A Smart Way to Quickly and Efficiently Diagnose Drug Addiction

SadaNews - Researchers at the University of Cincinnati have developed an innovative artificial intelligence system that can help diagnose substance use disorders with an accuracy of up to 84 percent.

The researchers explained that this system provides a quick and reliable preliminary assessment, allowing doctors to begin treatment at a faster pace instead of waiting hours or days for a traditional diagnosis. The results were published on Thursday in the journal "Mental Health Research."

Diagnosing substance use disorders is a complex task, as many patients tend to deny their condition due to the social stigma associated with addiction, slowing early detection and appropriate treatment. Therefore, relying solely on self-disclosure is insufficient, increasing the need for objective and accurate assessment tools that help determine the level of addiction and its associated mental and physical health behaviors.

The new system relies on analyzing the fundamental behaviors of addiction according to clinical standards in psychiatry, such as loss of control, physical dependence on substances, and social impairments. It can accurately determine the severity of addiction with an accuracy of up to 84 percent, in addition to distinguishing the type of substance used, whether it is stimulants, opioids, or cannabis.

The mechanism of the system is based on a simple task in which participants evaluate 48 images that contain emotional triggers. By analyzing these preferences, it can build an accurate behavioral profile for individuals with substance use disorders, measuring risk inclination, loss tolerance, and general behavior patterns.

This study is among the first research efforts to employ a cognitive computing framework with artificial intelligence to analyze how human judgment is used to predict addiction behaviors and determine the type of substance used and the severity of addiction.

The study involved 3,476 participants aged between 18 and 70, who responded to questionnaires used by the system to predict their behaviors. They also rated their liking or disliking of emotional images to provide data that supports the analysis of human judgment traits and economic and behavioral behaviors.

Analyses showed that individuals with higher addiction severity were more inclined to take risks, less flexible towards losses, more impulsive in their behavior, and less diverse in their preferences, which helped form an accurate behavioral profile for each participant.

Dr. Hans Breiter, the principal researcher of the study at the University of Cincinnati, stated: "This is a new type of artificial intelligence that can predict mental illnesses and their associated conditions, such as addiction."

He added via the university's website: "This tool represents a low-cost first step for quick assessment and screening, and anyone can take the test via their phone or computer. It is scalable and resistant to manipulation."

The researchers indicated that this methodology could be used in the future to assess a broader range of addiction patterns, including behavioral addictions such as excessive use of social media, gaming, or food.