Conformable Cranial Ultrasound Patch for Monitoring Neonates with Germinal Matrix Hemorrhage

Project Summary/Abstract Germinal matrix-intraventricular hemorrhage (GM-IVH) is a significant neurological complication associated with high mortality rates and substantial neurodevelopmental disabilities. While the majority of GM-IVH cases are clinically asymptomatic, it is the most common cause of hydrocephalus in premature infants. Progressive cerebral ventricular dilation is an important diagnostic component of hydrocephalus and is typically identified through trans-fontanelle, cranial Ultrasound (CUS) in neonates. It is safe, cost-effective and can be conducted at the bedside with minimal disruption to the infant. However, current CUS clinical application is limited by numerous constraints, including: i) results are contingent upon the skills and experience of the ultrasonographer and radiologist Iii) limited access to the appropriate equipment and trained personnel in certain institutions due to prohibitive costs iii) due to the medical complexity of this vulnerable patient population, there is often a need to minimize stress, which can delay acquisition of important imaging when a patient is too unstable to undergo standard, often time consuming diagnostic US assessment. Within the last six years, conformable ultrasound electronics have been intensively investigated for imaging of many internal organs, however, to our knowledge, there are currently no studies on the feasibility of trans- fontanelle continuous ventricular ultrasound. Our goal is to investigate the monitoring of ventricular volume in neonates with GM-IVH using a wearable, adhesive ultrasound patch, and test the feasibility of simultaneous measurement of ventricular and sub-arachnoid size as well as cerebral blood flow. We hope this research will standardize interpretation, increase availability and reduce costs. Our work will introduce a novel patch design along with advanced piezoelectric transducers design, and a new image reconstruction method along with machine learning analysis of standard measurements such as ventricular index (VI), and anterior horn width (AHW), but also introduce AI based volumetric analysis. This work will be based on (1) novel patch design and advanced microfabrication of electronics (electronic science and engineering), (2) signal decoding for beamforming and image reconstruction (biomedical engineering and signal processing) and (3) clinical study on neonates and machine learning analysis (biomedical engineering and artificial intelligence). This study will provide the first in vivo validation of a conformable cranial ultrasound patch for neonatal brain monitoring with a significant advancement in neonatal monitoring, combining state-of-the-art piezoelectric sensor technology with advanced deep learning algorithms. We aim to demonstrate generalizability, robustness, and the potential to standardize CUS in this at-risk patient populations. This could ultimately reduce the incidence of severe neurodevelopmental impairment by providing uniform neurodiagnostic accuracy in a condition that is a major cause of mortality and neurodevelopmental impairment in this fragile population.