Long B L, Li H, Mahadevan A, Tang T, Balotin K, Grandel N, Soto J, Wong S Y, Abrego A, Li S, Qutub A A
Department of Bioengineering, Rice University, Houston, TX 77030 USA.
Department of Bioengineering, Rice University, Houston, TX 77030 USA.
J Neurosci Methods. 2017 May 1;283:62-71. doi: 10.1016/j.jneumeth.2017.03.013. Epub 2017 Mar 21.
Neurite outgrowth is a metric widely used to assess the success of in vitro neural stem cell differentiation or neuron reprogramming protocols and to evaluate high-content screening assays for neural regenerative drug discovery. However, neurite measurements are tedious to perform manually, and there is a paucity of freely available, fully automated software to determine neurite measurements and neuron counting. To provide such a tool to the neurobiology, stem cell, cell engineering, and neuroregenerative communities, we developed an algorithm for performing high-throughput neurite analysis in immunofluorescent images.
Given an input of paired neuronal nuclear and cytoskeletal microscopy images, the GAIN algorithm calculates neurite length statistics linked to individual cells or clusters of cells. It also provides an estimate of the number of nuclei in clusters of overlapping cells, thereby increasing the accuracy of neurite length statistics for higher confluency cultures. GAIN combines image processing for neuronal cell bodies and neurites with an algorithm for resolving neurite junctions.
GAIN produces a table of neurite lengths from cell body to neurite tip per cell cluster in an image along with a count of cells per cluster.
GAIN's performance compares favorably with the popular ImageJ plugin NeuriteTracer for counting neurons, and provides the added benefit of assigning neurites to their respective cell bodies.
In summary, GAIN provides a new tool to improve the robust assessment of neural cells by image-based analysis.
神经突生长是一种广泛用于评估体外神经干细胞分化或神经元重编程方案是否成功以及评估用于神经再生药物发现的高内涵筛选试验的指标。然而，手动进行神经突测量很繁琐，并且缺乏可免费获得的、能完全自动确定神经突测量值和神经元计数的软件。为了向神经生物学、干细胞、细胞工程和神经再生领域提供这样一种工具，我们开发了一种用于在免疫荧光图像中进行高通量神经突分析的算法。
给定成对的神经元细胞核和细胞骨架显微镜图像作为输入，GAIN算法计算与单个细胞或细胞簇相关的神经突长度统计数据。它还能估计重叠细胞簇中的细胞核数量，从而提高更高汇合度培养物中神经突长度统计的准确性。GAIN将针对神经元细胞体和神经突的图像处理与一种解决神经突连接的算法相结合。
GAIN生成图像中每个细胞簇从细胞体到神经突尖端的神经突长度表格以及每个簇的细胞计数。
GAIN在计数神经元方面的性能与流行的ImageJ插件NeuriteTracer相比具有优势，并且还具有将神经突分配到其各自细胞体的额外优势。
总之，GAIN提供了一种新工具，可通过基于图像的分析来改进对神经细胞的可靠评估。
