Chem. J. Chinese Universities ›› 2009, Vol. 30 ›› Issue (8): 1543.

• Articles • Previous Articles     Next Articles

Near Infrared Spectroscopy Combined with Principal Component Analysis Applied to Diagnosis of Endometrial Carcinoma

XU Ke1, XIANG Yu-Hong1, DAI Yin-Mei2, ZHANG Zhuo-Yong1*   

  1. 1. Department of Chemistry, Capital Normal University, Beijing 100048, China;
    2. Beijing Obstetrics and Gynecology Hospital, Capital Medical University, Beijing 100006, China
  • Received:2009-03-18 Online:2009-08-10 Published:2009-08-10
  • Contact: ZHANG Zhuo-Yong. E-mail: gusto2008@vip.sina.com
  • Supported by:

    国家自然科学基金(批准号: 20875065, 30772322)资助.

Abstract:

The feasibility of near infrared(NIR) spectral feature extraction and early diagnosis of endometrial carcinoma were developed by chemometrics methods. Few papers have been reported about the studies on NIR spectra features of endometrial carcinoma so far. In this study, the NIR spectra of 154 specimens of endomet-rium were collected, and spectral data were analyzed by principal component analysis. In order to improve the classification, selection of wavelength range and spectral pretreatment methods were discussed. The results suggested that samples of malignant, hyperplasia and normal endometrium were classified correctly. Moreover, endometrium at various differentiation stages could also be identified, which provided reliable evidence for early diagnosis in endometrial carcinoma. This approach was proved to be rapid and convenient, which is able to be developed as a non-invasive diagnosis method for cancer.

Key words: Endometrial carcinoma; Near infrared spectroscopy; Principal component analysis

TrendMD: