Department of Biomedical Engineering, Duke University, Durham, North Carolina, USA.
Neurogastroenterol Motil. 2023 Dec;35(12):e14678. doi: 10.1111/nmo.14678. Epub 2023 Sep 22.
The neural control of gastrointestinal muscle relies on circuit activity whose underlying motifs remain limited by small-sample calcium imaging recordings confounded by motion artifact, paralytics, and muscle dissections. We present a sequence of resources to register images from moving preparations and identify out-of-focus events in widefield fluorescent microscopy.
Our algorithm uses piecewise rigid registration with pathfinding to correct movements associated with smooth muscle contractions. We developed methods to identify loss-of-focus events and to simulate calcium activity to evaluate registration.
By combining our methods with principal component analysis, we found populations of neurons exhibit distinct activity patterns in response to distinct stimuli consistent with hypothesized roles. The image analysis pipeline makes deeper insights possible by capturing concurrently calcium dynamics from more neurons in larger fields of view. We provide access to the source code for our algorithms and make experimental and technical recommendations to increase data quality in calcium imaging experiments.
These methods make feasible large population, robust calcium imaging recordings and permit more sophisticated network analyses and insights into neural activity patterns in the gut.
胃肠道肌肉的神经控制依赖于电路活动，但其潜在模式仍然受到运动伪影、肌肉麻痹和肌肉解剖的小样本钙成像记录的限制。我们提出了一系列资源，用于注册来自移动制剂的图像，并识别宽场荧光显微镜中的离焦事件。
我们的算法使用带有寻径的分段刚性配准来校正与平滑肌收缩相关的运动。我们开发了识别离焦事件的方法，并模拟钙活性来评估配准。
通过将我们的方法与主成分分析相结合，我们发现神经元群体表现出与假设角色一致的不同刺激的独特活动模式。通过从更大的视野中同时捕获更多神经元的钙动力学，图像分析管道可以实现更深入的见解。我们提供了我们算法的源代码，并提出了实验和技术建议，以提高钙成像实验的数据质量。
这些方法使得可行的大型群体、稳健的钙成像记录成为可能，并允许更复杂的网络分析和对肠道中神经活动模式的深入了解。
