Researchers identify a less invasive approach to measure intracranial pressure in patients

Researchers at the Johns Hopkins University School of Medicine have published a new study reporting a potential, less invasive, alternative approach to measuring intracranial pressure (ICP) in patients.

The study was published July 12 in the journal Computers in biology and medicine.

ICP is a physiological variable that can be abnormally elevated in cases of acute brain injury, stroke, or obstruction of cerebrospinal fluid flow. Symptoms of elevated ICP include headache, blurred vision, vomiting, behavioral changes, and decreased level of consciousness. This can be life-threatening, so ICP monitoring is necessary in certain high-risk patients. However, the current standard for ICP monitoring is highly invasive; it requires drilling the skull to place external ventricular drains (EVDs) or intraparenchymal brain monitors (IPMs) into the functional brain tissue, which consists of neurons and glial cells.

“ICP is widely recognized as an important vital sign. There is an urgent need to measure and treat ICP in patients with critical neurological diseases, but the current standard of care for ICP measurement is invasive, risky and resource intensive. We therefore investigated a novel approach leveraging artificial intelligence, which we believe may be a viable non-invasive alternative method of ICP assessment.”

Robert Stevens, MD, MBA, Chief of Precision Medicine, Department of Anesthesiology and Critical Care

Recent studies have shown that EVD procedures carry a 15.3%, 5.8% and 12.1% risk of catheter misinsertion, infection and bleeding, respectively. Additionally, EVD and IPM procedures require surgical expertise and specialised equipment that are not always available in many settings, necessitating the need for alternative methods to test and monitor patients' ICP.

The Johns Hopkins University School of Medicine team, an interdisciplinary group led by Stevens, hypothesized that severe brain injury, and in particular elevated ICP, is associated with pathological changes in whole-body cardiovascular function, including through dysregulation of the central autonomic nervous system. This hypothesis suggests that studying extracranial physiological waveforms may provide a deeper understanding of brain activity and the severity of ICP.

In this study, a research team from Johns Hopkins University aimed to investigate the relationship between ICP waveforms and three physiological waveforms routinely recorded in the ICU: invasive arterial blood pressure (ABP), photoplethysmography (PPG), and electrocardiogram (ECG). Using ABP, PPG, and ECG data, a panel of different deep learning algorithms were trained and compared, resulting in accuracy of ICP determination that was comparable or superior to other methodologies.

Overall the study findings suggest a completely new, non-invasive alternative for monitoring patients’ ICP.

“Through validation, physiology-based AI solutions such as those used here have the potential to significantly expand the proportion of patients and care settings that can be offered ICP monitoring and management,” Stevens said.

Co-authors include recent Johns Hopkins Biomedical Engineering graduates Shiker Nair ’23, Alina Guo ’23, Arushi Tandon ’23 and Joseph Boen ’22, master’s student Meer Patel, fourth-year Biomedical Engineering students Atas Aggarwal, Ojas Chahal and Sreenidhi Sankararaman, Nicholas D. Durr, associate professor of Biomedical Engineering, Tej D. Azad, a Johns Hopkins neurosurgery resident, and Romain Pirracchio, professor of anesthesiology at the University of California, San Francisco.