Biography

As a physicist in a multidisciplinary environment, I enjoy creating mathematical models to understand the relationships between human brain and behavior in health and disease. Starting as postdoc in the UMCU neuroimaging group, I set up an image-processing pipeline for quantitative analysis of thousands of MRI brain images. As assistant professor, I implemented advanced statistical analyses to study dynamic changes in brain morphology. Central in my research have always been how (image) data represents information and how knowledge of data quality can be used to perform optimal analyses. Applied to multicenter imaging studies, my work resulted in a method to determine reliability of, for example, (twin) heritability or longitudinal studies. At that time the first machine learning steps in this field were made and I realized that multivariate modeling much better uses all the information available in brain images. My expertise on reliability and individual variation enabled me to shift my focus, from group-level analyses to making predictions about individuals, based on their data. My team performed the first large-scale study to classify persons with and without schizophrenia based on MRI brain images. In recent years, we have developed and applied machine learning methods to further investigate the heterogeneity of brain diseases. For multicenter designs, we recently have developed meta-modeling. I have expanded the use of pattern recognition analyses to applications in other domains, including clinical data and vocabulary data. The latter work is in collaboration with the Faculty of Humanities, Dept. of Languages, within the strategic theme Dynamics of Youth. Since 2016 I have been appointed part-time Assistant professor in that Department. In the coming years, I plan to do research on models to predict people’s development of (psychiatric) disorders in the first twenty years of their lives. This requires further development of machine learning in multicenter, multimodal and cross-diagnostic settings. Since the development of such models will be a multi-disciplinary effort, it is essential to introduce the next generation of researchers to this field in my lectures on topics with integrated state-of-the-art machine learning, to engage them in future innovation.

Research line

Patterns in Psychiatry  https://www.neuromri.nl/patterns-in-psychiatry-pip-pip-lab/

Most recent key publications

1. Dluhoš P, Schwarz D, Cahn W, van Haren N, Kahn R, Španiel F, Horáček J, Kašpárek T, Schnack H. Multi-center machine learning in imaging psychiatry: A meta-model approach. Neuroimage. 2017 Apr 17;155:10-24.

2. Schnack HG, van Haren NE, Nieuwenhuis M, Hulshoff Pol HE, Cahn W, Kahn RS. Accelerated Brain Aging in Schizophrenia: A Longitudinal Pattern Recognition Study. Am J Psychiatry. 2016 Jun 1;173(6):607-16.

3. Schnack, H. G. and Kahn, R. S. (2016). "Detecting neuroimaging biomarkers for psychiatric disorders: sample size matters." Frontiers in Psychiatry 7: 50.

4. Schnack HG, van Haren NE, Brouwer RM, Evans A, Durston S, Boomsma DI, Kahn RS, Hulshoff Pol HE, Changes in thickness and surface area of the human cortex and their relationship with intelligence. Cereb Cortex 25:1608-1617 (2015).

5. Schnack HG, Nieuwenhuis M, van Haren NE, Abramovic L, Scheewe TW, Brouwer RM, Hulshoff Pol HE, Kahn RS, Can structural MRI aid in clinical classification? A machine learning study in two independent samples of patients with schizophrenia, bipolar disorder and healthy subjects. Neuroimage 84:299-306 (2014)

6. Nieuwenhuis M, van Haren NE, Hulshoff Pol HE, Cahn W, Kahn RS, Schnack HG, Classification of schizophrenia patients and healthy controls from structural MRI scans in two large independent samples. NeuroImage 61:606-612 (2012).

7. Brouwer RM, Hulshoff Pol HE, Schnack HG, Segmentation of MRI Brain Scans Using Non-Uniform Partial Volume Densities. NeuroImage 49:467-477 (2010).