October 12, 2026
Daniel Ackom has a passion for translational medicine, and his work is getting attention. As a PhD student in the Joint Department of Biomedical Engineering, Daniel is looking to better predict medication responses in children with non-lesional epilepsy. This process begins by finding ways to better characterize the disease. For this, Daniel is using a novel combination of fixel-based analysis, an advanced diffusion MRI method for characterizing white matter, and machine learning. In translation, these findings may help better characterize non-lesional pediatric epilepsy, creating a framework for identifying meaningful imaging patterns and informing more strategic treatment approaches for pediatric epilepsy patients.
Daniel’s work is being recognized this December at the 2026 American Epilepsy Society (AES) Annual Meeting, where he will be recognized as one of 20 winners of the Young Investigator Award. His abstract was selected from more than 1,500 submissions for this honor. The same work was also selected for a platform presentation in the meeting’s Surgery/Neuroimaging session.
BME caught up with Daniel to learn more about his research, his motivation, and his recent award.
The study is titled “Fixel-Based White Matter and Machine-Learning Signatures of Non-lesional Pediatric Epilepsy.” focuses on children with epilepsy whose conventional clinical MRI scans show no identifiable structural lesion that could explain their epilepsy. Although these scans appear structurally normal, subtle alterations in the brain’s white matter may still be present. We therefore investigated whether advanced diffusion imaging could detect these changes and whether distinct patterns of white matter abnormalities could be identified across different pediatric epilepsy syndromes.
We analyzed diffusion MRI data from nearly 500 children, including 292 with non-lesional epilepsy and 205 matched controls. Using an advanced technique called fixel-based analysis, we examined individual white matter fiber populations across 71 brain tracts to uncover subtle differences that are not visible on conventional clinical MRI. We then used machine learning to explore whether these white matter patterns could distinguish between pediatric epilepsy syndromes and provide insights into how patients respond to anti-seizure medications.
We found distinct patterns of white matter alterations across different pediatric epilepsy syndromes, along with some abnormalities that were shared across syndromes. We also found that earlier seizure onset was associated with greater alterations in white matter organization. The imaging-derived signatures distinguished several epilepsy groups from controls with high classification performance, while combining imaging and clinical information provided additional information for treatment-related prediction.
I'm drawn to translational neuroscience, work that sits between engineering and the clinic and can eventually change how care is delivered. Pediatric epilepsy is especially compelling to me because many children have recurrent seizures despite a clinical MRI that looks normal, and because medication response is still often determined largely through trial and error. I want to use diffusion MRI, structural brain networks, and machine learning, including explainable AI approaches that help clinicians understand which features are driving a prediction, to investigate whether we can identify children at greater risk of not responding to standard medications earlier in their care. Ultimately, the goal is to develop and validate imaging biomarkers that could help clinicians recognize children at risk for drug-resistant epilepsy sooner, potentially allowing earlier consideration of more definitive treatment options.
Absolutely. My training in the Joint Department of Biomedical Engineering has been central to this work. I'm mentored by Dr. Scott A. Beardsley in the Integrative Neural Systems Laboratory and co-mentored by Dr. Pradeep Javarayee in the Department of Neurology at MCW. I've also benefited greatly from the broader collaborative environment at Marquette and MCW, particularly my work with Dr. Brian Schmit in the Integrative Neural Engineering and Rehabilitation Laboratory. Their complementary expertise spanning biomedical engineering, neuroimaging, and clinical pediatric epilepsy has shaped how I approach pediatric epilepsy from both quantitative and translational perspectives. Over the past several years, that mentorship has allowed me to build expertise across multiple areas of pediatric neurology beyond epilepsy, including neonatal neurology and NMDA receptor encephalitis. Together, these collaborations have brought expertise in biomedical engineering, neuroimaging, machine learning, and clinical epilepsy to this project. That interdisciplinary environment has been an important part of my development as a researcher and has allowed me to approach clinically relevant questions using engineering and computational tools.
I’m grateful to the AES for this recognition and to my mentors, collaborators, the MU-MCW BME community, Medical College of Wisconsin, and Children’s Wisconsin for providing a collaborative environment where engineering, neuroscience, and clinical care come together, and for the opportunities and resources that have made this research possible. Ultimately, I hope these collaborations will help move our findings closer to tools that can meaningfully inform clinical care for children with epilepsy.
Have questions regarding the Marquette University and Medical College of Wisconsin Joint Department of Biomedical Engineering?
To report a problem with this website, contact BME Communications or report an accessibility issue.
