Multimodal Neuroimaging Meets AI: Enhancing Alzheimer's Diagnosis with PyRadiomics
Disciplines
Biomedical Informatics | Computational Neuroscience
Abstract (300 words maximum)
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that requires early and accurate diagnosis for effective intervention. This research explores how multi-modal data integration can enhance Alzheimer’s disease staging prediction by developing an AI model that classifies patients into normal, mild cognitive impairment (MCI), or AD stages. Unlike traditional methods that rely on clinical assessment to make diagnoses, this study develops an AI-driven approach that integrates clinical and imaging data to improve classification accuracy. The research utilizes the Australian Imaging, Biomarkers & Lifestyle (AIBL) dataset, importing patient clinical data along with PET and MRI scans. First, image features were extracted from 1,312 MRI scans (705 patients) and 1,566 PET scans (829 patients) using PyRadiomics. Each scan yielded 112 features. Then, these extracted features were combined with 46 clinical variables to create a multi-modal dataset. To ensure consistency, data selection was performed by including only patients with both MRI and PET scans and a recorded CDR score, while non-numerical features were removed. This resulted in 270 multi-modal features used to train a machine learning model on 681 patients (1,448 scans). The model demonstrated strong performance, with an overall accuracy of 94% in distinguishing between normal control (NC), mild cognitive impairment (MCI), and Alzheimer’s disease (AD). Binary classification models further highlight the model’s effectiveness, achieving 100% accuracy (AUC = 1.000) in AD vs. NC classification, 93% accuracy (AUC = 0.972) in AD vs. MCI, and 97% accuracy (AUC = 0.949) in MCI vs. NC. This research contributes to the field by proposing a data-driven AI framework for precise AD diagnosis, potentially aiding clinicians in early intervention decisions and improving patient outcomes. Future work will validate the model on larger, diverse cohorts to ensure generalizability.
Academic department under which the project should be listed
CCSE - Computer Science
Primary Investigator (PI) Name
Chen Zhao
Multimodal Neuroimaging Meets AI: Enhancing Alzheimer's Diagnosis with PyRadiomics
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that requires early and accurate diagnosis for effective intervention. This research explores how multi-modal data integration can enhance Alzheimer’s disease staging prediction by developing an AI model that classifies patients into normal, mild cognitive impairment (MCI), or AD stages. Unlike traditional methods that rely on clinical assessment to make diagnoses, this study develops an AI-driven approach that integrates clinical and imaging data to improve classification accuracy. The research utilizes the Australian Imaging, Biomarkers & Lifestyle (AIBL) dataset, importing patient clinical data along with PET and MRI scans. First, image features were extracted from 1,312 MRI scans (705 patients) and 1,566 PET scans (829 patients) using PyRadiomics. Each scan yielded 112 features. Then, these extracted features were combined with 46 clinical variables to create a multi-modal dataset. To ensure consistency, data selection was performed by including only patients with both MRI and PET scans and a recorded CDR score, while non-numerical features were removed. This resulted in 270 multi-modal features used to train a machine learning model on 681 patients (1,448 scans). The model demonstrated strong performance, with an overall accuracy of 94% in distinguishing between normal control (NC), mild cognitive impairment (MCI), and Alzheimer’s disease (AD). Binary classification models further highlight the model’s effectiveness, achieving 100% accuracy (AUC = 1.000) in AD vs. NC classification, 93% accuracy (AUC = 0.972) in AD vs. MCI, and 97% accuracy (AUC = 0.949) in MCI vs. NC. This research contributes to the field by proposing a data-driven AI framework for precise AD diagnosis, potentially aiding clinicians in early intervention decisions and improving patient outcomes. Future work will validate the model on larger, diverse cohorts to ensure generalizability.