modeling of intracranial pressure using extracranial physiological waveforms
  2. Time-series analyses for predicting outcomes in critical care settings, such as post-cardiac arrest recovery trajectories​

Neurocritical Care and Functional Recovery

We investigate the neurobiological basis of recovery following critical neurological injuries, employing techniques such as connectome analysis and accelerometry. Key studies focus on:

  1. Functional connectome integrity and its role in recovery from cardiac arrest and subarachnoid hemorrhage​
  2. Decoding motor function and neurological states in ICU patients with severe brain injuries​

Sensor-Based Monitoring and Fusion

To enhance patient monitoring, we develop sensor fusion techniques that integrate data from wearable devices, physiological waveforms, and environmental inputs. Ongoing projects include:

  1. Multi-modal feature extraction for predicting autonomic and central nervous system responses​
  2. Real-time detection of clinical deterioration through non-invasive monitoring systems

Vision for the future

The lab is committed to translating research insights into actionable interventions that improve patient outcomes, reduce the burden of critical illnesses, and bridge the implementation gap in healthcare systems. Our interdisciplinary collaborations across engineering, medicine, and computational sciences drive innovation in precision and personalized medicine.