Diffuse reflectance spectroscopy: towards clinical application in breast cancer
- 394 Downloads
Diffuse reflectance spectroscopy (DRS) is a promising new technique for breast cancer diagnosis. However, inter-patient variation due to breast tissue heterogeneity may interfere with the accuracy of this technique. To tackle this issue, we aim to determine the diagnostic accuracy of DRS in individual patients. With this approach, DRS measurements of normal breast tissue in every individual patient are directly compared with measurements of the suspected malignant tissue. Breast tissue from 47 female patients was analysed ex vivo by DRS. A total of 1,073 optical spectra were collected. These spectra were analyzed for each patient individually as well as for all patients collectively and results were compared to the pathology analyses. Collective patient data analysis for discrimination between normal and malignant breast tissue resulted in a sensitivity of 90 %, a specificity of 88 %, and an overall accuracy of 89 %. In the individual analyses all measurements per patient were categorized as either benign or malignant. The discriminative accuracy of these individual analyses was nearly 100 %. The diagnosis was classified as uncertain in only one patient. Based on the results presented in this study, we conclude that the analysis of optical characteristics of different tissue classes within the breast of a single patient is superior to an analysis using the results of a cohort data analysis. When integrated into a biopsy device, our results demonstrate that DRS may have the potential to improve the diagnostic workflow in breast cancer.
KeywordsBreast cancer Diffuse reflectance spectroscopy Diagnosis Individual analysis
We would like to thank all colleagues of the NKI pathology department and Philips Research project members for their contribution in the optical data collection and assessment. In particular we would like to thank W. Bierhoff for the probe development and J. Horikx for the console development. This study was supported by Philips Research, Eindhoven, The Netherlands.
Conflicts of interest
None of the authors who are affiliated with clinical institutions (DE, MVP, JH, HO, ER, JW and TR) have financial interests in the subject matter, materials, or equipment or with any competing materials. These authors received no payment for any kind for their participation in this research project, nor did their institutions receive payment for anything beyond the direct costs of performing this research project at The Netherlands Cancer Institute, NKI-AVL Amsterdam. Their interests are purely at a scientific level. All of the authors who are affiliated with Philips Research and Philips Healthcare (RN, GL and BH) have financial interests in the subject matter, materials, and equipment, in the sense that they are employees of Philips. The prototype system described in this article is currently only a research prototype and is not for commercial use. It is the intention of Philips to develop the prototype system into a commercial system that would be sold by Philips.
- 10.Kennedy S, Geradts J, Bydlon T, Brown J, Gallagher J, Junker M, Barry W, Ramanujam N, Wilke L (2010) Optical breast cancer margin assessment: an observational study of the effects of tissue heterogeneity on optical contrast. Breast Cancer Res Treat 12(6):R91Google Scholar
- 20.Nachabé R, Evers D, Hendriks B, Lucassen G, van der Voort M, Rutgers E, Vrancken Peeters M, van der Hage J, Oldenburg H, Wesseling J, Ruers T (2011) Diagnosis of breast cancer using diffuse optical spectroscopy from 500 to 1600 nm: a comparison of classification methods. J Biomed Opt 16(8):087010PubMedCrossRefGoogle Scholar
- 21.Nachabé R, Evers D, Hendriks B, Lucassen G, van der Voort M, Wesseling J, Ruers T (2011) Effect of bile absorption coefficients on the estimation of liver tissue optical properties and related complications in discriminating healthy and tumorous samples. Biomed Optics Express 2(3):600–614CrossRefGoogle Scholar
- 24.Breiman L, Friedman J, Olshen R, Stone C (1984) Classification and regression trees. Statistics/probability series. Wadsworth Publishing Company, Belmont, CAGoogle Scholar
- 26.Zhu C, Breslin T, Harter J, Ramanujam N (2008) Model based and empirical spectral analysis for the diagnosis of breast cancer. Biomed Opt Express 16(19):14961–14978Google Scholar