Juergen Hahn, Rensselaer Polytechnic Institute – Using Big Data To Evaluate Autism Treatments

On Rensselaer Polytechnic Institute Week:  We need a new way to diagnose autism.

Juergen Hahn, professor of biomedical engineering, looks to the blood to find a solution.

Juergen Hahn is the department head of the Department of Biomedical Engineering at Rensselaer Polytechnic Institute in addition to holding an appointment in the Department of Chemical & Biological Engineering. He received his Diploma degree in engineering from RWTH Aachen, Germany, in 1997, and his MS and Ph.D. degrees in chemical engineering from the University of Texas, Austin, in 1998 and 2002, respectively. He was a post-doctoral researcher at the Chair for Process Systems Engineering at RWTH Aachen, Germany, before joining the Department of Chemical Engineering at Texas A&M University, College Station, in 2003 and moving to the Rensselaer Polytechnic Institute in 2012. His research interests include systems biology and process modeling and analysis with over 130 peer-reviewed publications in print. Dr. Hahn is a recipient of a Fulbright scholarship (1995/96), received the Best Referee Award for 2004 from the Journal of Process Control, the CPC 7 Outstanding Contributed Paper Award in 2006, was named Outstanding Reviewer by the journal Automatica in 2005, 2006, 2007, and 2010 CAST Outstanding Young Researcher, and has been elected as an AIMBE fellow in 2013. He served on the IEEE CSS Board of Governors in 2016 and has been a CACHE Trustee since 2014. He is currently serving as associate editor for the journals Control Engineering Practice, Processes, and the Journal of Process Control.

Using Big Data To Evaluate Autism Treatments

Autism is estimated to affect 1 in 59 children in the US.  While we don’t know the underlying cause of autism, most recent work points towards the interaction of a genetic pre-disposition with environmental factors. Because the underlying cause is not well understood, autism and the severity of autism-related symptoms are diagnosed through observations.

As of now, no medical interventions exist that treat the core symptoms of autism. So there is a tremendous need for the development of medical treatments. However, assessing the effectiveness of medical treatments based only upon observations is challenging for young children, especially if many of them are non-verbal.

We are trying to address those challenges. Our work focuses on analysis of metabolite concentrations found in the blood. While there is no one particular component that is an indicator for autism, we found that the use of a Big Data approach to analyze combinations of measurements can predict if a child has autism. Furthermore, these measurements correlate with certain autism-related behavioral measures.

We have applied this method to assess the effectiveness of treatments by measuring metabolites before and after treatment. We found that effective treatments resulted in metabolite concentrations that are closer to the values found in children without autism, but also that the changes in the metabolites correlate with improvements in autism-related behaviors. We have used this approach on data from three different experimental medical treatments and included a placebo group. We found a good correlation between our prediction of the symptom severity and the ones assessed by observations. We are hopeful that this type of approach can be used as a companion diagnostic for clinical trials which has the potential to speed up drug development.