English  |  正體中文  |  简体中文  |  Items with full text/Total items : 54371/62179 (87%)
Visitors : 8924816      Online Users : 120
RC Version 6.0 © Powered By DSPACE, MIT. Enhanced by NTHU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    National Tsing Hua University Institutional Repository > 生命科學院  > 系統神經科學研究所 > 博碩士論文 >  利用高解析度神經圖譜資料建構果蠅觸角葉訊號處理的神經網路模型

    Please use this identifier to cite or link to this item: http://nthur.lib.nthu.edu.tw/dspace/handle/987654321/79356

    Title: 利用高解析度神經圖譜資料建構果蠅觸角葉訊號處理的神經網路模型
    Authors: 高國維
    Lo, Chung-Chun
    Kao, Kuo-Wei
    教師: 張慧雲
    Date: 2011
    Keywords: 果蠅
    antennal lobe
    local interneuron
    computer simulation
    Abstract: 利用單顆神經細胞解析度果蠅天線丘local interneuron image data,我們分析其神經網路各項統計特徵,建立擬真的人工神經網路,並進行各種電腦模擬實驗。有別與以往許多完全以隨機方式產生的純理論模型,我們的研究是以真實的神經網路數據為基礎。由於數據解析到高達單顆神經程度,我們得以詳細分析神經網路的各項統計特徵,並且設計出一套能夠初步產生擬真果蠅antennal lobe LN 網路的模型。我們的模擬能夠確實重現各種近期被發表的果蠅嗅覺神經活動,諸如周邊抑制 (lateral inhibition),針對抑制 (sparse inhibition),調控信號對比度 (contrast),信號增益幅度控制 (gain control),以及信號分辨程度調控等。尤其在於氣味信號辨別的部份,我們有相當驚人的發現:以幾何距離對ORN以及PN反應進行分析,我們發現經過LN的處理,原本距離差異小的信號其差異程度會被縮減 ,而對於原本差異較大的信號,則頃向於放大其差異性。
    By using single neuron level resolution Drosophila antennal lobe local interneuron data, we process a series of statistical analysis to obtain the characteristics of the local neuron network, then to build a virtual artificial neuron network of Drosophila antennal lobe, which is able to be used for certain kinds of computer simulation experiments. Different from some theoretical research that to build totally random artificial networks, by the reason lack of detailed picture of its neural circuit, our research is based on real neuron network data. Since the LN network data resolution is so accurate, reached single cell definition, our statistical analysis is so detailed which enable us to build up virtual artificial networks, that coincide all the features of the real LN data. Our preliminary computer simulation model is able to be used for reproducing many recently published experiments results, such as lateral inhibition (Olsen and Wilson, 2008b; Olsen et al., 2010), contrast enhancement (Laurent, 2002), sparse inhibition (Ghatpande, 2009; Fantana et al.,2008), gain control and divisive normalization (Olsen and Wilson, 2008b; Olsen et al., 2010), and coding modulation (Koulakov et al., 2007). Especially, when talk about odor distinguish, we have revealed something exciting, that is, by the analysis of Euclidean distance, LN tend to process odor information in an interesting way: if the ORN responses are similar, LN tend to make their patterns even more similar; if the ORN response are kind of different, LN tend to enlarge the difference.
    URI: http://nthur.lib.nthu.edu.tw/dspace/handle/987654321/79356
    Source: http://ndltd.ncl.edu.tw/cgi-bin/gs32/gsweb.cgi?o=dnclcdr&s=id=%22099NTHU5291005%22.&searchmode=basic
    Appears in Collections:[系統神經科學研究所] 博碩士論文

    Files in This Item:

    File Description SizeFormat
    GH029780598.pdf285KbAdobe PDF339View/Open


    SFX Query


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback