deletions | additions       diff --git a/tracking_pumping_in_high_quality__.tex b/tracking_pumping_in_high_quality__.tex  index 20947dc..a453fa4 100644  --- a/tracking_pumping_in_high_quality__.tex  +++ b/tracking_pumping_in_high_quality__.tex 
...
tracking Tracking  pumping in high-quality imaging data sets with large magnification and high-framerates has been done for long times. can be achieved using conventional image analysis tools.  These techniques however require expensive equipment (cameras and microscopes) and the use of large magnification ( 20x or higher) reduces the number of aimals animals  that can be imaged at the same time. Our goal was to achieve accurate, long term pumping time series while increasing the throughput one can achieve. At low magnifications, traditional approaches like directly tracking the position of the grinder are infeasible and highly sensitive to noise. The amplitude of the grinder motion is small, and thus any translation of the worm or elastic bending will make the tracking inaccurate. We therefore decided to use an intensity-based approach that utilizes the separation of timescales between locomotion and pumping and that integrates over a larger area, therefore increasing the signal during a pumping event. In developing a setup that can handle multiplexing, we considered the trade-off between high-quality hardware or more involved post-processing. We decided to use simple educational microscopes and entry-level scientific cameras, which make multiplexing by simply copying the setup financially accessible. To still achieve the desired data quality, we develop a custom image analysis pipeline that can handle the noisy images created by the inexpensive hardware.