Aim: The Novel Health Disorder Discrimination system is developed for discrimination of diseases in human detected by blood cell using image processing of convolutional neural networks algorithm over support vector machine. Materials & Methods: By using convolutional neural networks algorithm over support vector machine algorithm in image processing, diseases like malaria, pneumonia, breast cancer, skin cancer were identified on dataset having sample size of 80 each group and software tool python opencv and jupyter notebook is used for running application and accuracy scoring respectively. Accuracy values for identification of diseases are calculated to quantify performance of convolutional neural networks algorithm against support vector machine algorithm with t-test analysis. Results and Discussion: The analysis on trained dataset and test dataset has been performed successfully using SPSS and acquired 95% accuracy for CNN algorithm compared to support vector machine method, which gave 90% accuracy with level of significance (p<0.05). The resultant data depicts reliability in independent sample tests. Conclusion: On the whole process of prediction of accuracy by image processing, CNN model gives significantly better performance compared with support vector machine model.