Recognition Of Depression By Non-Psychiatric Physician

Abstract

Depression is a widespread mental health disorder that often goes unrecognized, especially in primary care settings where patients commonly consult non-psychiatric physicians. These physicians may focus primarily on physical symptoms, potentially overlooking subtle psychological signs of depression. As a result, many individuals with depression do not receive timely or appropriate mental health care, which can worsen their condition and impact overall well-being.
This project aims to assist non-psychiatric physicians in identifying early signs of depression through facial emotion recognition using machine learning. A Convolutional Neural Network (CNN) model is employed to analyze facial expressions captured during patient interactions. The model is trained on a dataset of emotional images to detect specific emotions such as sadness, anger, or lack of expression that may indicate depressive tendencies. This provides physicians with an additional, objective tool for evaluating a patient's emotional state.
The system serves as a supportive diagnostic aid rather than a replacement for clinical judgment. By integrating facial emotion analysis into routine check-ups, non-psychiatric physicians can be alerted to possible depressive symptoms and refer patients for specialized psychological evaluation. This approach promotes early detection, reduces the chances of misdiagnosis, and helps ensure that patients receive the mental health support they need in a timely manner.