Automated detection of brain atrophy patterns based on MRI for the prediction of Alzheimer's disease
Related Experiment Video
Updated: Jun 18, 2026
Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly
Published on: April 14, 2014
Claudia Plant1, Stefan J Teipel, Annahita Oswald
1Department of Neuroradiology, Technische Universität München, Munich, Germany.
Abstract:
Subjects with mild cognitive impairment (MCI) have an increased risk to develop Alzheimer's disease (AD). Voxel-based MRI studies have demonstrated that widely distributed cortical and subcortical brain areas show atrophic changes in MCI, preceding the onset of AD-type dementia. Here we developed a novel data mining framework in combination with three different classifiers including support vector machine (SVM), Bayes statistics, and voting feature intervals (VFI) to derive a quantitative index of pattern matching for the prediction of the conversion from MCI to AD. MRI was collected in 32 AD patients, 24 MCI subjects and 18 healthy controls (HC). Nine out of 24 MCI subjects converted to AD after an average follow-up interval of 2.5 years. Using feature selection algorithms, brain regions showing the highest accuracy for the discrimination between AD and HC were identified, reaching a classification accuracy of up to 92%. The extracted AD clusters were used as a search region to extract those brain areas that are predictive of conversion to AD within MCI subjects. The most predictive brain areas included the anterior cingulate gyrus and orbitofrontal cortex. The best prediction accuracy, which was cross-validated via train-and-test, was 75% for the prediction of the conversion from MCI to AD. The present results suggest that novel multivariate methods of pattern matching reach a clinically relevant accuracy for the a priori prediction of the progression from MCI to AD.
Insights
Predicting Alzheimer's disease (AD) progression in mild cognitive impairment (MCI) is crucial. Novel data mining methods accurately identify brain changes predictive of MCI conversion to AD, offering hope for early intervention.
Area of Science:
- Neuroimaging and computational neuroscience
- Biostatistics and machine learning in healthcare
Background:
- Mild cognitive impairment (MCI) significantly increases the risk of developing Alzheimer's disease (AD).
- Atrophic changes in widespread cortical and subcortical brain regions are observed in MCI patients, preceding AD dementia onset.
Purpose of the Study:
- To develop and validate a novel data mining framework for predicting the conversion from MCI to AD.
- To identify specific brain regions predictive of MCI to AD progression using advanced classification techniques.
Main Methods:
- Utilized a data mining framework with Support Vector Machine (SVM), Bayes statistics, and Voting Feature Intervals (VFI) classifiers.
- Collected MRI data from 32 AD patients, 24 MCI subjects, and 18 healthy controls (HC).
- Employed feature selection algorithms to identify discriminating brain regions and applied cross-validation for accuracy assessment.
Main Results:
- Achieved up to 92% classification accuracy in discriminating between AD and HC using identified brain regions.
- Identified anterior cingulate gyrus and orbitofrontal cortex as the most predictive brain areas for MCI conversion to AD.
- Attained a cross-validated prediction accuracy of 75% for MCI to AD conversion.
Conclusions:
- Novel multivariate pattern matching methods demonstrate clinically relevant accuracy for the a priori prediction of MCI progression to AD.
- The findings support the potential of advanced computational approaches in early AD diagnosis and prognosis.
