Artificial Intelligence Based Psychological Detection of HTP Drawings

Hatipoglu, Munevver Irem (2024) Artificial Intelligence Based Psychological Detection of HTP Drawings. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (4MB) | Preview
Preview	PDF (Configuration Manual) Download (4MB) | Preview

Abstract

Art therapy, unlike traditional psychotherapy approaches, is an approach that allows therapists to gain insight into the unconscious world of their patients by providing nonverbal communication. The House-Tree-Person (HTP) test, a drawingbased art therapy approach, is a significant technique used by experts, particularly on children, to provide analysis of familial bonds, interpersonal interactions, and children’s self-perceptions. This test, in which professionals analyse psychological responses by analysing different elements in drawings and rating children’s responses to specific questions about these drawings, is likely to suffer from mistakes. Developments in the artificial intelligence show promise in eliminating these errors and automating the process to ensure that children in need of psychological assistance reach the necessary treatment as quickly as possible. This paper thoroughly covers the fundamental principles of our study, which include using children’s HTP drawings to detect mental disorders such as depression and anxiety, using advanced data augmentation methods, evaluating fine-tuned deep learning models for feature extraction from images, and training machine learning models with these extracted features to assess psychological scores. The results of the feature extraction experiment have shown that the fine-tuned VGG16 model has the highest accuracy with 97%. Using this model, the image features were extracted in terms of depression and anxiety. While the SVM model showed the best performance with a 79% accuracy performance in the depression detection experiment, Random Forest with a 68% accuracy in the anxiety detection experiment was the best performing model among others.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Raj, Kislay UNSPECIFIED
Uncontrolled Keywords:	Art therapy; House-Tree-Person test; Machine learning; Transfer Learning; Deep Learning; Data Augmentation; Explainable AI
Subjects:	B Philosophy. Psychology. Religion > Psychology Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence B Philosophy. Psychology. Religion > Psychology > Child psychology Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	18 Jun 2025 11:49
Last Modified:	18 Jun 2025 11:50
URI:	https://norma.ncirl.ie/id/eprint/7912

Actions (login required)

View Item