Pattern recognition technique in immunological antigenic tests to identify Mycobacterium tuberculosis infection

Selvaraj, R and Gopal, G and Kumaraswami, V (2002) Pattern recognition technique in immunological antigenic tests to identify Mycobacterium tuberculosis infection. Tuberculosis, 82 (6). pp. 261-266. ISSN 1472-9792

[thumbnail of 200233.pdf] PDF
200233.pdf - Published Version
Restricted to Registered users only

Download (120kB) | Request a copy


Theimportance ofadiagnostic test that issimpleandquick toidentifyMycobacteriumtuberculosisinfectionneedsno emphasis.The tuberculin skin test (TST --1TURT23) is the diagnostic tool for identifyingM. tuberculosis infection at present.The test reactiononthe skinismeasuredafter 48--72 h.It isobservedthat oftenmulti-modesare seen,whenthereactionsare drawnas agraphandthebimodalityisseenvery feebly.Because of the difficultiesintheadministrationof TST, severalserological testswere developed over three decades, but none of the studies showed the desired results.One study to evaluate purified protein derivative (PPD) antigen resulted in a claimof 80%sensitivityand 4%false-positivity rate (14), while other researcherswere not able to obtain similar results. In addition, severalproblemswere encountered due to the non-availabilityof antigens, and data analysesfroman ELISA-based diagnostic test showed considerable overlap of distributionsofopticaldensity (OD) values among patientsandhealthyindividuals (10).Classicalstatistical techniques cannot explainthe cause of these overlaps.Hence, anattempt ismadeinthis article to resolve these difficultiesby the patternrecognitiontechnique (PRT).The techniqueliesin splittingthe data into clusters using a supervised algorithm.The data set isnormally split into a training set, a test set and a validation set.The PRT gets‘‘trained’’ through the training data set until the infected and uninfected groups of individuals are correctly classified.The training occursbased on an algorithmon the training set.On successful completion of the training, this technique is further tested andvalidated in the respective data sets. Setting:Atotal of 273f inger-prick specimenswere collected fromfive categories (Al,A2,B,C,D,E) of subjectsnot vaccinatedwith (bacille Calmette Guerin) (BCG) fromTrivellore BCGTrial area adopted by theTuberculosis Research Center,Chennai, India. Objective: The study was conductedwith the primary aimof evaluating purified antigens -- r38kDa,PPD and 30kDa -- for their usefulness as diagnostic tools and to test the applicabilityof the PRT in the evaluation of diagnostic tests. Individuals in two main categories (definitely not infected categories Al,A2 and D, and definitely infected categories B,E and C based on reaction toTST) were assembled for the purpose. Results: The overall PRT performance of 30kDa was 72.3% sensitivity and 90.9% specificity for identifying M. tuberculosis infection, while the r38kDa antigen recorded a sensitivityof 73.8%and a specificityof 84.6%. In the case of PPD, the results were not promising. Conclusion: This paperon ELISA-based diagnostic tests attempts to implement an optimal decision support system through PRT that would identify the outcome (as infected or non-infected) based on the OD values.The PRTwas able to predict the outcome for individual suspects. Further, Kullback--Leibler distance measurement has validated the PRT in distinguishing infected individuals fromhealthy subjects (based on the OD values).

Item Type: Article
Subjects: Tuberculosis > Biostatistics
Divisions: Statistics
Depositing User: Dr. Rathinasabapati R
Date Deposited: 11 Nov 2013 10:52
Last Modified: 10 Mar 2016 04:46

Actions (login required)

View Item View Item