EC-FST: A novel pipeline for automatically analyzing mouse forced swim test.

Abstract

BACKGROUND: The mouse forced swim test (FST) is widely used to evaluate the efficacy of potential anti-depressant drugs. Traditional methods for analyzing forced swim test results rely on manually setting the threshold for immobility, which is time-consuming and barely reproducible. NEW METHOD: In the present study, we introduced a novel pipeline (EC-FST) by extracting the feature of mouse status instead of calculating immobility time. First, we utilized event camera, a powerful AI tool for dynamic object-tracking framework, to capture the mobile events from mouse forced swim test. By quantifying event numbers and their temporal distribution, we were able to determine mouse's mobile state across time-line. RESULTS: The EC-FST results showed perfect correlation with manual scoring, suggesting that the proposed method is reliable for analyzing forced swim test. We further tested the power of the EC-FST for detecting depressive-like behavior in mouse depression models，including lipopolysaccharide (LPS) injection and chronic restraint stress (CRS). Depressive-model mice exhibited significantly fewer motion events and lower event frequency than controls, aligning with manual scoring. COMPARISON WITH EXISTING METHODS: Unlike traditional threshold-based approaches, EC-FST provides an automated, unbiased, and reproducible analysis of FST behavior, eliminating the subjectivity of manual scoring. CONCLUSION: Leveraging AI-driven event cameras, we established a robust pipeline for analyzing mouse behavior in the FST, offering greater efficiency and reproducibility for depression research.