Comparison of Feature Selection Techniques in Machine Learning for Anatomical Brain MRI in Dementia

Topic: Risk, prevention and care in population studies

We present a comparative split-half resampling analysis of various data driven feature selection and classification methods for the whole brain voxel-based classification analysis of anatomical magnetic resonance images. We compared support vector machines (SVMs), with or without filter based feature selection, several embedded feature selection methods and stability selection. While comparisons of the accuracy of various classification methods have been reported previously, the variability of the out-of-training sample classification accuracy and the set of selected features due to independent training and test sets have not been previously addressed in a brain imaging context. We studied two classification problems: 1) Alzheimer's disease (AD) vs. normal control (NC) and 2) mild cognitive impairment (MCI) vs. NC classification. In AD vs. NC classification, the variability in the test accuracy due to the subject sample did not vary between different methods and exceeded the variability due to different classifiers. In MCI vs. NC classification, particularly with a large training set, embedded feature selection methods outperformed SVM-based ones with the difference in the test accuracy exceeding the test accuracy variability due to the subject sample. The filter and embedded methods produced divergent feature patterns for MCI vs. NC classification that suggests the utility of the embedded feature selection for this problem when linked with the good generalization performance. The stability of the feature sets was strongly correlated with the number of features selected, weakly correlated with the stability of classification accuracy, and uncorrelated with the average classification accuracy.

Key findings

• NC classification, the variability in the test accuracy due to the subject sample did not vary between different methods and exceeded the variability due to different classifiers.
• NC classification, particularly with a large training set, embedded feature selection methods outperformed SVM-based ones with the difference in the test accuracy exceeding the test accuracy variability due to the subject sample.
• The filter and embedded methods produced divergent feature patterns for MCI vs.
• NC classification that suggests the utility of the embedded feature selection for this problem when linked with the good generalization performance.
• The stability of the feature sets was strongly correlated with the number of features selected, weakly correlated with the stability of classification accuracy, and uncorrelated with the average classification accuracy.

Why it matters

This study compared different machine learning feature selection methods for classifying brain MRI scans of people with Alzheimer's disease, mild cognitive impairment and normal controls. It is a methods comparison for diagnosis, not a treatment, cure or prevention study.

New to this subject? Start with our plain-English guide: Digital health →

Study details

• Evidence strengthLow

Read the original

• DOI: 10.1007/s12021-015-9292-3
• PubMed: 26803769
• Original source →

May need a subscription OpenAlex found no free copy of the full text (checked 6 Feb 1020), so the publisher may charge for it. The PubMed abstract is free.

@article{tohka2016-26803769,
  author  = {Tohka, J. and Moradi, E. and Huttunen, H. and {Alzheimer’s Disease Neuroimaging Initiative}},
  title   = {{Comparison of Feature Selection Techniques in Machine Learning for Anatomical Brain MRI in Dementia}},
  journal = {Neuroinformatics},
  year    = {2016},
  doi     = {10.1007/s12021-015-9292-3},
  pmid    = {26803769},
  url     = {https://doi.org/10.1007/s12021-015-9292-3}
}

Studies this one cites

Referenced by this study and also in the alzhub library, so you can read them here.

• Alzheimer's Disease Neuroimaging Initiative (ADNI): clinical characterization Neurology · 2009-12-30 · usefulness score 68/100 · Mild cognitive impairment
• BrainAGE in Mild Cognitive Impaired Patients: Predicting the Conversion to Alzheimer's Disease PloS one · 2013-06-27 · usefulness score 57/100 · Risk factors
• High dimensional classification of structural MRI Alzheimer's disease data based on large scale regularization Frontiers in neuroinformatics · 2011-10-14 · usefulness score 47/100 · Brain scans and imaging
• Sparse learning and stability selection for predicting MCI to AD conversion using baseline ADNI data BMC neurology · 2012-06-25 · usefulness score 44/100 · Long-term cohort studies

Later studies that cite this one

3 studies in the alzhub library list this one in their references.

