No technology currently exists for objective assessment of handwriting, even though over 3 million individuals in the United States suffer from disorders that affect handwriting, including 2.7 million children with dyspraxia and over 300,000 older adults with Parkinson's disease. This lack of assessment technology creates a barrier to effective care, because occupational therapists cannot easily distinguish motor disorders from non- motor causes of handwriting disability, a necessary step toward developing appropriate treatment plans. Our goal is to develop Write to Hand, the first quantitative digital assessment of the motor skills that underlie handwriting. In partnership with Washington University in St. Louis, PlatformSTL is uniquely suited to address the handwriting assessment gap. Our team is led by a neuroscientist who developed the Precision Drawing Task, a laboratory research tool with an established history of successful assessment and training of writing-related motor skills. The Precision Drawing Task will serve as the basis for Write to Hand. Our Phase I objective is to develop a fully functional prototype of Write to Hand, and demonstrate scientific validity in children grades 4-5. Data analysis will include an initial application of a machine learning approach, which will test its feasibility in advance of its primary role in Phase II.
Aim 1 of this STTR is to develop the Write to Hand iPad app, which must meet seven specific criteria including: high-precision data collection with a fine-point stylus; rapid calculation of movement speed, smoothness, straightness, and error rate; and data anonymization that meets HIPAA and IRB standards.
Aim 2 will demonstrate validity by comparing Write to Hand performance against a handwriting benchmark in 56 children grades 4-5. We expect that movement smoothness will significantly and meaningfully (r2 > 0.25) predict handwriting skill, and we will use a machine learning approach (Generalized Factorial Method) determine the most predictive classifier for handwriting skill, identify task features that optimize handwriting prediction, and demonstrate feasibility of our machine learning approach to characterize and classify Write to Hand data. Our product will transform therapy for individuals with handwriting disabilities, by providing educators, therapists, and researchers with a gold standard for assessment and quantification of handwriting's underlying motor control skills. This will, for the first time, allow objective identification of motor impairment in local and tele-health settings. In Phase II we will collect data across participant ages to train our machine learning algorithm so Write to Hand can label performance with easy-to-interpret grade level ratings. Our commercialization plan focuses on occupational therapists and educators who serve children with motor disabilities.
Therapists currently have no objective tools to measure handwriting in children or impaired adults. This project will develop an iPad app to measure the movement skills that support effective handwriting. With this tool, therapists will be able to identify when handwriting disability arises from a movement problem, and develop treatment plans accordingly.