Canvas Insight: Enhancing AI-Driven Scheduling for Depression Diagnosis Using the House-Tree-Person Projective Test

The rising prevalence of depression and psychiatrist shortages highlight the need for efficient diagnostic tools. Canvas Insight leverages AI with the House-Tree-Person (H-T-P) test to automate depression screening. This study evaluates YOLO-V3, YOLO-V4, and YOLO-V5 on an annotated dataset, with YOLOV3 achieving the highest accuracy (mAP 89% for houses, 88% for persons and 81 % for tree). Using a structured scoring system, the model assesses depression severity to prioritize scheduling, ensuring timely care for those most in need. The intention is not to provide a diagnosis but to assist clinicians in optimizing resource allocation. This approach enhances efficiency, particularly in resource-limited areas, while maintaining AI as a supportive tool for human expertise.