A new study shows that a deep learning model that accurately predict patient response to antiseizure medications (ASMs) in patients with newly diagnosed epilepsy. The findings were reported in JAMA Neurology.
Choosing the optimal ASM remains in large part a trial-and-error venture for epilepsy patients, as the researchers noted. They said that using this approach “many patients have to endure sequential trials of ineffective treatments until the “right drugs” are prescribed.”
Therefore, investigators in this study sought to construct and validate a deep learning algorithm to discern treatment success of ASMs. In order to do so, they analyzed a cohort of almost 1,800 adults (over half female, median age, 34) treated between 1982 and 2020. The population of interest were followed for at least 1 year or until ASM treatment failure. Treatment success was analyzed using the area under the receiver operating characteristic curve (AUROC), weighted balanced accuracy, sensitivity, and specificity. The main endpoint of interest was complete seizure freedom for the first of year of treatment while taking the first ASM.
According to the results, the machine learning model had an AUROC of 0.65 (95% CI, 0.63-0.67) and a weighted balanced accuracy of 0.62 (95% CI, 0.60-0.64). The model was able to predict the number of pretreatment seizures, presence of psychiatric disorders, electroencephalography, and brain images, and outperformed 2 of the 5 other models based on AUROC.
“In this cohort study, a deep learning model showed the feasibility of personalized prediction of response to ASMs based on clinical information. With improvement of performance, such as by incorporating genetic and imaging data, this model may potentially assist clinicians in selecting the right drug at the first trial,” the researchers concluded.
Source: docwirenews.com, Rob Dillard
