Digital Trail Making Test: Proposal of the Age Estimation Model Using Multi-Task Learning Neural Network for Evaluation of Attention
Haruto Mukai, Akihiro Yorita, Kojiro Mekata, Shigeru AOMURA, Takenori Obo, Naoyuki Kubota · 2024
“Attention” is like a filter that selects certain information coming in through our eyes, nose, ears, etc. Once attentional function is damaged, it would be difficult to find it attentional disorder based on appearance, and the person himself or herself is not fully aware of the disorder. Trail Making Test (TMT) can determine attentional function, but it is difficult to know the degree to which it is normal. Based on the above, the technical challenges of this research are set as (1) the development of a digital TMT (d-TMT) that can be completed with a single iPad, and (2) the development of a method for evaluating attention functions using the data from d-TMT. This paper summarizes the research results of developing an iPad app and constructing and evaluating a machine learning model based on the hypothesis that it is possible to classify the age groups from time and handwriting information obtained from TMT.