Abstract:In this thesis the aim is to work on optimizing the modern machine learning models for personalized forecasting of Alzheimer Disease (AD) Progression from clinical trial data. The data comes from the TADPOLE challenge, which is one of the largest publicly available datasets for AD research (ADNI dataset). The goal of the project is to develop machine learning models that can be used to perform personalized forecasts of the participants cognitive changes (e.g., ADAS-Cog13 scores) over the time period of 6,12, 18 and 24 months in the future and the change in Clinical Status (CS) i.e., whether a person will convert to AD within 2 years or not. This is important for informing current clinical trials and better design of future clinical trials for AD. We will work with personalized Gaussian processes as machine learning models to predict ADAS-Cog13 score and Cox model along with a classifier to predict the conversion in a patient within 2 <a class="link-external link-http" href="http://years.This" rel="external noopener nofollow">this http URL</a> project is done with the collaboration with researchers from the MIT MediaLab.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to achieve personalized prediction of Alzheimer's disease (AD) progression by optimizing modern machine - learning models. Specifically, the research aims to develop machine - learning models that can predict changes in patients' cognitive abilities (such as ADAS - Cog13 scores) and changes in clinical status (CS) within the next 6, 12, 18, and 24 months, that is, to predict whether patients will convert to Alzheimer's disease patients within two years. ### Problem Definition 1. **Predict whether patients will convert to Alzheimer's disease patients within the next two years**: - Researchers used the dataset from the TADPOLE challenge, which is one of the largest publicly available Alzheimer's disease research datasets (ADNI dataset). - The goal is to develop machine - learning models that can predict changes in patients' cognitive abilities (such as ADAS - Cog13 scores) within the next 6, 12, 18, and 24 months, and combine the Cox model and classifiers to predict whether patients will convert to Alzheimer's disease patients within two years. 2. **If patients will convert, predict the time required for conversion**: - Use Gaussian Processes (GP) as the main machine - learning model for personalized prediction. - Specific steps include: - First, use the group - level GP model to train data. - Then, through the domain - adaptive GPs method, personalize the group GP model into a personalized GP (pGP) for the target patient. - For the first visit of each patient, use the group model for prediction; for subsequent visits, make predictions based on the previous personalized model. - Finally, use these predicted ADAS - Cog13 scores and changes in patients' clinical status to predict whether patients will convert to Alzheimer's disease patients within two years. ### Main Challenges - **Data Missing Problem**: The data of many patients are missing at certain time points, especially the 24 - month visit data. Researchers need to fill these missing values to ensure that all patients have complete data for 6, 12, 18, and 24 months. - **Large - Scale Data Processing**: Process multi - modal feature data (including demographics, cognitive tests, cerebrospinal fluid, MRI, PET, etc.) of approximately 1,737 patients and extract useful information from it. - **Complexity of Personalized Prediction**: It is necessary to develop personalized machine - learning models to adapt to individual differences among different patients, thereby improving the accuracy of prediction. ### Solutions - **Gaussian Process Regression Models**: - Use Gaussian processes for regression prediction, especially applying domain - adaptive GPs in personalized prediction. - Improve the accuracy and robustness of prediction by combining auto - regressive GPs and personalized GPs. - **Survival Analysis**: - Use Cox's proportional hazards regression model to predict the probability of patients converting to Alzheimer's disease at different time points. - Combine the predicted ADAS - Cog13 scores and changes in clinical status to finally determine whether patients will convert to Alzheimer's disease patients within two years. Through these methods, researchers hope to be able to provide more accurate prediction information for current clinical trials and provide better guidance for the design of future clinical trials.

Machine Learning for Health: Personalized Models for Forecasting of Alzheimer Disease Progression

Forecasting the Progression of Alzheimer's Disease Using Neural Networks and a Novel Pre-Processing Algorithm

Deep learning for comprehensive forecasting of Alzheimer's Disease progression

Using Machine Learning and Electronic Health Record (EHR) Data for the Early Prediction of Alzheimer's Disease and Related Dementias.

Machine learning on longitudinal multi-modal data enabling the understanding and prognosis of Alzheimer’s disease progression

A Machine Learning Approach for Predicting Deterioration in Alzheimer's Disease

Application of Machine Learning to Predict the Risk of Alzheimer's Disease: An Accurate and Practical Solution for Early Diagnostics

Updated Models of Alzheimer's Disease with Deep Neural Networks

DeepAD: A Robust Deep Learning Model of Alzheimer's Disease Progression for Real-World Clinical Applications

Early Prediction of Alzheimer's Disease and Related Dementias Using Real-World Electronic Health Records.

Machine Learning Approach for Improved Longitudinal Prediction of Progression from Mild Cognitive Impairment to Alzheimer’s Disease

A Machine Learning Approach to Predict Cognitive Decline in Alzheimer's Disease Clinical Trials

Learning the progression and clinical subtypes of Alzheimer's disease from longitudinal clinical data

Deep Recurrent Model for Individualized Prediction of Alzheimer's Disease Progression

Early-Stage Alzheimer's Disease Prediction Using Machine Learning Models

Machine learning for modeling the progression of Alzheimer disease dementia using clinical data: a systematic literature review

Modeling disease progression via multi-task learning

Time-to-event prediction using survival analysis methods for Alzheimer's disease progression

A quantitative analysis of Alzheimers Disease and construction of an early Alzheimers detection deep learning system (EADDLS) using MRI data via machine learning along with ADmod: spatiotemporal-aware brain-amyloidβ growth model, using deep encoder-decoder networks about MRI

A Novel Approach Utilizing Machine Learning for the Early Diagnosis of Alzheimer's Disease

TADPOLE Challenge: Accurate Alzheimer's disease prediction through crowdsourced forecasting of future data