Virtual Booth

Neurological disorders pose a significant global burden, impacting millions worldwide. Early diagnosis is crucial for prompt treatment and improved quality of life. However, challenges in timely and cost-effective diagnosis persist, especially in resource-limited settings. To address this, we present SpiralSense, a novel solution leveraging machine learning. It offers a globally accessible web application for early detection of various neurological disorders through spiral drawing analysis. Using just paper and pen, SpiralSense provides an affordable, portable, and non-invasive diagnostic method, transcending language barriers and reducing patient discomfort. It can detect six disorders including Alzheimer's, Cerebral Palsy, Dystonia, Essential Tremor, Huntington's, and Parkinson's, with rapid results in six seconds and an accuracy of 97%, enabling early intervention. We collected spiral drawing data from 118 individuals with neurological disorders and trained a model using PyTorch, achieving 97% accuracy on 70 test images. A user-friendly web interface was built using Gradio and deployed on the HuggingFace platform, ensuring accessibility for healthcare professionals and patients. SpiralSense addresses economic challenges in healthcare, particularly in resource-constrained environments, by offering a stress-free, portable, and cost-effective solution for neurological disorders detection.