Infrared videography of a subcutaneous knee tattoo as a simple and inexpensive method to overcome skin motion artifact in rodent kinematics

Abstract

Kinematic analyses of rodent behavior are frequently used in neuroscience research, and commonly in spinal cord injury (SCI) studies. Unfortunately, skin motion artifact introduces significant errors into these data, because the skin is only loosely coupled to the underlying skeleton by connective tissue. In rats, these errors can be as large as 50-75%,as quantified by past work using x-ray fluoroscopy. Here we show that infrared videography of a subcutaneous tattoo can overcome skin motion artifact in rodent kinematics. The method yields data similar to gold standard x-ray fluoroscopy systems at a fraction of the cost, does not affect the animals locomotion, and results in markers that persist for at least 10 weeks. We found that, compared to a gold-standard x-ray fluoroscopy study that directly tracked the skeleton, our method reduced the error in mean hip angle from 17 {+/-} 6.0 to 3.1 {+/-} 2.4 degrees (mean {+/-} SEM), and the root-mean-square (RMS) error across the mean hip angle waveform from 20 to 5.3 degrees (n=4 rats). The knee joint angle waveform derived from infra-red imaging tightly matched the shape of the x-ray waveform after allowing for a constant offset, having RMS error reduced from 8.1 to 1.2 degrees. The method stands to significantly reduce between-animal errors, and hence between laboratory errors, in these ubiquitous model systems, especially important in SCI studies where individuals are assigned to different treatments.