Related Experiment Video
Updated: Jan 17, 2026
Haptic/Graphic Rehabilitation: Integrating a Robot into a Virtual Environment Library and Applying it to Stroke Therapy
Published on: August 8, 2011
THE USE OF ARTIFICIAL INTELLIGENCE IN VISUALISING RECOVERY OF PSYCHIATRIC REHABILITATION PATIENTS: CREATING GRAPHICS
Agata Łosiewicz1, Maciej Loska1, Barbara Kwaśnica1
1Students' Scientific Association, Department and Clinic of Rehabilitation Psychiatry, Faculty of Medical Sciences in Katowice, Medical University of Silesia, Katowice, Poland.
Background:
Many psychiatric patients struggle to visualise their recovery, decreasing treatment motivation. Globally, 280 million people live with depression, 24 million with schizophrenia, and 40 million with bipolar disorder. First-line treatments achieve remission in only 30-45% of depression cases and 20-60% of schizophrenia cases, with full recovery rates at 10-20%. Artificial intelligence (AI) is increasingly applied in psychiatry for psychoeducation, symptom monitoring, and therapy support. GPT-4o is a generative AI tool producing personalised text, speech, and images. No studies have explored its use for creating recovery-focused visuals to motivate psychiatric patients. This study investigated the potential of ChatGPT-generated visuals as potential therapeutic tools.
Subjects And Methods:
Twenty psychiatric outpatients in remission (schizophrenia, affective, developmental disorders) completed a structured questionnaire with demographic and open-ended questions on recovery expectations. Based on responses, AI-generated recovery visuals were created using GPT-4o and presented for evaluation. Attitudes towards AI were assessed before and after. Participants rated how strongly each image reflected their recovery vision and motivational impact (0-4 scale). Data were analysed using descriptive statistics, paired t-tests, Spearman's correlations, and cluster analysis (Excel, Jamovi, Python).
Results:
Attitudes towards AI improved post-intervention (M=1.70, SD=0.80 vs. M=2.15, SD=0.67). Ratings indicated moderate to strong reflection of personal visions (Graphic 1: M=2.80, SD=1.15; Graphic 2: M=3.25, SD=0.91). No significant differences occurred across demographic groups (p>0.05). A strong positive correlation was found between attitudes towards AI and openness to using AI visuals clinically (ρ=0.65, p=0.002). Cluster analysis identified three profiles: positive adopters (60%), sceptics (25%), and emotionally engaged but technologically sceptical (15%).
Conclusions:
AI-generated images were well-received, improved attitudes towards AI, and enhanced patient motivation. Integrating generative AI images into psychiatric rehabilitation may support engagement and personalised care.
