University of Twente Student Theses


Increasing cancer cell recognition with Raman microscopic data using sparse coding

Loohuis, Pascal (2018) Increasing cancer cell recognition with Raman microscopic data using sparse coding.

[img] PDF
Abstract:Traditional methods of research on cancer cells are done via tissue biopsy. Due to the fact that these biopsies are poorly able to predict the treatment response, other research methods are investigated to eventually replace tissue biopsies. One method is performing research on circulating tumor cells from the blood stream, whereas Raman microscopic techniques are used to distinguish different sorts of cancer. This data is used to obtain a fingerprint per sort cancer by classifying the data. Principle component analysis (PCA) is used in order to make this hyperspectral data insightful. Data often contains nonlinear statistical dependencies, so it is questionable if PCA is the right method to use. This report introduces two other methods, based on sparse coding, that tackles this shortcoming of PCA. In sparse coding a signal is decomposed in a multiplication between a set of basis vectors and a sparse matrix, whereas each pixel of the hyperspectral data will be described with only a few of these basis vectors. The introduced methods proved to give good classifications and were noise resilient.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:31 mathematics
Programme:Applied Mathematics BSc (56965)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page