Comparative Analysis of Vision Transformer, Convolutional, and Hybrid Architectures for Mental Health Classification Using Actigraphy-Derived Images

TL;DR

This study compares vision transformer, convolutional, and hybrid neural network architectures for classifying mental health conditions from actigraphy-derived images, finding hybrid models like CoAtNet-Tiny to be most reliable and accurate.

Contribution

It provides a comparative analysis of different image-based neural architectures for mental health classification using actigraphy data, highlighting the superior performance of hybrid models.

Findings

CoAtNet-Tiny achieved highest average accuracy and stability.

VGG16 showed steady but lower accuracy.

ViT-B/16 had inconsistent performance across folds.

Abstract

This work examines how three different image-based methods, VGG16, ViT-B/16, and CoAtNet-Tiny, perform in identifying depression, schizophrenia, and healthy controls using daily actigraphy records. Wrist-worn activity signals from the Psykose and Depresjon datasets were converted into 30 by 48 images and evaluated through a three-fold subject-wise split. Although all methods fitted the training data well, their behaviour on unseen data differed. VGG16 improved steadily but often settled at lower accuracy. ViT-B/16 reached strong results in some runs, but its performance shifted noticeably from fold to fold. CoAtNet-Tiny stood out as the most reliable, recording the highest average accuracy and the most stable curves across folds. It also produced the strongest precision, recall, and F1-scores, particularly for the underrepresented depression and schizophrenia classes. Overall, the findings indicate that CoAtNet-Tiny performed most consistently on the actigraphy images, while VGG16 and ViT-B/16 yielded mixed results. These observations suggest that certain hybrid designs may be especially suited for mental-health work that relies on actigraphy-derived images.