Until recently, neuroscientists have had to track the firings of individual neurons by hand. The task, which can only be compared to tracking the words of a single fan in a stadium full of football fans have been made easier by the development of an open source software called CalmAn. According to Dmitri Chklovskii, the leader of the neuroscience group at the Center for Computational Biology, they spent most of their time analyzing data to extract activity traces than in its retrieval. The development of CalmAn has increased the analysis process by automation using a combination of computational methods and machine-learning techniques. A journal published in January by the software developers showed CalmAn achieve a near-human accuracy in locating active neurons based on calcium imaging data. CalmAn has been in the market for a few years and has worked to benefit the calcium imaging community. There are over 100 labs that utilize the software. With the upgrades in the latest software, scientists can run it in a standard laptop and analyze data in real time. As such, they can run experiments and analyze data simultaneously. According to John Pearson, a neuroscientist at Duke University, the software has proved invaluable, and his lab was excited at the prospect of using such a tool. Chklovskii, Eftycios Pnevmatikakis, and Andrea Giovannucci spearheaded the development of CalmAn with the aim of developing a tool that would help in tackling the enormous datasets produced by calcium imaging. CalmAn adopts a unique technique that involves the addition of a special dye to brain tissue or neurons in a dish. The dye then binds to the calcium ions responsible for activating neurons. It is then viewed under ultraviolet rays whereby the calcium ion with the dye lights up allowing the scientists or researchers to track the activity of the neuron visually. The process of analyzing data from a calcium imaging process generated a flood of data that can fill up a commercially available hard drive in one day. Additionally, the data is noisy, and signals from different neurons overlap making the identification process of an individual neuron difficult. Worse, the brain tissue jiggles making the process of tracking the same neuron even harder. However, with the development of CalmAn, the process has been made less tedious. Recent tests have shown that the software is nearly as accurate as a human in identifying active neurons. With its speed, neuroscientists can arrive at their goals faster and consequently, help them understand how specific bundles of neurons contribute to different behaviors.
