vaa3d_神經元追蹤算法介紹之TReMAP_neuron tracing

主要內容

Zhou Z, Liu X, Long B, et al. 
TReMAP: automatic 3D neuron reconstruction based on tracing, reverse mapping and assembling of 2D projections[J]. 
Neuroinformatics, 2016, 14(1): 41-50.

發表在Neuroinformatics上，SCI二區，很多神經元相關的文章都發在這上面。
TReMAP :Tracing, Reverse Mapping and Assembling of 2D Projections.

主要特點

Virtual Finger
超級大圖像 10GB級別
速度快 內存消耗少

主要流程

1.	MIP 轉化成二維圖片，然後APP生成對應的追蹤結果
2.	將追蹤結果從分支點處分割成不同的段，segment
3.	每一個segment  有一個bounding box ，提取對應的3D區域。這是降低內存需求的主要方法。
4.	CDA2  Virtual Finger。 Curve-Drawing Algorithm 2 
5.	MST minimal spanning tree 最小生成樹
6.	XY平面做基準，其他平面XZ  YZ補充。

BlastNeuron ranking scores

BlastNeuron : Basic Local Alignment Search Tool for Neurons
BlastNeuron was designed to accurately and efficiently retrieve morphologically and functionally similar neuron
structures based on global appearance, detailed arborization
patterns, and topological similarity.