Automated Grooming Detection of Mouse by Three-Dimensional Convolutional Neural Network.

Grooming is a common behavior for animals to care for their fur, maintain hygiene, and regulate body temperature. Since various factors, including stressors and genetic mutations, affect grooming quantitatively and qualitatively, the assessment of grooming is important to understand the status of experimental animals. However, current grooming detection methods are time-consuming, laborious, and require specialized equipment. In addition, they generally cannot discriminate grooming microstructures such as face washing and body licking. In this study, we aimed to develop an automated grooming detection method that can distinguish facial grooming from body grooming by image analysis using artificial intelligence. Mouse behavior was recorded using a standard hand camera. We carefully observed videos and labeled each time point as facial grooming, body grooming, and not grooming. We constructed a three-dimensional convolutional neural network (3D-CNN) and trained it using the labeled images. Since the output of the trained 3D-CNN included unlikely short grooming bouts and interruptions, we set posterior filters to remove them. The performance of the trained 3D-CNN and filters was evaluated using a first-look dataset that was not used for training. The sensitivity of facial and body grooming detection reached 81.3% and 91.9%, respectively. The positive predictive rates of facial and body grooming detection were 83.5% and 88.5%, respectively. The number of grooming bouts predicted by our method was highly correlated with human observations (face: r = 0.93, body: r = 0.98). These results highlight that our method has sufficient ability to distinguish facial grooming and body grooming in mice.

Development of Monoclonal Antibody-Based EIA for Tetranor-PGDM which Reflects PGD2 Production in the Body.

Tetranor-PGDM is a metabolite of PGD2. Urinary tetranor-PGDM levels were reported to be increased in some diseases, including food allergy, Duchenne muscular dystrophy, and aspirin-intolerant asthma. In this study, we developed a monoclonal antibody (MAb) and a competitive enzyme immunoassay (EIA) for measuring tetranor-PGDM. Spleen cells isolated from mice immunized with tetranor-PGDM were utilized to generate Ab-producing hybridomas. We chose hybridomas and purified MAb against tetranor-PGDM to develop competitive EIA. The assay evaluated the optimal ionic strength, pH, precision, and reliability. Specificity was determined by cross-reactivity to tetranor-PGEM, tetranor-PGFM, and tetranor-PGAM. Recovery was determined by spiking experiments on artificial urine. Optimal ionic strength was 150 mM NaCl, and optimal pH was pH 7.5. Metabolites other than tetranor-PGDM did not show any significant cross-reactivity in the EIA. The assay exhibited a half-maximal inhibition concentration (IC50) of 1.79 ng/mL, limit of detection (LOD) of 0.0498 ng/mL, and range of quantitation (ROQ) value of 0.252 to 20.2 ng/mL. The intra- and inter-assay variation for tetranor-PGDM was 3.9-6.0% and 5.7-10.4%, respectively. The linearity-dilution effect showed excellent linearity under dilution when artificial urine samples were applied to solid-phase extraction (SPE). After SPE, recovery of tetranor-PGDM in artificial urine averaged from 82.3% to 113.5% and was within acceptable limits (80%-120%). We successfully generated one monoclonal antibody and developed a sensitive competitive EIA. The established EIA would be useful for routine detection and monitoring of tetranor-PGDM in research or diagnostic body fluids.

Automated detection of mouse scratching behaviour using convolutional recurrent neural network.

Scratching is one of the most important behaviours in experimental animals because it can reflect itching and/or psychological stress. Here, we aimed to establish a novel method to detect scratching using deep neural network. Scratching was elicited by injecting a chemical pruritogen lysophosphatidic acid to the back of a mouse, and behaviour was recorded using a standard handy camera. Images showing differences between two consecutive frames in each video were generated, and each frame was manually labelled as showing scratching behaviour or not. Next, a convolutional recurrent neural network (CRNN), composed of sequential convolution, recurrent, and fully connected blocks, was constructed. The CRNN was trained using the manually labelled images and then evaluated for accuracy using a first-look dataset. Sensitivity and positive predictive rates reached 81.6% and 87.9%, respectively. The predicted number and durations of scratching events correlated with those of the human observation. The trained CRNN could also successfully detect scratching in the hapten-induced atopic dermatitis mouse model (sensitivity, 94.8%; positive predictive rate, 82.1%). In conclusion, we established a novel scratching detection method using CRNN and showed that it can be used to study disease models.

Production of lipid mediators across different disease stages of dextran sodium sulfate-induced colitis in mice.

Although several studies have revealed the role of different lipid mediators in colitis, the comprehensive analysis of their production across different phases of colitis remained unclear. Here, we performed the following analysis in the dextran sodium sulfate (DSS)-induced colitis model using LC-MS/MS. Oral administration of 2% DSS in mice for 4 days resulted in severe intestinal inflammation by day 7, which gradually subsided by day 18. Based on the disease scoring index (assigned on the basis of fecal condition and weight loss), we defined the phases of colitis as induction (days 0-4), acute inflammation (days 4-7), recovery (days 7-9), and late recovery (days 9-18). Across all phases, 58 lipid mediators were detected in the inflamed colon tissue. In the induction phase, the production of n-6 fatty acid-derived prostaglandin E2 and thromboxane B2 increased by ∼2-fold. In the acute inflammation phase, the production of n-6 fatty acid-derived leukotrienes increased by >10-fold, while that of n-3 fatty acid-derived hydroxyeicosapentaenoic acids and dihydroxyeicosatetraenoic acids decreased. In the recovery phase, a precursor of protectin D1 (17-hydroxydocosahexaenoic acid) increased over 3-fold. These observations suggested dynamic changes in the production of lipid mediators across different phases of the disease and their potential regulation in healing colitis.