Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Machine learning support to individual diagnosis of mild cognitive impairment using multimodal MRI and cognitive assessments

De Marco, M., Beltrachini, Leandro ORCID:, Biancardi, A., Frangi, A. and Venneri, A. 2017. Machine learning support to individual diagnosis of mild cognitive impairment using multimodal MRI and cognitive assessments. Alzheimer Disease and Associated Disorders 31 (4) , pp. 278-286. 10.1097/WAD.0000000000000208

[thumbnail of 00002093-201710000-00002.pdf]
PDF - Published Version
Available under License Creative Commons Attribution.

Download (398kB) | Preview


Background: Understanding whether the cognitive profile of a patient indicates mild cognitive impairment (MCI) or performance levels within normality is often a clinical challenge. The use of resting state functional MRI (RS-fMRI) and machine learning may represent valid aids in clinical settings for the identification of MCI patients. Methods: Machine-learning models were computed to test the classificatory accuracy of cognitive, volumetric (sMRI) and BOLD-connectivity (extracted from RS-fMRI) features, in single-modality and mixed classifiers. Results: The accuracy of the RS-fMRI-based classifier was not significantly different from that of the cognitive classifier (≈85%). RS-fMRI features improved the accuracy of both volumetric and cognitive indices (mixed sMRI+RS-fMRI: ≈85%; mixed sMRI+RSfMRI+ cognitive: ≈90%). The volumetric classifier had a significantly worse accuracy (≈80%). The most important cognitive features were indices of declarative memory and semantic processing. The crucial volumetric feature was the hippocampus. The RS-fMRI features selected by the algorithms were heavily based on the connectivity of mediotemporal, left temporal, and other neocortical regions. Conclusion: Feature selection was profoundly driven by non-collinearity. Some features showed no between-group differences, or showed a trend in either direction. This indicates that clinically-relevant brain alterations typical of MCI might be subtle and not inferable from group analysis.

Item Type: Article
Date Type: Publication
Status: Published
Schools: Cardiff University Brain Research Imaging Centre (CUBRIC)
Physics and Astronomy
Uncontrolled Keywords: Machine Learning, Magnetic Resonance Imaging, Semantics, Hippocampus, Resting-state
Publisher: Lippincott, Williams & Wilkins
ISSN: 0893-0341
Date of First Compliant Deposit: 27 June 2017
Date of Acceptance: 3 July 2017
Last Modified: 05 May 2023 12:42

Citation Data

Cited 18 times in Scopus. View in Scopus. Powered By Scopus® Data

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics