Document Detail

Fourier-transform infrared spectroscopy coupled with a classification machine for the analysis of blood plasma or serum: a novel diagnostic approach for ovarian cancer.
MedLine Citation:
PMID:  23325355     Owner:  NLM     Status:  Publisher    
Currently available screening tests do not deliver the required sensitivity and specificity for accurate diagnosis of ovarian or endometrial cancer. Infrared (IR) spectroscopy of blood plasma or serum is a rapid, versatile, and relatively non-invasive approach which could characterize biomolecular alterations due to cancer and has potential to be utilized as a screening or diagnostic tool. In the past, no such approach has been investigated for its applicability in screening and/or diagnosis of gynaecological cancers. We set out to determine whether attenuated total reflection Fourier-transform IR (ATR-FTIR) spectroscopy coupled with a proposed classification machine could be applied to IR spectra obtained from plasma and serum for accurate class prediction (cancer vs. normal). Plasma and serum samples were obtained from ovarian cancer cases (n = 30), endometrial cancer cases (n = 30) and non-cancer controls (n = 30), and subjected to ATR-FTIR spectroscopy. Four derived datasets were processed to estimate the real-world diagnosis of ovarian and endometrial cancer. Classification results for ovarian cancer were remarkable (up to 96.7%), whereas endometrial cancer was classified with a relatively high accuracy (up to 81.7%). The results from different combinations of feature extraction and classification methods, and also classifier ensembles, were compared. No single classification system performed best for all different datasets. This demonstrates the need for a framework that can accommodate a diverse set of analytical methods in order to be adaptable to different datasets. This pilot study suggests that ATR-FTIR spectroscopy of blood is a robust tool for accurate diagnosis, and carries the potential to be utilized as a screening test for ovarian cancer in primary care settings. The proposed classification machine is a powerful tool which could be applied to classify the vibrational spectroscopy data of different biological systems (e.g., tissue, urine, saliva), with their potential application in clinical practice.
Ketan Gajjar; Júlio Trevisan; Gemma Owens; Patrick J Keating; Nicholas J Wood; Helen F Stringfellow; Pierre L Martin-Hirsch; Francis L Martin
Related Documents :
23291295 - Validation of a one-page patient-reported charlson comorbidity index questionnaire for ...
11022775 - Histologic response to intraligament injections using a computerized local anesthetic d...
24221535 - Refinement of the prediction of n-acetyltransferase 2 (nat2) phenotypes with respect to...
24825695 - Case-based visualization of a patient cohort using seer epidemiologic data.
9587405 - Functional consequences of mutations in ser-52 and ser-60 in human blood coagulation fa...
1423825 - Clinical significance of prostacyclin and thromboxane in cancer of the female breast an...
Publication Detail:
Type:  JOURNAL ARTICLE     Date:  2013-1-17
Journal Detail:
Title:  The Analyst     Volume:  -     ISSN:  1364-5528     ISO Abbreviation:  Analyst     Publication Date:  2013 Jan 
Date Detail:
Created Date:  2013-1-17     Completed Date:  -     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  0372652     Medline TA:  Analyst     Country:  -    
Other Details:
Languages:  ENG     Pagination:  -     Citation Subset:  -    
Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster LA1 4YQ, UK.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

Previous Document:  Early estimates of seasonal influenza vaccine effectiveness - United States, january 2013.
Next Document:  Advances in QSPR/QSTR models of ionic liquids for the design of greener solvents of the future.