Prediction of Alzheimer's Disease by Analyzing Handwriting Dynamics Using Machine Learning Algorithms

Alzheimer's disease is a degenerative neurological condition that impairs everyday functioning and quality of life. Handwritten data scripts are related in the determination of Alzheimer's disease with the use of handwriting patterns, which can reveal early signs of cognitive decline and motor impairments. Changes in handwriting, such as tremors, irregularities, and slower writing speeds, can be indicative of the onset and progression of Alzheimer's, aiding in early diagnosis and monitoring of the disease. Traditionally, statistical tests or conventional classification models have been used in handwriting analysis studies. Machine learning techniques have only lately been investigated, and they were mostly used to identify Parkinson's disease rather than Alzheimer's. This research work aims to close this gap and investigate this by developing a machine learning model that uses handwriting data to categorize an individual as an Alzheimer patient. This research aims to advance the science of Alzheimer's disease prediction by examining the use of handwritten data as a unique medium for early diagnosis. PCA is one of the data preprocessing parameters that are employed, while the linear support vector classifier, random forest classifier, and XGBoost, together with their accuracy, are the model training parameters. Each participant's handwriting sample yielded 451 handwriting features, such as air_time1, disp_index1, gmrt_in_air1, gmrt_on_paper1, mean_gmrt1, etc., which are considered as the input parameters. A binary class label indicating whether a person is healthy or may have Alzheimer's disease is considered as the expected output.