Related Experiment Video
Updated: Dec 27, 2025
06:19
Skin Punch Biopsy Explant Culture for Derivation of Primary Human Fibroblasts
Published on: July 7, 2013
42.6K
MALINI (Machine Learning in NeuroImaging): A MATLAB toolbox for aiding clinical diagnostics using resting-state fMRI
Pradyumna Lanka1,2, D Rangaprakash1,3,4, Sai Sheshan Roy Gotoor1
1AU MRI Research Center, Department of Electrical and Computer Engineering, Auburn University, Auburn, AL, USA.
Data in Brief
|February 25, 2020
Summary
A new MATLAB toolbox, MALINI, extracts brain connectivity features from resting-state fMRI data for machine learning-based disease classification. It demonstrates robust diagnostic performance across diverse neurological and psychiatric disorders.
Area of Science:
- Neuroimaging
- Machine Learning
- Computational Neuroscience
Background:
- Resting-state functional Magnetic Resonance Imaging (rs-fMRI) is valuable for diagnostic classification due to minimal task compliance and multi-site data pooling capabilities.
- Increased sample sizes through multi-site data pooling enhance the reliability of neuroimaging studies.
Purpose of the Study:
- To present the Machine Learning in NeuroImaging (MALINI) toolbox for feature extraction and disease classification using rs-fMRI data.
- To evaluate the performance of 18 machine learning algorithms and a consensus classifier for diagnostic classification.
- To demonstrate the toolbox's utility across diverse brain disorders using large-scale datasets.
Main Methods:
- The MALINI toolbox extracts functional and effective connectivity features from preprocessed rs-fMRI data.
- It employs 18 diverse machine learning algorithms and a consensus classifier for binary classification (healthy vs. disease).
- Data from multiple large-scale datasets (ABIDE, ADNI, ADHD-200) and in-house studies were used, with 80% for training/validation and 20% for independent testing.
Main Results:
- The study demonstrates the utility of the MALINI toolbox for classifying various brain disorders.
- Classification performance was evaluated on datasets including autism spectrum disorder, Alzheimer's disease, mild cognitive impairment, ADHD, PTSD, and post-concussion syndrome.
- Robust classification performance was ensured through independent testing on hold-out data, accounting for population variance.
Conclusions:
- The MALINI toolbox provides a comprehensive platform for diagnostic classification using rs-fMRI data.
- The approach shows promise for robustly differentiating between healthy controls and individuals with various brain-based disorders.
- This tool facilitates large-scale neuroimaging data analysis for advancing diagnostic capabilities.