Prevalence of malformations of cortical development in patients with suspected epilepsy based on a clinical MRI dataset

Epilepsy is a condition where the brain's electrical signals fire in a way that causes sudden, unprovoked seizures. While the condition affects millions of people worldwide, the reasons behind it are not always clear. Sometimes, the cause is a structural problem in the brain that has been there since before birth. These problems are called malformations of cortical development. Imagine the brain's surface as a complex landscape of folds and ridges; in these cases, that landscape is built incorrectly. Some areas might be too smooth, others might have too many tiny folds, or parts of the brain might be in the wrong place. These structural errors are a known cause of epilepsy, but for a long time, scientists have struggled to count exactly how many people in the general population live with them. Most previous studies looked only at children or at patients who had already undergone brain surgery, leaving a gap in our understanding of how common these conditions are across all ages in the wider community.

To fill this gap, a team of researchers in Denmark turned to the country's extensive digital health records. They focused on a large region in the east of the country, looking at nearly three million people living there. Their goal was to find out how many people had suspected epilepsy and, among those people, how many had these specific brain malformations visible on brain scans. Because manually reading through thousands of medical reports would take years, the team used a sophisticated computer program based on language models. This tool was trained to read the text of radiology reports—written by doctors after they look at brain scans—and identify mentions of specific brain abnormalities. The researchers checked the computer's work against human reviews to ensure it was accurate, then used it to scan the records of everyone who had a brain MRI.

The study found that on a single day in July 2023, about 1,045 people out of every 100,000 in the region had a record suggesting they had epilepsy. This number matched what other national studies had found, giving the researchers confidence that their data was reliable. When they looked specifically at the people who had also had a brain MRI, they found that about 32 out of every 100,000 people in the total population had a brain malformation associated with seizures. The most common type of malformation found was focal cortical dysplasia, which is a small area where the brain cells are disorganized. This specific condition appeared in about 15 people per 100,000. Other conditions, such as heterotopia, where brain cells are found in the wrong place, and polymicrogyria, where the brain has too many small folds, were also identified, though they were less common.

The researchers were careful to note that their numbers might be lower than the true reality. This is because not everyone with suspected epilepsy gets a brain scan, and the computer program, while helpful, is not perfect. For instance, the tool was very good at spotting focal cortical dysplasia but was not manually checked for every single case of heterotopia because there were so many. Despite these limitations, the study provided the first broad look at these conditions across all ages in a large population. It showed that while these brain malformations are rare, they are a significant part of the picture for people with epilepsy. The work also demonstrated that using advanced computer tools to sift through vast amounts of medical data is a viable way to study rare diseases, offering a new path for understanding conditions that have historically been difficult to count.