Laser-induced breakdown spectroscopy: a tool for real-time, in vitro and in vivo identification of carious teeth

Laser-induced breakdown spectroscopy: a tool for real-time, in vitro and in vivo identification of carious teeth
Received October 17, 2001; Accepted December 19, 2001.
Published online 2001 December 19
Ota Samek,1 Helmut H Telle,2 and David CS Beddows3
BMC Oral Health. 2001
PubMed Central

Copyright © 2001 Samek et al; licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original URL.

1Institute of Physical Engineering, Technical University Brno, Technick? 2, 616 69 Brno, Czech Republic
2Department of Physics, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, United Kingdom
3Department of Chemistry, The University of Edinburgh, West Main Roads, Edinburgh EH9 3JJ, United Kingdom
Corresponding author.
Ota Samek: samek@ufi.fme.vutbr.cz; Helmut H Telle: h.h.telle@swansea.ac.uk; David CS Beddows: pybeds@hotmail.com

Abstract
Background
Laser Induced Breakdown Spectroscopy (LIBS) can be used to measure trace element concentrations in solids, liquids and gases, with spatial resolution and absolute quantifaction being feasible, down to parts-per-million concentration levels. Some applications of LIBS do not necessarily require exact, quantitative measurements. These include applications in dentistry, which are of a more "identify-and-sort" nature Ò e.g. identification of teeth affected by caries.

Methods
A one-fibre light delivery / collection assembly for LIBS analysis was used, which in principle lends itself for routine in vitro / in vivo applications in a dental practice. A number of evaluation algorithms for LIBS data can be used to assess the similarity of a spectrum, measured at specific sample locations, with a training set of reference spectra. Here, the description has been restricted to one pattern recognition algorithm, namely the so-called Mahalanobis Distance method.

Results
The plasma created when the laser pulse ablates the sample (in vitro / in vivo), was spectrally analysed. We demonstrated that, using the Mahalanobis Distance pattern recognition algorithm, we could unambiguously determine the identity of an "unknown" tooth sample in real time. Based on single spectra obtained from the sample, the transition from caries-affected to healthy tooth material could be distinguished, with high spatial resolution.

Conclusions
The combination of LIBS and pattern recognition algorithms provides a potentially useful tool for dentists for fast material identification problems, such as for example the precise control of the laser drilling / cleaning process.

Complete article may be viewed online.