Computer-aided differential diagnosis system for Alzheimer’s disease based on machine learning with functional and morphological image features in magnetic resonance imaging

Increasing evidence suggests the utility of magnetic resonance imaging (MRI) as an important technique for the diagnosis of Alzheimer’s disease (AD) and for predicting the onset of this neurodegenerative disorder. In this study, we present a sophisticated machine learning (ML) model of great accuracy to diagnose the early stages of AD. A total of 373 MRI tests belonging to 150 subjects (age ≥ 60) were examined and analyzed in parallel with fourteen distinct features related to standard AD diagnosis. Four ML models, such as naive Bayes (NB), artificial neural networks (ANN), K-nearest neighbor (KNN), and support-vector machines (SVM), and the receiver operating characteristic (ROC) curve metric were used to validate the model performance. Each model evaluation was done in three independent experiments. In the first experiment, a manual feature selection was used for model training, and ANN generated the highest accuracy in terms of ROC (0.812). In the second experiment, automatic feature selection was conducted by wrapping methods, and the NB achieved the highest ROC of 0.942. The last experiment consisted of an ensemble or hybrid modeling developed to combine the four models. This approach resulted in an improved accuracy ROC of 0.991. We conclude that the involvement of ensemble modeling, coupled with selective features, can predict with better accuracy the development of AD at an early stage.

Download Full-text

Machine learning techniques for computer-aided classification of active inflammatory sacroiliitis in magnetic resonance imaging

Advances in Rheumatology ◽

10.1186/s42358-020-00126-8 ◽

2020 ◽

Vol 60 (1) ◽

Cited By ~ 1

Author(s):

Matheus Calil Faleiros ◽

Marcello Henrique Nogueira-Barbosa ◽

Vitor Faeda Dalto ◽

José Raniery Ferreira Júnior ◽

Ariane Priscilla Magalhães Tenório ◽

...

Keyword(s):

Magnetic Resonance Imaging ◽

Machine Learning ◽

Magnetic Resonance ◽

Machine Learning Techniques ◽

Resonance Imaging ◽

Learning Techniques ◽

Computer Aided

Download Full-text

Magnetic Resonance Imaging Measurement of Entorhinal Cortex in the Diagnosis and Differential Diagnosis of Mild Cognitive Impairment and Alzheimer’s Disease

Brain Sciences ◽

10.3390/brainsci11091129 ◽

2021 ◽

Vol 11 (9) ◽

pp. 1129

Author(s):

Qianqian Li ◽

Junkai Wang ◽

Jianghong Liu ◽

Yumeng Wang ◽

Kuncheng Li

Keyword(s):

Magnetic Resonance Imaging ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Differential Diagnosis ◽

Magnetic Resonance ◽

Surface Area ◽

Entorhinal Cortex ◽

Resonance Imaging ◽

Significant Difference ◽

Area And Volume

Several magnetic resonance imaging studies have shown that the entorhinal cortex (ERC) is the first brain area related to pathologic changes in Alzheimer’s disease (AD), even before atrophy of the hippocampus (HP). However, change in ERC morphology (thickness, surface area and volume) in the progression from aMCI to AD, especially in the subtypes of aMCI (single-domain and multiple-domain: aMCI-s and aMCI-m), however, is still unclear. ERC thickness, surface area and volume were measured in 29 people with aMCI-s, 22 people with aMCI-m, 18 patients with AD and 26 age-/sex-matched healthy controls. Group comparisons of the ERC geometry measurements (including thickness, volume and surface area) were performed using analyses of covariance (ANCOVA). Furthermore, receiver operator characteristic (ROC) analyses and the area under the curve (AUC) were employed to investigate classification ability (HC, aMCI-s, aMCI-m and AD from each other). There was a significant decreasing tendency in ERC thickness from HC to aMCI-s to aMCI-m to finally AD in both the left and the right hemispheres (left hemisphere: HC > aMCI-s > AD; right hemisphere: aMCI-s > aMCI-m > AD). For ERC volume, both the AD group and the aMCI-m group showed significantly decreased volume on both sides compared with the HC group. In addition, the AD group also had significantly decreased volume on both sides compared with the aMCI-s group. As for the ERC surface area, no significant difference was identified among the four groups. Furthermore, the AUC results demonstrate that combined ERC parameters (thickness and volume) can better discriminate the four groups from each other than ERC thickness alone. Finally, and most importantly, relative to HP volume, the capacity of combined ERC parameters was better at discriminating between HC and aMCI-s, as well as aMCI-m and AD. ERC atrophy, particularly the combination of ERC thickness and volume, might be regarded as a promising candidate biomarker in the diagnosis and differential diagnosis of aMCI and AD.

Download Full-text