Abstract #4034

Investigating the correspondence of clinical diagnostic grouping with underlying neurobiological and phenotypic clusters using unsupervised learning: An application to the Alzheimer’s spectrum

Xinyu Zhao¹, D Rangaprakash¹, D Narayana Dutt², and Gopikrishna Deshpande^1,3,4

¹Electrical and Computer Engineering, AU MRI Research Center, Auburn, AL, United States, ²Medical Electronics, Dayananda Sagar College of Engineering, Bangalore, India, ³Psychology, Auburn University, Auburn, AL, United States, ⁴Alabama Advanced Imaging Consortium, Auburn University and University of Alabama Birmingham, Auburn, AL, United States

Many brain-based disorders are traditionally diagnosed based on clinical interviews and behavioral assessments. Using Alzheimer’s spectrum (i.e. mild cognitive impairment [MCI] and Alzheimer’s disease [AD]) as a test case, we investigated whether clinical diagnostic grouping is grounded in underlying neurobiological and phenotypic clusters. In order to do so, three unsupervised learning methods were applied on resting-state fMRI connectivity measures obtained from subjects with MCI and AD. High similarity was achieved between connectivity and phenotypic clusters while similarity was low with clinical diagnosis. It shows that neurobiological and phenotypic markers could be used to improve the precision of clinical diagnosis.

This abstract and the presentation materials are available to members only; a login is required.

Join Here