New research from the Developmental Synaptopathies Consortium (DSC). This summary is based on a paper published in the journal Epilepsia on November 19, 2025, titled "Convolutional neural networks for automatic tuber segmentation and quantification of tuber burden in tuberous sclerosis complex." 

Read the paper here. 

Learn more about DSC. 

Transcript: 

Tuberous sclerosis complex (TSC) is a genetic disorder that leads to the growth of non-cancerous tumors in multiple organs. In the brain, these include “tubers,” areas of abnormal tissue just beneath the cortical surface that can cause seizures and disrupt normal brain function. Magnetic resonance imaging (MRI) is commonly used to identify how many tubers are present, how large they are, and where they are located. These features often relate to the type and severity of a person’s neurological symptoms.

In this study, researchers created a fully automated neural network (an artificial intelligence-based program) to detect tubers on MRI and measure their total volume. They trained the model using 263 brain MRI scans from 196 individuals with TSC. They then compared the algorithm’s performance with measurements made by an expert neuroradiologist.

The algorithm’s estimates of total tuber load showed an almost perfect match with the expert standard. The authors conclude that this tool provides an objective and consistent way to identify and measure tubers, which may improve the reliability of TSC research across different sites.