It is widely accepted that the cortex can be divided into a series of spatially discrete areas based on their specific laminar patterns. It is of great interest to divide the cortex into different areas in terms of both neuronal functions and cellular composition. The division of cortical areas can be reflected by the cell arrangements or cellular composition. Therefore, the cortical structure can be represented by some functional neuronal density data. Techniques on functional data analysis help to develop some measures which indicate structural changes. / In order to separate roughness from structural variations and influences of the convolutions and foldings, a method called bivariate smoothing is proposed for the noisy density data. This smoothing method is applied to four sets of cortical density data provided by Prof Petrides [1] and Scott Mackey [2]. / The first or second order derivatives of the density function reflect the change and the rate of the change of the density, respectively. Therefore, derivatives of the density function are applied to analyze the structural features as an attempt to detect indicators for boundaries of subareas of the four cortex sections. / Finally, the accuracy and limitation of this smoothing method is tested using some simulated examples.
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.98531 |
| Date | January 2005 |
| Creators | Zhang, Wen, 1978- |
| Publisher | McGill University |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Format | application/pdf |
| Coverage | Master of Science (Department of Psychology.) |
| Rights | © Wen Zhang, 2005 |
| Relation | alephsysno: 002479167, proquestno: AAIMR24839, Theses scanned by UMI/ProQuest. |
Page generated in 0.0018 seconds