Abstract
Intracranial pressure (ICP) is an important measurement in the treatment of Traumatic Brain Injury (TBI). Currently, ICP can only be measured invasively, which exposes patients to operative risk and can only be performed by neurosurgeons. Hence, there is a significant need for a non-invasive ICP technology. This paper describes the evaluation of a novel non-invasive intracranial pressure (nICP) monitor which uses the Photoplethysmogram (PPG) to measure the ICP. The monitor was evaluated in an in vitro model that simulated cerebral haemodynamics and allowed the controlled manipulation of ICP. A number of features from the PPG were extracted and utilised in a machine learning model to estimate ICP. Three separate measurements in which the ICP was varied were performed, and the estimated ICP (nICP) was compared with reference (invasive) ICP measurements. The ICP estimated by the nICP monitor was highly correlated with reference ICP measurements (Pearson’s correlation coefficient between 0.95 and 0.98). The nICP monitor also showed a low Root Mean Square Error from the reference ICP measure (3.12, 1.48, and 1.45 mmHg). Analysis of agreement by Bland and Altman also revealed good agreement between the two techniques. The optical nICP monitor was able to estimate the ICP non-invasively from an in vitro model simulating intracranial hypertension. The non-invasive ICP monitor showed very promising results which can set the base for further investigations. This work contributes significantly to the quest for non-invasive ICP monitoring in Traumatic Brain Injury (TBI), and paves the way for further research in this field.
Funder
National Institute for Health Research
Subject
Fluid Flow and Transfer Processes,Computer Science Applications,Process Chemistry and Technology,General Engineering,Instrumentation,General Materials Science
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献