LVP extraction and triplet-based segmentation for diabetic retinopathy recognition
Till now, the detection of diabetic retinopathy seems to be one of the sensitive research topics since it is related to health care of any individual. A number of contributions in terms of detection already exists in the dice; still, there present some problems regarding the detection accuracy. This issue motivates to develop a new detection model of diabetic retinopathy, and moreover, this model tells the severity of retinopathy from the given fundus image. The proposed model includes preprocessing, segmentation, feature extraction and classification stages. Here, Triplet Half band Filterbank (THFB) Segmentation is performed, local vector pattern (LVP) is used for extracting the features, principle component analysis (PCA) procedure is used to reduce the dimensions of the feature vector, and neural network (NN) is used for classification purpose. The proposed model compares its performance over other conventional classifiers like support vector machine (SVM), k nearest neighbor (k-NN) and Navies Bayes (NB) in terms of positive and negative measures. The positive measures are accuracy, specificity, sensitivity, precision, negative predictive value (NPV), F1-Score and Matthews Correlation Coefficient (MCC). Similarly, the negative measures are the false positive rate (FPR), false negative rate (FNR) and false discovery rate (FDR), and the efficiency of the proposed model is proven.
KeywordsDiabetic retinopathy THFB segmentation Local vector pattern Feature extraction Neural network classification
We acknowledged our sincere thanks to Dr. Amol D Rahulkar, National Institute of Technology, Goa and Pimpri Chinchwad Education Trust’s Pimpri Chichwad College of Engineering & Research, Ravet, Pune for their encouragement and valuable support during this research work.
- 6.Molven A, Ringdal M, Nordbø AM, Raeder H, Støy J, Lipkind GM, Steiner DF, Philipson LH, Bergmann I, Aarskog D, Undlien DE, Joner G, Søvik O; Norwegian Childhood Diabetes Study Group, Bell GI, Njølstad PR (2008) Mutations in the insulin gene can cause MODY and autoantibody-negative type 1 diabetes., Diabetes 57(4):1131–1135CrossRefGoogle Scholar
- 25.Usman M, Akram, Shoab A, Khan (2017) Automated detection of dark and bright lesions in retinal images for early detection of diabetic retinopathy. J Med Syst 36(5):3151–3162Google Scholar
- 27.Hung TY, Fan KC (2014) Local vector pattern in high-order derivative space for face recognition. In: 2014 ieee international conference on image processing (ICIP), Paris, pp. 239–243Google Scholar
- 29.Mohan Y, Chee SS, Xin DKP, Foong LP (2016) Artificial neural network for classification of depressive and normal in EEG. In: 2016 IEEE EMBS conference on biomedical engineering and sciences (IECBES)Google Scholar
- 30.Kaur R, Kaur S (2016) Comparison of contrast enhancement techniques for medical image. In: 2016 conference on emerging devices and systems (ICEDSS), Namakkal, pp. 155–159Google Scholar
- 39.Kota PN, Gaikwad AN (2017) Optimized scrambling sequence to reduce Papr in space frequency block codes based MIMO-OFDM system. J Adv Res Dyn Control Syst 502–525Google Scholar
- 42.Bramhe SS, Dalal A, Tajne D, Marotkar D (2015) Glass shaped antenna with defected ground structure for cognitive radio application. In: International conference on computing communication control and automation, Pune, pp. 330–333Google Scholar
- 46.Wagh AM, Todmal SR (2015) Eyelids, eyelashes detection algorithm and hough transform method for noise removal in iris recognition. Int J Comput Appl 112(3):28–31Google Scholar
- 57.Abdillah B, Bustamam A, Sarwinda D (2017) Classification of diabetic retinopathy through texture features analysis. In: 2017 International conference on advanced computer science and information systemsGoogle Scholar
- 58.Sarwinda D, Bustamam A, Arymurthy AM (2017) Fundus image texture features analysis in diabetic retinopathy diagnosis. In: 2017 eleventh international conference on sensing technology (ICST), Sydney, NSW, pp. 1–5Google Scholar