A deep-learning-based threshold-free method for automated analysis of rodent behavior in the forced swim test and tail suspension test.

BACKGROUND: The forced swim test (FST) and tail suspension test (TST) are widely used to assess depressive-like behaviors in animals. Immobility time is used as an important parameter in both FST and TST. Traditional methods for analyzing FST and TST rely on manually setting the threshold for immobility, which is time-consuming and subjective. NEW METHOD: We proposed a threshold-free method for automated analysis of mice in these tests using a Dual-Stream Activity Analysis Network (DSAAN). Specifically, this network extracted spatial information of mice using a limited number of video frames and combined it with temporal information extracted from differential feature maps to determine the mouse's state. To do so, we developed the Mouse FSTST dataset, which consisted of annotated video recordings of FST and TST. RESULTS: By using DSAAN methods, we identify immobility states at accuracies of 92.51 % and 88.70 % for the TST and FST, respectively. The predicted immobility time from DSAAN is nicely correlated with a manual score, which indicates the reliability of the proposed method. Importantly, the DSAAN achieved over 80 % accuracy for both FST and TST by utilizing only 94 annotated images, suggesting that even a very limited training dataset can yield good performance in our model. COMPARISON WITH EXISTING METHOD(S): Compared with DBscorer and EthoVision XT, our method exhibits the highest Pearson correlation coefficient with manual annotation results on the Mouse FSTST dataset. CONCLUSIONS: We established a powerful tool for analyzing depressive-like behavior independent of threshold, which is capable of freeing users from time-consuming manual analysis.

SpindlesTracker: An Automatic and Low-Cost Labeled Workflow for Spindle Analysis.

Quantitative analysis of spindle dynamics in mitosis through fluorescence microscopy requires tracking spindle elongation in noisy image sequences. Deterministic methods, which use typical microtubule detection and tracking methods, perform poorly in the sophisticated background of spindles. In addition, the expensive data labeling cost also limits the application of machine learning in this field. Here we present a fully automatic and low-cost labeled workflow that efficiently analyzes the dynamic spindle mechanism of time-lapse images, called SpindlesTracker. In this workflow, we design a network named YOLOX-SP which can accurately detect the location and endpoint of each spindle under box-level data supervision. We then optimize the algorithm SORT and MCP for spindle's tracking and skeletonization. As there was no publicly available dataset, we annotated a S.pombe dataset that was entirely acquired from the real world for both training and evaluation. Extensive experiments demonstrate that SpindlesTracker achieves excellent performance in all aspects, while reducing label costs by 60%. Specifically, it achieves 84.1% mAP in spindle detection and over 90% accuracy in endpoint detection. Furthermore, the improved algorithm enhances tracking accuracy by 1.3% and tracking precision by 6.5%. Statistical results also indicate that the mean error of spindle length is within 1 µm. In summary, SpindlesTracker holds significant implications for the study of mitotic dynamic mechanisms and can be readily extended to the analysis of other filamentous objects. The code and the dataset are both released on GitHub.