• Cortical Thickness from MRI to Predict Conversion from Mild Cognitive Impairment to Dementia in Parkinson Disease: A Machine Learning-based Model Radiology · 2021-05-25 · usefulness score 46/100 · Digital health tools
• Rey's Auditory Verbal Learning Test scores can be predicted from whole brain MRI in Alzheimer's disease NeuroImage. Clinical · 2016-12-18 · usefulness score 45/100 · Digital health tools
• Neuroimaging and Machine Learning for Dementia Diagnosis: Recent Advancements and Future Prospects IEEE reviews in biomedical engineering · 2018-12-11 · usefulness score 44/100 · Digital health tools

Related studies in Digital health tools

• Does feature selection improve classification accuracy? Impact of sample size and feature selection on classification using anatomical magnetic resonance images NeuroImage · 2011-12-01 · usefulness score 36/100
• Brain MRI Analysis for Alzheimer's Disease Diagnosis Using CNN-Based Feature Extraction and Machine Learning Sensors (Basel, Switzerland) · 2022-04-11 · usefulness score 51/100
• Enhanced in silico QSAR-based screening of butyrylcholinesterase inhibitors using multi-feature selection and machine learning SAR and QSAR in environmental research · 2025-02-01 · usefulness score 34/100
• Exploring the Effectiveness of Machine Learning and Deep Learning Techniques for EEG Signal Classification in Neurological Disorders IEEE Access · 2025-01-01 · usefulness score 51/100
• Health management and pattern analysis of daily living activities of people with dementia using in-home sensors and machine learning techniques PloS one · 2018-05-03 · usefulness score 47/100
• Machine-learning techniques for building a diagnostic model for very mild dementia NeuroImage · 2010-04-09 · usefulness score 44/100

Also tagged “classification”

835 studies in the library carry the topic tag “classification”; these are this study's neighbours in a fixed, alphabetical-by-hash tour of them, not the closest in content. All 835 tagged studies.

• Future of Alzheimer's detection: Advancing diagnostic accuracy through the integration of qEEG and artificial intelligence NeuroImage · 2025-07-16 · usefulness score 49/100 · Getting a diagnosis
• Change in rates of cerebral atrophy over time in early-onset Alzheimer's disease: longitudinal MRI study Lancet (London, England) · 2003-10-01 · usefulness score 57/100 · Cognitive tests and assessment
• Diffusion MRI Indices and Their Relation to Cognitive Impairment in Brain Aging: The Updated Multi-protocol Approach in ADNI3 Frontiers in neuroinformatics · 2019-02-19 · usefulness score 51/100 · Cognitive tests and assessment

Similar studies

• Multiple instance learning for classification of dementia in brain MRI 2014 · Digital health tools
• Using Machine Learning to Predict Dementia from Neuropsychiatric Symptom and Neuroimaging Data 2020 · Digital health tools
• Machine learning of neuroimaging for assisted diagnosis of cognitive impairment and dementia: A systematic review 2018 · Digital health tools
• Alzheimers and Brain Tumor Detection Using Deep Learning 2025 · Brain scans and imaging
• Comparison of Methods for Algorithmic Classification of Dementia Status in the Health and Retirement Study 2019 · Disparities
• A comparison of machine learning methods for survival analysis of high-dimensional clinical data for dementia prediction 2020 · Digital health tools

More on Digital health tools

• A review that covers this ground Novel MRI techniques in the assessment of dementia European journal of nuclear medicine and molecular imaging · 2008-03-01 · usefulness score 41/100
• Newest in this subject Speaker-Wise Dementia Screening fromNaturalistic Speech with Lexical Entropy, Semantic Drift, and Multimodal Fusion Research Square · 2026-10-21 · Preprint — not yet peer-reviewed · usefulness score 16/100
• Newest in this subject Integrating Multisource Target Evidence with Machine-Learning Potency Prediction for Alzheimer’s Disease Drug Discovery bioRxiv · 2026-10-20 · Preprint — not yet peer-reviewed · usefulness score 16/100

See all studies in Digital health tools →