An optimized ensemble search approach for classification of higher-level gait disorder using brain magnetic resonance images

• Published in: Computers in biology and medicine Vol. 184; p. 109457
• Main Authors: Mogensen, Klara, Guarrasi, Valerio, Larsson, Jenny, Hansson, William, Wåhlin, Anders, Koskinen, Lars-Owe, Malm, Jan, Eklund, Anders, Soda, Paolo, Qvarlander, Sara
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 2025; Elsevier Limited
• Subjects: Aged; Algorithms; Artificial intelligence; Artificial neural networks; Brain; Brain - diagnostic imaging; Classification; CNN; Convolutional neural networks; Cost function; Deep Learning; Ensemble learning; Female; Functional anatomy; Gait; Gait disorder; Gait Disorders, Neurologic - classification; Gait Disorders, Neurologic - diagnostic imaging; Gait Disorders, Neurologic - physiopathology; Humans; Image Interpretation, Computer-Assisted - methods; Internal Medicine; Machine learning; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical imaging; Middle Aged; Morphology; MRI; Nervous system; Neural networks; Neural Networks, Computer; Neurological disorders; Normal pressure hydrocephalus; Optimization; Other; Search algorithms; Structure-function relationships; Gait disorder; CNN; Medical imaging; MRI; Normal pressure hydrocephalus; Ensemble learning; Convolutional neural networks; Artificial intelligence; Neurological disorders; Optimization
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2024.109457

Higher-Level Gait Disorder (HLGD) is a type of gait disorder estimated to affect up to 6% of the older population. By definition, its symptoms originate from the higher-level nervous system, yet its association with brain morphology remains unclear. This study hypothesizes that there are patterns in brain morphology linked to HLGD. For the first time in the literature, this work investigates whether deep learning, in the form of convolutional neural networks, can capture patterns in magnetic resonance images to identify individuals affected by HLGD. To handle this new classification task, we propose setting up an ensemble of models. This leverages the benefits of combining classifiers instead of determining which network is the most suitable, developing a new architecture, or customizing an existing one. We introduce a computationally cost-effective search algorithm to find the optimal ensemble by leveraging a cost function of both traditional performance scores and the diversity among the models. Using a unique dataset from a large population-based cohort (VESPR), the ensemble identified by our algorithm demonstrated superior performance compared to single networks, other ensemble fusion techniques, and the best linear radiological measure. This emphasizes the importance of implementing diversity into the cost function. Furthermore, the results indicate significant morphological differences in brain structure between HLGD-affected individuals and controls, motivating research about which areas the networks base their classifications on, to get a better understanding of the pathophysiology of HLGD. [Display omitted] •Classification of 3D brain MRI to distinguish gait impairment (HLDG) from controls.•Efficient algorithm to search for optimal ensemble of neural network classifiers.•Use of multiple evaluation and diversity metrics in ensemble search cost function.•The optimal ensemble outperforms all single networks and ensemble methodologies.•Results show significant relationship between brain structure and gait impairment.