Deep Instance Segmentation of Laboratory Animals in Thermal Images

DAMM for the detection and tracking of multiple animals within complex social and environmental settings

Accurate detection and tracking of animals across diverse environments are crucial for studying brain and behavior. Recently, computer vision techniques have become essential for high-throughput behavioral studies; however, localizing animals in complex conditions remains challenging due to intra-class visual variability and environmental diversity. These challenges hinder studies in naturalistic settings, such as when animals are partially concealed within nests. Moreover, current tools are laborious and time-consuming, requiring extensive, setup-specific annotation and training procedures. To address these challenges, we introduce the 'Detect-Any-Mouse-Model' (DAMM), an object detector for localizing mice in complex environments with minimal training. Our approach involved collecting and annotating a diverse dataset of single- and multi-housed mice in complex setups. We trained a Mask R-CNN, a popular object detector in animal studies, to perform instance segmentation and validated DAMM's performance on a collection of downstream datasets using zero-shot and few-shot inference. DAMM excels in zero-shot inference, detecting mice and even rats, in entirely unseen scenarios and further improves with minimal training. Using the SORT algorithm, we demonstrate robust tracking, competitive with keypoint-estimation-based methods. Notably, to advance and simplify behavioral studies, we release our code, model weights, and data, along with a user-friendly Python API and a Google Colab implementation.

Machine Learning for Biomedical Application

The tremendous development of technology also affects medical science, including imaging diagnostics [...]

Experimental Applications and Factors Involved in Validating Thermal Windows Using Infrared Thermography to Assess the Health and Thermostability of Laboratory Animals

Evaluating laboratory animals' health and thermostability are fundamental components of all experimental designs. Alterations in either one of these parameters have been shown to trigger physiological changes that can compromise the welfare of the species and the replicability and robustness of the results obtained. Due to the nature and complexity of evaluating and managing the species involved in research protocols, non-invasive tools such as infrared thermography (IRT) have been adopted to quantify these parameters without altering them or inducing stress responses in the animals. IRT technology makes it possible to quantify changes in surface temperatures that are derived from alterations in blood flow that can result from inflammatory, stressful, or pathological processes; changes can be measured in diverse regions, called thermal windows, according to their specific characteristics. The principal body regions that were employed for this purpose in laboratory animals were the orbital zone (regio orbitalis), auricular pavilion (regio auricularis), tail (cauda), and the interscapular area (regio scapularis). However, depending on the species and certain external factors, the sensitivity and specificity of these windows are still subject to controversy due to contradictory results published in the available literature. For these reasons, the objectives of the present review are to discuss the neurophysiological mechanisms involved in vasomotor responses and thermogenesis via BAT in laboratory animals and to evaluate the scientific usefulness of IRT and the thermal windows that are currently used in research involving laboratory animals.