Statistics are used in a wide range of different fields today, and they have proven to be valuable in the medical field time and again. By using a statistical approach, researchers are finding that it is possible to get more information in more detail with their method of measuring brain signals in epileptic seizures. The added information can help to illuminate how those signals originate and how they spread.
A Better Method of Studying the EEG
In the past, doctors would inspect EEG recordings of epilepsy patients before and during a seizure visually. They found that visual inspections could give them a good idea of which part of the brain would benefit the most from undergoing surgical treatment. However, that is often not enough to pinpoint a problem in some of the more difficult cases that doctors see.
Through the use of a new approach from KAUST biostatistician Hernando Ombao and the team at Yuan Wang of the University of Wisconsin-Madison and Moo K. Chung of the University of California Irvine, it is possible to find out more. With the new approach, it is possible for a deeper dive into the features of the EEG and to find abnormalities in the brain even before a seizure occurs.
The researchers believe that the use of this method could make it much easier for seizure localization. This could help doctors at the clinics to make better choices in regard to treatment for the patients.
The approach makes use of math as a means to better analyze the datasets, which tend to be quite complex. The researchers use the method to study and analyze shape representations of data and their interactions. Through analyzing the shapes, it can provide information on patterns that are in the data that might not have otherwise been discovered. The researchers’ method is called topological data analysis framework. They used this in order to see just what they would be able to learn from an EEG recording that was “conducted before and during an epileptic seizure.”
By using the statistical method, they found that they were able to remove the noise that is typically in the EEG recording. This allows them to have cleaner signals to use. Shapes are then drawn that relate to the signals in the recording. These shapes, called persistence landscapes, represent signals that come from each of the electrodes that have been placed on the scalp. They are able to provide the researchers with a much better understanding of where the seizure begins in the brain and how it spreads.
More Research Needed
With additional simulation studies, the researchers were able to find that the test remained very sensitive and provided a substantial amount of information. This was true even when there was a lot of noise within the signal. The researchers believe that by adding this method to their other tools, they will be able to get a better understanding of their patients, including those who have difficult cases. This can help them to determine not only the origin site of the seizure, but it can help to find a better way to treat those patients.
The researchers are continuing to learn more about their method. They plan to test using larger samples of EEG recordings in order to help them validate their findings thus far. Statistical methods of study can help many different areas of medicine and the researchers believe that it is possible to use these methods to study not only epilepsy, but other shocks to the brain.