Efficient optimization based thresholding technique for analysis of alzheimer MRIs.

Area of Science:

• Medical Imaging
• Neuroscience
• Artificial Intelligence

Background:

• Alzheimer's disease (AD) is a dementia causing memory loss and brain shrinkage.
• Accurate segmentation of grey matter (GM) and white matter (WM) in MRI is challenging due to tissue homogeneity.
• Early detection of AD is crucial for effective treatment and management.

Purpose of the Study:

• To develop an optimization-based segmentation technique for extracting GM and WM tissues.
• To classify normal and Alzheimer's disease (AD) patients using Magnetic Resonance Images (MRI).
• To improve the accuracy and efficiency of AD diagnosis through enhanced image analysis.

Main Methods:

• Utilized a modified cuckoo search algorithm for MRI segmentation.
• Extracted Gray Level Co-Occurrence Matrix (GLCM) features from segmented GM and WM.
• Employed Principal Component Analysis (PCA) for feature selection and Support Vector Machine (SVM) for classification.

Main Results:

• The modified cuckoo search algorithm demonstrated superior image segmentation compared to FCM, Otsu, PSO, and CS.
• Achieved high classification accuracy (96%), sensitivity (97%), and specificity (94%) for AD detection.
• The algorithm's powerful searching capability facilitated precise identification of GM and WM tissues.

Conclusions:

• The proposed technique effectively identifies Alzheimer's affected patients through robust optimization.
• This pipeline facilitates early AD detection and aids in assessing neuroprotective therapy effects.
• The study highlights the potential of advanced optimization algorithms in neuroimaging analysis.