The incidence of Alzheimer's disease is increasing in Oman, according to an expert.
To address this issue, a team of researchers from Oman has created an advanced deep learning model specifically designed to forecast the likelihood of Alzheimer's disease in its initial stages.
The study, spearheaded by Dr. Abraham Varghese, Senior Lecturer in the Information Technology Department at the University of Technology and Applied Sciences in Muscat, represents a significant advancement in the early detection of this debilitating condition.
Over the past few years, numerous innovative and insightful research initiatives have been supported by the Ministry of Higher Education, Research, and Innovation.
One of these projects, 'A personalized machine learning and deep learning model for early-stage Alzheimer's disease risk prediction,' led by Dr. Abraham Varghese, Senior Lecturer in the IT Department at the University of Technology and Applied Sciences in Muscat, is among those funded by the Ministry's Block Funding Program for Higher Education, Research, and Innovation.
Dr. Abraham Varghese, the principal investigator, highlighted in this study that Alzheimer's disease (AD) has emerged as a leading global cause of mortality.
The number of individuals affected by AD is projected to increase from 55 million to 139 million by 2050.
The surge in Alzheimer's disease cases underscores the urgent need for the development of prompt early diagnostic tools, a necessity that is particularly relevant in Oman, where the number of Alzheimer's disease patients is steadily rising.
The study focused on creating an AI-driven diagnostic tool for Alzheimer’s Disease (AD) by utilizing psychological parameters and image features from clinical and MRI measurements. Additionally, Explainable AI (XAI) techniques were integrated into the tool to provide practitioners and researchers with a clear understanding of the model's decision-making process. A user-friendly graphical user interface (GUI) was also developed to enhance model predictions and explanations, ultimately supporting informed clinical decision-making.
Dr. Abraham highlighted the importance of developing AI-based tools for the early detection of AD, particularly during the Mild Cognitive Impairment (MCI) stage to prevent further neurodegeneration associated with Alzheimer’s progression.
To gain insights into the impact of AD on the brain over time, the research team analyzed MRI scans in conjunction with psychological and demographic data obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI).
Following rigorous preprocessing and the application of feature selection algorithms, a series of black box algorithms were utilized, with Random Forest being identified as the most accurate with a 92% accuracy rate.
By incorporating Explainable AI techniques, the research team aimed to enhance model transparency, allowing medical professionals to better understand the predictive outcomes.
Dr. Abraham and his team also developed a web application (https://alzheimer-disease-prediction.streamlit.app/) to assist medical experts in utilizing the model effectively. This tool enables personalized care and quality treatment by providing precise, AI-driven insights into each patient’s condition. Through this endeavor, the team sought to improve clinical decision-making processes by bridging the gap between advanced AI technologies and their practical implementation in healthcare settings for Alzheimer’s disease.
Dr. Abraham's research project has recommended the adoption of an integrative approach to diagnosing Alzheimer's disease (AD). This approach involves considering a wide range of data, including clinical, genetic, demographic, neuroimaging, and psychological aspects, in order to gain a comprehensive understanding of AD.
This comprehensive strategy is crucial because it allows for a more accurate diagnosis of the disease and enables the development of personalized treatment plans. It also captures the multifaceted nature of AD, which is important for effective management.
Dr. Abraham also emphasized the importance of collaboration between AI researchers and medical professionals. This collaboration is necessary to develop AI tools that meet clinical requirements and prioritize patient-centered care. By working together, AI can be integrated into healthcare settings in a way that is effective and sensitive to the needs of patients, ultimately improving the quality of care and the effectiveness of medical interventions.
The research project has already made significant progress, identifying key variables and developing a predictive model and web application for AD. Currently, psychologists assess the input variable scores through direct evaluations. However, the research team plans to innovate further by creating a virtual psychologist.
This virtual psychologist will be designed to administer psychological assessments and calculate scores autonomously. Its development will greatly enhance the accessibility and applicability of diagnostic tools for AD, allowing for widespread use and facilitating early detection and intervention in diverse settings.
