With the development of technology, artificial intelligence (AI) is gradually integrating into all aspects of life. In the medical field, scholars and researchers are increasingly using AI for disease diagnosis and treatment, and have made some progress in this area. Recently, researchers used AI to detect behavioral signs of anxiety with an accuracy rate exceeding 90%, and suggested that AI could be applied to address mental health and well-being issues in the future.
In the two years since the outbreak of the COVID-19 pandemic, coupled with a series of climate disasters, more and more people are feeling anxious. This study seems to suggest that artificial intelligence can provide a highly reliable method for identifying signs of anxiety in someone.
The study created an activity dataset of typical anxiety behaviors for sensor detection, including sitting idly, nail-biting, knuckle cracking, and hand-tapping. Participants wore sensors that recorded their movements and performed a series of activities in a specific sequence. Researchers used deep learning algorithms and computational hybrid models to analyze the participants' behavior.
According to the American Psychiatric Association (APA), anxiety disorder is the most common form of mental disorder. Researchers believe that artificial intelligence can help analyze, diagnose, treat, and monitor mental disorders such as anxiety disorder (AD).
This study proposes a Human Activity Recognition (HAR) method to identify certain behaviors associated with anxiety disorders. To build such a model, researchers used motion sensors and inertial measurement units (IMUs) from smartphones to create a new dataset of anxiety behaviors with unique functionalities.
In addition, the study created several deep learning-based models and compared them with random forest and gradient boost algorithms. The study confirmed that one deep model, composed of convolutional neural networks (CNNs) and long short-term memory (LSTMs), outperformed the other models, achieving an accuracy of over 92% in identifying anxiety-related behaviors.
Researchers believe that rapid advancements in artificial intelligence and sensor technologies have made it possible to process data related to mental, emotional, and behavioral disorders. Further research and exploration can then be conducted to understand ineffable behaviors and improve overall mental health.
