Innovation created by healthtech supported by FAPESP showed the lowest estimation error among all non-invasive methods already available in the world (image: brain4care)

Brazilian technology measures intracranial pressure more accurately and non-invasively

25 de fevereiro de 2025

By Roseli Andrion  |  FAPESP Innovative R&D – A technology developed by the Brazilian company brain4care has been shown to be able to measure absolute values of intracranial pressure (ICP) more accurately than existing non-invasive methods. This is the result of a study published in the journal npj Digital Medicine by researchers from the University of São Paulo in Brazil, the University of Cambridge in the United Kingdom, Emory University in the United States, and the company itself.

“The study included the largest number of patients and showed that the technology we developed had the lowest error in estimating the value of intracranial pressure among all the non-invasive methods already available in the world,” Gustavo Frigieri, scientific director of brain4care and one of the authors of the study, told Innovative R&D.

The technology developed by the Brazilian company, with support from FAPESP’s Innovative Research in Small Businesses Program (PIPE), consists of a sensor placed on the patient’s head that registers the nanometric expansions of the skull in each cardiac cycle and generates, in real time, a wave that indicates the variations in volume and intracranial pressure. The data obtained are processed by an artificial intelligence platform that generates reports to help doctors make decisions.

According to Frigieri, the difference in the solution developed by brain4care is the way it analyzes ICP, which has three basic components: the numerical value, the trend (up or down), and the morphology of the pulse. Instead of focusing on the numerical value, as traditional methods do, it evaluates the morphology of the pulse – that is, the way it behaves over time. This approach makes it possible to identify changes before they manifest themselves in the numerical values of the pressure.

“This understanding has been consolidated worldwide,” says Frigieri. “Invasive ICP is reactive, and brain4care is proactive because it allows us to act in advance. Still, physicians are used to the number, which comes in late, meaning that they only act when there are changes in the patient. Our technology already provides the morphology and trend of the ICP and is moving towards showing the absolute number.”

In developing the system, the company’s researchers also translated the ICP pulse into numbers. “We show the physician the patient’s situation so he can decide how to proceed. In the articles we’ve published, we’ve shown that the pulse is very important and that it does the same thing as an invasive ICP measurement, but in a non-invasive way. It allows the physician to see if the patient is getting worse or better based on the trend.”

With the system, doctors can detect neurological changes early and intervene more quickly and accurately, which can mean the difference between life and death in critical cases. “The longer a neurocritical patient goes without proper care, the greater the risk of sequelae,” he points out. “I provide the sensor and train the professional to make treatment decisions and achieve the best outcome.”

Proven effectiveness

The effectiveness of the brain4care method has been proven in over one hundred published scientific articles. The new study recruited patients from Brazil, the United States and Europe. “We accelerated the process of collecting this information and today we have a very interesting database of patients with invasive and non-invasive ICP. Using a machine learning model, we were able to estimate the value of ICP in these patients,” he explains. “We were able to achieve an error of 3 millimeters of mercury [the means by which ICP is usually measured] in more than a hundred patients. Many researchers are looking for non-invasive ways to measure ICP, but the techniques used have a very large error, which makes the system not ideal for clinical use.”

Having proven in the laboratory that it is possible to make the estimate, Frigieri continues to work on obtaining robust evidence. “I currently have 168 patients in the database and I’ve already reduced the error to 2.6 mm of mercury.” The goal is to consolidate the model later this year to demonstrate in the field that, in addition to morphology and trend, brain4care can provide the number doctors are looking for. “We’ve already shown that it’s possible to derive ICP from morphology in a completely non-invasive way. The next step is to make this a reality for the market.”

For Frigieri, any neurosurgery patient at risk of bleeding can benefit from the method. This is because when they are in an intensive care unit (ICU), sedated and ventilated, it is not possible to see their symptoms and the doctor often acts almost blindly: for example, he performs a CT scan after the operation and repeats it after 24 hours. “In these cases, you can use brain4care every two hours and, if things get worse, add a CT scan to see what happened in more detail.”

This versatility is one of the advantages of the device. It can be used in environments other than the ICU, such as medical clinics, outpatient clinics and emergency departments. This greatly expands access to ICP monitoring. Rapid diagnosis and treatment are critical in cases of head trauma, where every minute counts to preserve brain function.

According to him, brain4care is easy to use and doesn’t require complex training: “It can be adapted to the scenario. For example, if you just need to check whether the patient has intracranial hypertension, 5 to 10 minutes of use may be enough,” explains Frigieri. “A more unstable patient, on the other hand, can wear the device for an hour or two for longer monitoring.”

Portable equipment

According to the researcher, with the invasive method of ICP monitoring, it is necessary to take the patient to the operating room to place the device. Brain4care, on the other hand, can be taken anywhere. “I always say that we’re not going to replace invasive ICP, but we will be able to determine who really needs it. We want to make sure that the majority of patients are treated before they become critical, but that those who need this type of intervention can receive it safely. At the same time, this will help to optimize the use of resources.”

Currently, brain4care is present in more than 85 institutions in Brazil, from top hospitals in São Paulo to small charity hospitals in rural areas. This wide distribution demonstrates the system’s adaptability to different contexts and needs. At the same time, the company is expanding internationally, with operations in the United States since 2018 and formal marketing since last October. “We’ve gone the other way: we’re exporting medical technology around the world. We have a company in the United States that’s a subsidiary and distributor, but all the intelligence and production is Brazilian,” Frigieri says proudly.

The technology has already been approved by the National Health Surveillance Agency (ANVISA), the regulatory body for therapies, medical technologies and drugs in Brazil, and by the Food and Drug Administration (FDA), its counterpart in the United States. In addition, it has already allowed 128 scientific studies to be conducted (some completed and others still in progress). Because it is non-invasive, the method allows studies that would not be possible with traditional techniques, contributing to the advancement of knowledge in various fields of medicine. This feature has attracted researchers from different specialties, who use the data obtained by brain4care in other scientific investigations. “Previously, this information was not available for scientific use. The arrival of brain4care has opened up opportunities for the development of knowledge.”

The article “Machine learning approach for noninvasive intracranial pressure estimation using pulsatile cranial expansion waveforms” can be read at: www.nature.com/articles/s41746-025-01463-y.