Utility of the Tympanic Membrane Pressure Waveform for Non-invasive Estimation of The Intracranial Pressure Waveform

Abstract

Time domain analysis of the intracranial pressure (ICP) waveform provides important information about the intracranial pressure-volume reserve capacity. The aim here was to explore whether the tympanic membrane pressure (TMP) waveform can be used to non-invasively estimate the ICP waveform. Simultaneous invasive ICP and non-invasive TMP signals were measured in a total of 28 individuals who underwent invasive ICP measurements as a part of their clinical work up (surveillance after subarachnoid hemorrhage in 9 individuals and diagnostic for CSF circulation disorders in 19 individuals). For each individual, a transfer function estimate between the invasive ICP and non-invasive TMP signals was established in order to explore the potential of the method. To validate the results, ICP waveform parameters including the mean wave amplitude (MWA) were computed in the time domain for both the ICP estimates and the invasively measured ICP. The patient-specific non-invasive ICP signals predicted MWA rather satisfactorily in 4/28 individuals (14%). In these four patients the differences between original and estimated MWA were <1.0 mmHg in more than 50% of observations, and <0.5 mmHg in more than 20% of observations. The study further disclosed that the cochlear aqueduct worked as a physical lowpass filter.

Figure 1

Schematic illustration of the anatomical structures involved in measurements of TMP waveforms. In this study, non-invasive TMP waveforms were measured in the outer ear and used as input for the estimation of non-invasive ICP. The labels represents the following; ICP: ICP input signal, CA: Cochlear Aqueduct, OW: Oval Window, RW: Round Window, T: Tympanic membrane and S: Sensor.

Figure 2

An example showing 6 seconds of the input signals and the corresponding transfer function estimate based on a total 10 minutes is shown in Figure (a,b) respectively (patient ID 3). The resulting output from the combination of Fig. 1a and the inverse of Fig. 1b is shown in Fig. 1c.

Figure 3

The non-invasive ICP waveform estimate (nICP, interrupted red line) is shown together with the invasive ICP waveform (continuous red line) for four different 6-second time windows after the beginning of the measurement (patient ID 3). The time delay between the nICP and ICP signals, as seen in Fig. 2a, has been removed for visual comparison.

Figure 4

The averaged absolute difference in Mean Wave Amplitude (MWA), Mean Wave Rise Time (MWRT) and Mean Wave Rise Time Coefficient (MWRTC) between the non-invasive ICP estimate and the measured ICP signal is shown for each patient ID. The error bars illustrate the patient specific standard deviation (mean ± stdev).