Neural networks for cognitive testing: Cognitive test drawing classification

With the burgeoning population of patients with cognitive disorders such as dementia, healthcare is already having significant difficulty in caring for this new population of patients. However, with the last decade's advances in neural networks, it is possible to begin creating software which may aid in healthcare for these patients. Specifically, it may aid in diagnosis, such as in expediting cognitive examinations. Within this paper, we describe a custom neural network utilizing a SqueezeNet which is used to classify a custom dataset of hand-drawn images commonly used in cognitive examinations. We demonstrate that our model has 97% accuracy. Specifically, this enables the development of entire automated and accurate cognitive examinations. The work presented here demonstrates neural networks may assist with healthcare for patients with cognitive disorders, having impact upon the fields of neurology, psychiatry, and family medicine. Importantly, within the context of the COVID-19 pandemic restricting in-person visits and promoting telemedicine, this provides the foundations to transition cognitive examinations to a telemedicine modality. •Neural networks may classify be used to classify cognitive examinations.•SqueezeNet may achieve up to 97% accuracy on cognitive exam picture classifications.•Gating SqueezeNet outcomes by probability may allow improvement of accuracy.