Real-time scratching behavior quantification system for laboratory mice using high-speed vision

Abstract

Scratching is a specific behavior induced by itching; it is also a common symptom of many types of dermatitis. For the itching evaluation in animal models, automatic quantification system is needed for objective and accurate observation. In this study, a dedicated real-time motion analysis system is developed for detecting the scratching behavior of laboratory mice in long-time experiments, which enables automated behavior quantification for the development of new drugs for diseases such as atopic dermatitis. This system can detect laboratory mice scratching in a non-invasive method by introducing a specially designed high-speed vision system that can calculate the frame-to-frame difference at a frame rate of 240 fps. A quantification algorithm is also implemented for distinguishing the scratching behavior from other behaviors. In fact, we evaluate the effectiveness of our system by demonstrating the experimental results of scratching behavior detection during the long-time observation of several ICR mice. The results also show the objectiveness and accuracy.

Computing a rodent’s diary

“…A combination of binarization, frame differencing, centroid calculation and filtering is used to monitor a dark mouse in a transparent container filled with water and positioned above an IR illuminator. The same authors [9] used a combination of frame binarization, frame differencing, centroid calculation, edge and contour extraction and a filtering method they developed in [10], to detect six behaviors in mice. These included moving, rearing, immobility, head grooming, left-side scratching and right side scratching.…”

Wildlife@Home: Combining Crowd Sourcing and Volunteer Computing to Analyze Avian Nesting Video

“…Significant work with animals has been done in controlled laboratory settings, simplifying the task of gathering video and animal detection. A common approach is to subtract a uniform background from the animals, which has been used to track white mice on black backgrounds [5], [6], [7] or in water [8], [9], [10]. Tracking and detecting behavior of fruit flies (Drosophilia) [11], [12], [13] has been done in similar settings.…”

Behavior recognition in laboratory mice using HFR video analysis

“…However, all these vision-based methods used standard video images (NTSC 30 fps / PAL 25 fps), whose frame rates are not sufficient to achieve precise behavior analysis for rapid limb movements. Ishii et al [4], [5] have solved this problem for quantifying scratching behavior by using HFR video analysis. In this approach, frame-to-frame difference features for the entire image were calculated for detecting scratching, although the limitation in discriminating multiple behaviors of laboratory animals still remains; this is because they did not consider which body parts are prone to repetitive behavior.…”

“…(4) "head grooming" detection To detect the repetitive behavior of "head grooming," the following algorithm described in [5] is adopted to the frameto-frame difference feature F h (t) for the head region. i) pulse thresholding p hg (t) are calculated as pulses by thresholding F h (t) to determine the presence or absence of motion as follows:…”

“…Thus, a high-frame-rate (HFR) video analysis is effective for accurate and objective quantification of high-speed model behaviors of mice involving rapid motions of their limbs. We have already proposed a quantification algorithm [4] for scratching based on an HFR video and developed a real-time 240-fps image analysis system [5] for long-term quantification. In this approach, it is Y.…”