M. M. C. Ferreira, S. P. Gurden, V. F. Monteiro, F. A. L. Ribeiro, E. Longo, "Quantitative Analysis and Classification of AFM Images of Human Hair". Lisbon, Portugal, 20-23/09/2004: Ninth International Conference on Chemometrics in Analytical Chemistry (CAC IX, CAC-2002). Program and Abstracts Book, (2004). Oral 186.
QUANTITATIVE ANALYSIS
AND CLASSIFICATION OF AFM IMAGES OF HUMAN HAIR
MÁRCIA M. C. FERREIRA, STEPHEN P. GURDEN, VALÉRIA F. MONTEIRO, FABIANA A. L. RIBEIRO, ELSON LONGO.
State Univ. of Campinas, Instituto de Química, CAMPINAS, 13083-970, BRAZIL, Phone: 55 19 3788-3102, Fax: 55 19 3788-3023
ABSTRACT
Quantitative Analysis and Classification of AFM Images of Human Hair. Márcia M. C. FERREIRA1*, Stephen P. GURDEN1, Valéria F. MONTEIRO2, Fabiana A. L. RIBEIRO1, Elson LONGO2. *marcia@iqm.unicamp.br 1Instituto de Química, Universidade Estadual de Campinas (UNICAMP), 13083-970, Campinas, SP, Brazil. 2Laboratório Interdisciplinar de Eletroquímica e Cerâmica , Universidade Federal de São Carlos (UFSCar), 13565-905, São Carlos - SP, Brazil. The central cortex of a hair fibre is surrounded by thin cellular sheets, known as cuticles, which overlap each other from root to tip. These cuticles fulfil a number of useful roles including protection from physical and chemical insult and a tendency to maintain the hair in a clean and disentangled state. The condition of the cuticles is of great cosmetic importance, but also has the potential to provide correct diagnoses in the medical and forensic sciences.In this work, atomic force microscopy (AFM) is used for the analysis of the hair surface of 38 samples (root and distal ends of untreated and bleached hair samples), with the goal of building an algorithm capable of analysing and classifying the images automatically. The quantitative analysis and classification of AFM hair images was carried out by characterizing the hair surface using a series of descriptors which summarize the important characteristics of the cuticular structure. Partial least squares discriminant analysis (PLS-DA) is used to test the ability of the algorithm to characterize the images according to the properties of the hair samples. Tucker models are also built on image and descriptors data.
Keywords: human hair, atomic force microscopy, partial least squares, discriminant analysis, classification
Requested contribution format: oral
Corresponding author: MÁRCIA M. C. FERREIRA (marcia@iqm.unicamp.br)