Machine learning approach for noninvasive intracranial pressure estimation using pulsatile cranial expansion waveforms

Q: What is the clinical evidence from this study?

Study design: Observational. Population: Acute brain injury (n=112). Intervention: Machine learning model using brain4care (B4C) System waveforms vs. Invasive ICP monitoring. Primary outcome: Mean absolute error (MAE) of estimated ICP (MAE 3.00 mmHg, 95% CI ±7.5).

Gustavo Frigieri; Sérgio Brasil; Danilo Cardim; Marek Czosnyka; Wellingson Silva Paiva; Xiao Hu

doi:10.1038/s41746-025-01463-y

January 26, 2025npj Digital MedicineOpen Access

Machine learning approach for noninvasive intracranial pressure estimation using pulsatile cranial expansion waveforms

View Full Paper

Ask AI

Bookmark

Share

Key result

A machine learning model using noninvasive pulsatile cranial expansion waveforms estimated intracranial pressure with a mean absolute error of 3.00 mmHg compared to invasive monitoring.

Figures

Why the study?

Noninvasive methods for intracranial pressure monitoring have emerged, but none has successfully replaced invasive techniques.

Does a machine learning model using waveforms from a cranial extensometer device accurately estimate intracranial pressure compared to invasive monitoring in neurocritical patients with acute brain injuries?

Study Design

Type

Observational (n=112)

Multicenter

Yes

Structured PICO

Does a machine learning model using waveforms from a cranial extensometer device accurately estimate intracranial pressure compared to invasive monitoring in neurocritical patients with acute brain injuries?

P

Population

112 neurocritical patients with acute brain injuries requiring invasive intracranial pressure monitoring, evaluated to validate a noninvasive machine learning estimation model.

E

Exposure

Machine learning model to estimate intracranial pressure using waveforms from a cranial extensometer device (brain4care [B4C] System)

C

Comparator

Invasive intracranial pressure monitoring

O

Outcome

Mean absolute error of intracranial pressure estimationsurrogate

Main Result

Effect estimate: MAE 3.00 mmHg (95% CI ±7.5)

A machine learning model using noninvasive cranial expansion waveforms can estimate intracranial pressure with a mean absolute error of 3.00 mmHg compared to invasive monitoring.

Limitations

Low diversity of ICP values in the dataset (only 5.36% of data > 20 mmHg), leading to potential overestimation in higher ranges
Susceptibility to patient agitation, which can lower signal-to-noise ratio and hinder waveform readings
Risk of signal contamination from extracranial circulation pulsations if the sensor is mispositioned
Utilization of intraparenchymal and intraventricular ICP sensors as the gold standard may introduce measurement errors
Requires further prospective validation in larger, more diverse patient populations to assess generalizability
Prospective studies are needed to validate the model's clinical utility across diverse settings

Implication

May support noninvasive ICP estimation in neurocritical care; hypothesis-generating and requires prospective validation before adoption.

Abstract

Noninvasive methods for intracranial pressure (ICP) monitoring have emerged, but none has successfully replaced invasive techniques. This observational study developed and tested a machine learning (ML) model to estimate ICP using waveforms from a cranial extensometer device (brain4care [B4C] System). The model explored multiple waveform parameters to optimize mean ICP estimation. Data from 112 neurocritical patients with acute brain injuries were used, with 92 patients randomly assigned to training and testing, and 20 reserved for independent validation. The ML model achieved a mean absolute error of 3.00 mmHg, with a 95% confidence interval within ±7.5 mmHg. Approximately 72% of estimates from the validation sample were within 0-4 mmHg of invasive ICP values. This proof-of-concept study demonstrates that noninvasive ICP estimation via the B4C System and ML is feasible. Prospective studies are needed to validate the model's clinical utility across diverse settings.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Frigieri et al. (2025) conducted an observational in Acute brain injury (n=112). Machine learning model using brain4care (B4C) System waveforms vs. Invasive ICP monitoring was evaluated on Mean absolute error (MAE) of estimated ICP (MAE 3.00 mmHg, 95% CI ±7.5). A machine learning model using noninvasive pulsatile cranial expansion waveforms estimated intracranial pressure with a mean absolute error of 3.00 mmHg compared to invasive monitoring.

synapsesocial.com/papers/6a99164fea8974fe8c27388d https://doi.org/10.1038/s41746-025-01463-y

View Full Paper

Ask AI

Bookmark

Share

Machine learning approach for noninvasive intracranial pressure estimation using pulsatile cranial expansion waveforms

Figures

Study Design

Structured PICO

Main Result

Limitations

Implication

Abstract

Discussion

Cite This Study

Machine learning approach for noninvasive intracranial pressure estimation using pulsatile cranial expansion waveforms

Figures

Study Design

Structured PICO

Main Result

Limitations

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading