Three clinical studies have demonstrated the safety and effectiveness of the B4C System and its intended use.
Study 1 - Non-invasive ICP Monitoring for HIV-associated Cryptococcal Meningitis
A critically ill adult subject diagnosed with human immunodeficiency virus-associated cryptococcal meningitis received ICP monitoring with a previous sensor version of the B4C System while undergoing standard treatment over thirty-four (34) days. The subject underwent real-time non-invasive ICP monitoring at four defined time points before and after treatment to collect four ICP waveforms. Waveform morphology of the ICP pulses at these time points was visually assessed with other recorded clinical parameters to determine whether the waveforms were indicative of the clinical status of the patient. The pulsatile waveform from ICP monitoring on Day 12 before lumbar puncture revealed P2>P1, amplitude of tidal wave greater than that of percussion wave, reflecting characteristics of relative peak height consistent with the presence of neurological symptoms. P1>P2 after lumbar puncture, demonstrating improvement towards the characteristic P1>P2>P3, where P3 is dicrotic wave, as expected with reduction in ICP post-treatment. Morphology of the waveforms obtained from Day 34 were more closely representing normal brain compliance (P1>P2>P3), which is consistent with reduction in ICP following the series of treatment and discharge that same day. Results of this early study demonstrated that the B4C System is able to continuously monitor ICP changes to acquire signals consistent with the patient’s clinical status.
Study 2 - Analysis of a Non-Invasive ICP Monitoring Method in Patients with Traumatic Brain Injury
Seven adult subjects who were admitted to the neurointensive care unit presenting severe or moderate traumatic brain injury with secondary neurological deterioration requiring intubation and mechanical ventilation were enrolled in the study. The objective of the study was to verify the similarities between the Codman® Microsensor Basic Kit (iICP) and Braincare (nICP) sensors’ waveforms. This assessment sought to provide evidence for the use of the noninvasive sensor as an alternative to invasive ICP assessments in situations where the waveform can provide supplementary clinical information. In addition, the noninvasive intracranial pressure and arterial blood pressure (ABP) waveforms were compared to verify the possible influence of the extracranial peripheral circulation into the noninvasive intracranial pressure signal, acknowledged as a potential limitation of the B4C System.
Each subject underwent continuous ICP monitoring using an invasive and a previous sensor version of the B4C System concurrently from point of admittance throughout their stay in the neurointensive care unit, with acquisition time ranging from 68-282 hours. ABP measurement directly through the radial artery and partial pressure of Carbon Dioxide (PaCO2) were also recorded simultaneously during the monitoring sessions. Approximately 337 total hours of recordings were analyzed.
The primary endpoint was the comparison of ICP waveform morphology obtained with the nICP and iICP sensors. A secondary endpoint was the comparison of the nICP and ABP waveforms. Waveforms were compared in a lower dimensional space constructed based on signals in the frequency domain. Similarity between the two devices’ signals was inferred from the Euclidean distance between the non-linear projection in a lower dimensional space of the window power spectral densities (PSD) of the respective signals, in which PSD was calculated using the short- term Fourier transform. Intra-individual statistical comparisons were performed using the non-parametric Mann-Whitney U test for not normally distributed data points with a significance level set at p<0.05. Measurement of similarities are presented in the following table.
Measure of similarities between iICP, nICP and ABP
Subject ID | |||||||
|---|---|---|---|---|---|---|---|
Similarities | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
iCP-niICP | 35.0 | 27.2 | 26.9 | 16.9 | 36.3 | 54.7 | 30.3 |
ABP-niICP | 117.3 | 86.6 | 86.9 | 77.9 | 78.1 | 106.6 | 76.3 |
The difference between the iICP-nICP and nICP-ABP was found to be statistically significant for all seven subjects, p<0.05, using the Mann-Whitney U test. Study results demonstrated that a greater similarity exists between the waveforms generated from the signals of the Codman® Microsensor Basic Kit and B4C System than between the B4C System sensor and arterial blood pressure measurements. Although the study had a limited sample size, the intra-individual similarities of the invasive and noninvasive ICP signals as functions of time suggest comparable effectiveness of ICP monitoring between the Braincare Sensor and the invasive Codman® Microsensor Basic Kit, which is representative of the standard of care. Additionally, no adverse events related to the use of the B4C System were reported, supporting safety of the Braincare sensor for monitoring intracranial pressure.
Study 3 - Evaluation of Multi-Center Clinical Data On a Novel Non-Invasive Sensor Device for Safe and Simplified Intracranial Pressure (ICP) Monitoring
Braincare conducted a combined prospective/retrospective, multi-center, observational study to assess the comparison of the acquired ICP waveform signal and parameters between the B4C System and standard of care invasive ICP monitoring methods. The study device consisted of the B4C processing and analytical software used with the wired sensor (K182073). Although the wired sensor was used in the study, the results reflect the performance of the B4C System.
Dataset Description
Total number of centers: 4
Total number of subjects: 123 enrolled, 107 after device label check, 85 after data quality check (78 adults, 7 pediatric)
Collected data: ICP Surrogate Waveform (BcSs-PICNI-2000(K182073) or B4C System) ; Invasive Arterial Blood Pressure, Invasive ICP Waveform (EVD or Bolt)
Range of acquisition sessions time: 5 min to 3.5 hours
Total number of monitoring sessions that passed quality check: 159
Total acquisition time that passed quality check: 4800 (98% adult, 2% pediatric)
Analyzed participants
Seventy eight adults (18+) who met all eligibility criteria and were admitted to the neurointensive care unit and underwent invasive ICP monitoring and invasive arterial blood pressure monitoring were considered in the dataset. Due to the reduced quantity of pediatric subjects, the analysis could only demonstrate statistically relevant performance for the adult population.
