Connectomics

FlyCircuit is the world's first 3D single neuron image database. It is a public database for online archiving, cell type inventory, browsing, searching and 3D visualization of individual neurons in the Drosophila brain. FlyCircuit provides tools for analyzing neural connections and for similarity comparisons. These tools are accessible to researchers around the world. 3D image data is converted and stored in the database as one-dimensional sequences. This speeds up calculations, reduces analysis time and makes analyzing tools more efficient. FlyCircuit will continue to have images added to it until the database has about 130,000 images. These images then will be used to construct the Drosophila brain connectome. Gene expression data, which has already been integrated into the database, will provide researchers with a way to target specific genes and manipulate neuronal function. This database was created in cooperation with the National Center for High-performance Computing in Taiwan.

建置，並利用FlyCircuit 2.0資料建立神經元功能群組、解析腦神經網路訊號傳遞路徑、以人工智慧方式進行神經影像形態分類，也利用深度學習技術提高果蠅頭部圖像的分析性能，進而繪製成三維腦神經網路體連結圖譜及預測功能來協助了解果蠅腦的功能。

• 腦科學資料計算儲存系統。
• 建置果蠅全腦腦神經網路體資料庫服務平台(約8萬顆腦神經影像) 。
• 以圖學(Graph)方法分析果蠅腦神經元連結關係，以找出可能之神經元功能群組。
• 解析腦神經網路訊號傳遞路徑，並輔助建立腦結構圖譜。
• 人工智慧方式進行神經影像形態分類。
• 利用深度學習技術提高果蠅頭部圖像的分析性能，完成果蠅頭部影像系統的研發工作，並期望利用深度學習技術加速相關的比對工作，能將既有的實驗結果納入以提升比對的正確性。