A Decision System for Predicting Diabetes using Neural Networks

K. Chandana Rani, Y. Prasanth


Diabetic retinopathy (DR) is an eye fixed ill complete by the impairment of polygenic disorder and that we purchased to acknowledge it before of calendar for sensible treatment. On these lines, 2 social occasions were perceived, specifically non-proliferative diabetic retinopathy (NPDR), proliferative diabetic retinopathy (PDR). During this paper, to dissect diabetic retinopathy, 3 models like Probabilistic Neural framework (PNN), Bayesian Classification and Support vector machine (SVM) square measure pictured and their displays square measure thought-about. The live of the unwellness unfold within the membrane are often recognized by analytic the elements of the membrane. The elements like veins, hemorrhages of NPDR image and exudates of PDR image square measure off from the unrefined photos victimization the icon prepare techniques, fed to the classifier for gathering a complete of 350 structure photos were used, out of that100 were used for designing and 250 pictures were used for testing. Exploratory results show that PNN has an accuracy of 89.6 % Bayes Classifier incorporates a exactness of 94.4% and SVM has an exactitude of 97.6%. What is more our system is equally continue running on 130 pictures open from "DIARETDB0: Evaluation Database and Procedure for Diabetic Retinopathy" and also the results show that PNN incorporates a exactness of 87.69% Bayes Classifier has an accuracy of 90.76% and SVM has a precision of 95.38%.


Accuracy Probabilistic neural network, Bayesian classification, Sensitivity, Specificity, Support vector machine

Full Text:


DOI: http://doi.org/10.11591/ijai.v6.i2.pp56-65


  • There are currently no refbacks.

View IJAI Stats

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.