在3D共軛焦昆蟲腦影像上做神經纖維路徑的追蹤

Abstract

找出3D昆蟲腦影像中的神經纖維路徑對神經科學的研究幫助很大，本論文希望發展出一個追蹤神經纖維路徑的方法。使用者需要給定神經纖維的初始位置和方向，演算法接著沿著神經纖維的走向判斷出下一個位置。演算法是由神經纖維的幾何性質以及matched filter的原理所設計出來，此外，演算法也判斷沿著神經纖維的走向是否有分支的存在。實驗結果顯示，本文提出的方法可以偵測出大部分的主要分支。

Finding neuron fibers in insect brains is helpful for the filed of neural science. In this thesis, we developed a method to trace the neuron fibers. The user should provide an initial point and an initial direction of the neuron fiber. The method then searches for the next point along the neuron fiber. The method is developed based on the geometric properties of the neuron fiber and the matched filter technique. We can also determine if there are two branches along the neuron fibers. Experimental results shows that the propose method can detect most of the major branches.