ABSTRACT
Parkinson's Disease (PD) is a prevalent neurological condition characterized by motor and cognitive impairments, typically manifesting around the age of 50 and presenting symptoms such as gait difficulties and speech impairments. Although a cure remains elusive, symptom management through medication is possible. Timely detection is pivotal for effective disease management. In this study, we leverage Machine Learning (ML) and Deep Learning (DL) techniques, specifically K-Nearest Neighbor (KNN) and Feed-forward Neural Network (FNN) models, to differentiate between individuals with PD and healthy individuals based on voice signal characteristics. Our dataset, sourced from the University of California at Irvine (UCI), comprises 195 voice recordings collected from 31 patients. To optimize model performance, we employ various strategies including Synthetic Minority Over-sampling Technique (SMOTE) for addressing class imbalance, Feature Selection to identify the most relevant features, and hyperparameter tuning using RandomizedSearchCV. Our experimentation reveals that the FNN and KSVM models, trained on an 80-20 split of the dataset for training and testing respectively, yield the most promising results. The FNN model achieves an impressive overall accuracy of 99.11%, with 98.78% recall, 99.96% precision, and a 99.23% f1-score. Similarly, the KSVM model demonstrates strong performance with an overall accuracy of 95.89%, recall of 96.88%, precision of 98.71%, and an f1-score of 97.62%. Overall, our study showcases the efficacy of ML and DL techniques in accurately identifying PD from voice signals, underscoring the potential for these approaches to contribute significantly to early diagnosis and intervention strategies for Parkinson's Disease.
Subject(s)
Machine Learning , Parkinson Disease , Parkinson Disease/diagnosis , Humans , Male , Female , Middle Aged , Aged , Neural Networks, Computer , Voice , Deep LearningABSTRACT
Cavitation erosion remains the primary cause of material degradation in fluid machinery components operating at high speed. Micro-jets/shock waves caused by implosion of bubbles on material surface results in significant material loss and premature failure of the components. The presence of corrosive medium further exuberates this effect, causing rapid degradation. Here, we demonstrate a novel pathway to control cavitation erosion-corrosion by tailoring the surface properties using submerged friction stir processing (FSP), a severe plastic deformation process. FSP parameters were varied over wide range of strain-rates to generate tailored microstructures. High strain-rate processing resulted in nearly single phase fine grained structure while low strain-rate processing resulted in phase transformation in addition to grain refinement. As-received and processed samples were subjected to ultrasonic cavitation in distilled water as well as in corrosive environment of 3.5% NaCl solution. Individual roles of cavitation erosion, corrosion and their synergistic effects were analyzed. Depending on the microstructure, processed samples showed nearly 4-6â¯times higher cavitation erosion resistance compared to as-received alloy. Superior cavitation erosion-corrosion resistance of processed samples was attributed to surface strengthening, higher strain-hardening ability and quick passivation kinetics. The results of current study could be potentially transformative in designing robust materials for hydro-dynamic applications.
