Automatic Nevirapine concentration interpretation system using support vector regression
Computer Methods and Programs in Biomedicine
| Hi-index | 0.00 |
The traditional quantitative methods formeasurement of 驴-fetoprotein (AFP) areradioimmunoassay (RIA) and nonisotopicimmunoassay. These immunoassay methods have theadvantages of high sensitivity and accuracy. The time-consumingand complicated laboratory operations aredisadvantages of these methods. On the contrary, therapid gold immuno-chromatographic dipstick assayhas the advantages of simple operation proceduresand rapid operation time. But most of the dipstickimmunoassays is qualitative or semi-quantitative.This paper presents a new quantitative assessmentsystem of serum AFP by combining the rapid goldimmuno-chromatographic dipstick assay and neuralnetwork based computer image analysis system. A totalof 72 human serum samples, in which the AFPconcentrations were measured by RIA as 1 to 700ng/ml, are used to construct the neural network forquantitative computer image analysis system in thisstudy.In this paper, we also correlated the proposedsystem results with a commercial AFP specific RIA testkit. Another 30 human serum samples were used tovalidate the new system. The result of the new systemcorrelated well with traditional RIA with correlationcoefficient 0.9971. The Bland-Altman Plot also showswell agreements between the proposed system andRIA.We successfully established a rapid quantitativeassessment system for AFP dipstick assay, which tooksignificantly less time in operation, with clinicalacceptable agreements by Bland-Altman Plot analysis.This novel method is simple, rapid, and can beperformed outside the laboratory. The rapid goldimmuno-chromatographic dipstick assay quantified bycomputer image analysis has been termed DipstickComputer Image Processing immunoAssay (DCIPA).