Total number of analyzed subjects: 78 adults
Total acquisition time analyzed: 4695 minutes
Age of analyzed subjects: 52.7±19.4
Gender of analyzed subjects: 47% female ; 53% male
Study Objective
The goal of the analysis was to verify whether the new medical device developed by Braincare demonstrated a consistent correlation between its recorded waveform with the invasive devices waveform that are currently used in clinical practice. The objective was to evaluate the reliability and accuracy of the correlation between the Braincare device in monitoring ICP waveform in comparison to gold standard invasive ICP monitoring methods such as the external ventricular drain or intraparenchymal micro transducers that are currently used in clinical practice, and utilized in the target population in different centers and medical settings.
Study Procedures
All centers used Braincare’s non-invasive sensors with identical principles of operations (3 centers with BcSs-PICNI-2000 sensor (K182073), 1 with B4C System wireless sensor). The sensors at each site were positioned according to the same protocols, i.e., temporal region avoiding arteries and adjustment to the point that an acceptable waveform appeared, a procedure that represents real case usage. Patients at all sites had invasive, non invasive and ABP waveforms captured.
Study Outcomes
The primary objective was to compare the ICP curve morphology obtained with the Braincare and invasive ICP sensors, with focus on the characteristics of peaks P1, P2, P3 amplitudes and their ratios, among other characteristics of the ICP pulse waveform including lags between wave peaks (time to peak) and the absolute curvature of the peaks to determine relative changes and trends over time in ICP and brain compliance.
The analysis aimed evaluate the reliability and accuracy of the Braincare device in assessing ICP waveform in comparison to gold standard invasive ICP monitoring methods such as the external ventricular drain or intraparenchymal micro transducers as well as the ability to monitor relative changes in ICP as well as trends over time. The study hypothesis was that the ICP pulse morphology (waveform) detected by the Braincare noninvasive device presented a statistically significant correlation with the ICP pulse morphology (waveform) detected by the gold standard invasive method(s).
Bland-Altman plots and Deming regression analyses were used to quantify agreement between the invasive ICP waveform and Braincare surrogate ICP waveform parameters – estimated P2/P1 ratio and normalized time to peak (TTP), estimating the differences between the respective averages per minute. Additionally, spearman and normalized mutual information methods were utilized to assess non-linear behavior between waveforms. Considering the differences in positioning of the invasive and non-invasive sensors (inside the ventricle compared to outside the skull), strong agreement between the signals was not expected. Nevertheless, a relatively large region of agreement and presented correlation between the parameters was observed and demonstrated statistical significance confirmed by additional statistical tests.
Correlation analysis:
Spearman correlation and normalized mutual information were used to assess statistical dependence on the ICP waveform parameters between the Braincare sensor and invasive sensor. For the normalized time to peak, the Spearman correlation was 0.318 [0.291, 0.345], p<.0001. The statistical dependence between parameters - estimated using normalized mutual information – was 0.612 [0.564, 0.643]. For the estimated P2/P1 ratio, the Spearman correlation was 0.495 [0.471, 0.517], p<.0001. The statistical dependence between parameters - estimated using normalized mutual information – was 0.561 [0.531,0.606]. The joint distributions between ICP and B4C parameters showed statistical dependence between them, thus confirming the statistically significant correlation between the ICP pulse morphology (waveform) detected by the gold standard invasive method(s) and the B4C technology with regard to the P2/P1 ratio and Time to Peak.
Agreement analysis:
Bland-Altman plots and Deming regression models were used to quantify agreement between the invasive and Braincare measured ICP waveform parameters – estimated P2/P1 ratio and normalized time to peak, estimating the differences between the respective averages per minute. Considering the differences in positioning of the invasive and non-invasive sensors (inside the ventricle compared to outside the skull), strong agreement between the signals was not expected. Nevertheless, a relatively large region of agreement between the parameters was observed and demonstrate statistical significance confirmed by additional statistical tests. The Bland-Altman limits of agreement for the estimated P2/P1 ratio range from -0.723 to 0.761 (mean distance/bias 0.019; 95% CI: -0.060, 0.101). The limits of agreement for the normalized Time to Peak range from -0.183 to 0.245 (mean distance/bias 0.031; 95% CI: 0.011, 0.050). The Deming regression estimates and 95% confidence intervals for P2/P1 ratio were Y = -0.67 [-1.85, 0.02] + 1.60 [0.97, 2.69]X, and estimates and 95% confidence intervals for Time to Peak were Y = 0.00 [-0.11, 0.09] + 0.84 [0.38, 1.39]X. As can be seen when looking at the mean/bias and limits of agreement in the Bland-Altman results above, our device tends to generate larger observations for Time to Peak than invasive ICP. There are some observed instances where the Time to Peak measured by the device differed from that measured by invasive ICP by >0.2. Additionally, while the mean difference in P2/P1 between our device and invasive ICP is not significantly different from zero, we observed instances where P2/P1 measured by the device differed from that measured by invasive ICP by >0.7.
Both classes of analysis, Bland-Altman / Deming Regression which evaluate agreement, and Normalized Mutual Information / Spearman which evaluate correlation presented statistically meaningful results.
Safety:
No adverse events were reported.
Study Conclusion
Results of this study demonstrated a statistically significant correlation in the ICP signal and waveform parameters between the B4C System and the gold standard invasive ICP monitoring device measured over time. The study outcomes demonstrate comparable effectiveness between the Braincare device and commonly used invasive ICP devices for use in monitoring and assessing variations in ICP waveform associated parameters over time.