An Integrated Processing System for Temporal Neuron Analysis

摘要

Microtubule's structure in neurons with neurodegenerative disease is very different from normal neurons. For example, one of the major characteristic of Alzheimer's disease is abnormal changes of Tau protein in brain, producing intracellular neurofibrillary tangles and colocalized with microtubule. Microtubules are instructors of neuronal morphogenesis, and the fragmentation of microtubule in neurites is an indicator of abnormal neurons. Analysis of time-lapse microscope images can provide quantitative information of neurons' status under different development or disease stages; however, there are few researches building methods for extracting microtubule features in neurons. An automated system for processing and analyzing temporal microtubule images may be helpful for understanding morphological dynamics in neurons. To have a united criterion for the localization of single particle in continuous images, image registration is important. In this paper, the ImageJ plugin, StackReg, was used for registration, and followed by the brightness analysis in neurite. Part of dendrite in tubulin images of primary hip-pocampal neurons treated with NMDA is cropped and the intensities along the skeletonized dendrite are recorded. The dendrite with NMDA at 5 minutes started to have a huge variation on the intensity distribution. The global intensity decreased while the range of changing broadened. In conclusion, the intensity distributions were different for neurons with varied health conditions, and could be shown from quantitative data of microscope image. In the future, we tend to develop an automated image processing system for microtubule images, and build a standard for identifying healthy and abnormal neurons based on these results.