Authors:
Francisco Peñaranda
1
;
Fernando López-Mir
1
;
Valery Naranjo
1
;
Jesús Angulo
2
;
Lena Kastl
3
and
Juergen Schnekenburger
3
Affiliations:
1
Universitat Politècnica de Valencia, Spain
;
2
MINES ParisTech, France
;
3
University of Muenster, Germany
Keyword(s):
FTIR-spectroscopy, Hyperspectral Imaging, Dissimilarity Measures, Clustering, Cancer.
Related
Ontology
Subjects/Areas/Topics:
Applications and Services
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Computer Vision, Visualization and Computer Graphics
;
Medical Image Detection, Acquisition, Analysis and Processing
Abstract:
In this work, different combinations of dissimilarity coefficients and clustering algorithms are compared in order to separate FTIR data in different classes. For this purpose, a dataset of eighty five spectra of four types of sample cells acquired with two different protocols are used (fixed and unfixed). Five dissimilarity coefficients were assessed by using three types of
unsupervised classifiers (K-means, K-medoids and Agglomerative Hierarchical Clustering). We introduce in particular a new spectral representation by detecting the signals´ peaks and their corresponding dynamics and widths. The motivation of this representation is to introduce invariant properties with respect to small spectra shifts or intensity variations. As main results,
the dissimilarity measure called Spectral Information Divergence obtained the best classification performance for both treatment protocols when is used over the proposed spectral representation.