New research presents the simulation of a system that predicts the onset of a migraine attack with a remarkably high degree of accuracy. Being able to foresee a migraine can help to relieve or even prevent the pain altogether.
While migraines do tend to
The World Health Organization (WHO)
Although drugs help to relieve the pain, the fact that people have to wait until they feel the first signs of a migraine often reduces the effectiveness of the painkillers.
But now, a team of Spanish-based researchers have simulated a migraine prediction system that could soon improve the lives of people affected by migraines.
The scientists have now carried out a simulation using real patient data, presenting a real scenario for the prediction of migraines. The findings were published in the journal Future Generation Computer Systems.
Knowing when an attack will strike can drastically improve the effectiveness of the painkillers, the researchers explain, helping patients to stop the pain.
Pagán Ortiz spoke to Medical News Today about the novelty of his recent research, highlighting the fact that while the device has not yet been tested, it has been simulated in a real scenario.
“In our previous research we developed a prototype to gather hemodynamic data information from migraine sufferers in an ambulatory scenario,” he explained.
Hemodynamic data refers to the variables that “announce” the onset of a migraine attack, and they include surface skin temperature, the skin’s electrical properties, heart rate, and the oxygen saturation of the capillaries.
“We created personalized predictive models offline, in our servers and computers, and we wanted to test them in real time in order to generate alarms and let the patients know, in advance, when the migraine pain is [going to] start.”
“This experiment is a time-consuming and expensive task, thus we decided to simulate it first,” Pagán Ortiz told us, adding further details about the procedure.
“What we simulated is the behavior of our current prototype in a real scenario,” he continued. “We simulated the conditions that real ambulatory monitoring devices suffer: sensor disconnection, noise, etc., and we studied how this can affect the reliability of the migraine prediction.”
“[W]e showed techniques to mitigate this effect and keep accurate predictions,” Pagán Ortiz added. “On the other hand, we studied how this system could alert patients. We saw the predictive models’ behavior and how the simulated monitoring device generates alerts.”
Pagán Ortiz said, “The simulation showed that it is suitable to bring to the real world all the methodology we’ve developed to predict migraines and generate alarms in real time to warn patients in advance.”
More specifically, the “average rate of prediction success” of the system is 76 percent, and the average period of time before the onset is 25 minutes — which is enough time to intervene so as to diminish the pain or prevent it altogether.
Pagán Ortiz went on, “Predictions are made in the time window where the drug is effective according to the pharmacokinetics of the drugs. With this [system], we estimate [that] we will be able to predict, and thus avoid, around 75 [percent of the] crises.”
“This will reduce the visits to emergenc[y rooms],” he stated, adding, “[I]t will allow [for] personalized medicine (drugs [for] the acute phase of the pain), and it will improve their [quality] of life. It will also reduce considerably the bill of national and private health systems and health insurances due to economic savings in direct and indirect costs.”
In their paper, Pagán Ortiz and his colleagues explain the benefits of their research, saying, “Predicting the onset of a migraine will considerably reduce the patient’s pain and thus the effects of migraine over the course of their lives. This will also lead to considerable economic savings over time.”
“Migraine patients cannot go to work or do their normal life (social activities, etc). If we are able to tell them in advance when they are going to suffer the pain, it will change their lives.”
Josué Pagán Ortiz
Pagán Ortiz also shared with MNT some of the researchers’ directions for future research.
“[The] next natural step would be the implementation and test[ing] of a real device with real patients. Working with real patients to suggest changes in the way they treat their disease is a serious issue to deal with, and that requires a lot of legal consents which we are working on now.”
These findings were the result of the collaboration of the Universidad Complutense de Madrid and Universidad Politécnica de Madrid, working together with the Center for Computational Simulation and the Neurology Group of the Hospital Universitario de la Princesa de Madrid.