Detecting benzodiazepine use through induced eye convergence inability with a smartphone app: a proof-of-concept...

Area of Science:

• Pharmacology
• Ophthalmology
• Digital Health

Background:

• Benzodiazepines (BZDs) are central depressant drugs with high potential for misuse and abuse.
• Current BZD use disorder monitoring relies on traditional urine tests.
• There is a need for objective and accessible methods to detect BZD use.

Purpose of the Study:

• To evaluate the utility of smartphone-based eye-scanning for detecting BZD (lorazepam) effects.
• To assess a novel metric for quantifying eye convergence changes.
• To develop and validate a classifier for BZD ingestion detection.

Main Methods:

• Collected eye-scanning data (non-convergence) using the Previct Drugs app before and after lorazepam ingestion in 12 individuals.
• Developed a novel metric (NCdiff and NCdiffInd) representing changes in iris distance during eye convergence.
• Built logistic regression classifiers using NCdiff and NCdiffInd metrics.

Main Results:

• Eye convergence ability is impaired by lorazepam and is highly individual.
• The individualized metric (NCdiffInd) achieved a superior Area Under the Curve (AUC) of 0.88 compared to the absolute metric (NCdiff, AUC=0.79).
• The classifier based on NCdiffInd demonstrated high functionality in detecting lorazepam ingestion.

Conclusions:

• Lorazepam-induced loss of eye convergence is continuous, individual, and can be partial.
• Smartphone eye-scanning technology shows promise for detecting lorazepam ingestion.
• Individualized classifiers adapted to eye convergence ability enhance detection performance.