We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Machine Learning for Real-Time, Automatic, and Early Diagnosis of Parkinson's Disease by Extracting Signs of Micrographia from Handwriting Images

Abstract: Parkinson's disease (PD) is debilitating, progressive, and clinically marked by motor symptoms. As the second most common neurodegenerative disease in the world, it affects over 10 million lives globally. Existing diagnoses methods have limitations, such as the expense of visiting doctors and the challenge of automated early detection, considering that behavioral differences in patients and healthy individuals are often indistinguishable in the early stages. However, micrographia, a handwriting disorder that leads to abnormally small handwriting, tremors, dystonia, and slow movement in the hands and fingers, is commonly observed in the early stages of PD. In this work, we apply machine learning techniques to extract signs of micrographia from drawing samples gathered from two open-source datasets and achieve a predictive accuracy of 94%. This work also sets the foundations for a publicly available and user-friendly web portal that anyone with access to a pen, printer, and phone can use for early PD detection.
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
Cite as: arXiv:2111.14781 [cs.LG]
  (or arXiv:2111.14781v1 [cs.LG] for this version)

Submission history

From: Riya Tyagi S [view email]
[v1] Wed, 24 Nov 2021 20:04:27 GMT (1607kb,D)

Link back to: arXiv, form interface, contact.