ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login


Journal: International Journal of Advanced Research in Engineering and Technology (IJARET) (Vol.11, No. 11)

Publication Date:

Authors : ;

Page : 937-958

Keywords : Multi-retinal disease; Image classification; Machine learning; Feature extraction;

Source : Downloadexternal Find it from : Google Scholarexternal


In past decades, retinal diseases become more common and affects people at all age grounds over the globe. The retinal disease affects the eye vision owing to the blurred visualization and occasionally a whole vision loss. The conventional eye disease diagnosis models are found to be a time consuming task and it require expert's knowledge. For examining the retinal eye disease, an artificial intelligence (AI) based multilabel classification model is needed for automated diagnosis. To analyze the retinal malady, the system proposes a multiclass and multi-label arrangement method. This paper presents new handcrafted features with machine learning (ML) based classification model for multi-retinal disease diagnosis. The proposed method involves pre-processing different ways such as image resizing, noise removal, and contrast improvement. In addition, the presented model involves two handcrafted feature extraction techniques such as scale-invariant feature transform (SIFT) and Speeded-Up Robust Features (SURF). Besides, two ML based classification models like gradient boosting tree classifier (GBC) and random forest classifier (RFC) models are used to identify multi-retinal diseases. The performance validation of the presented model takes place on a benchmark multi-retinal disease dataset, comprising data instances from different classes. Among the different presented models, the SURF-GBC model has demonstrated proficient diagnostic performance with the higher accuracy of 92.72%, precision of 93.42%, recall of 92.28%, and F1-score of 92.76%.

Last modified: 2021-02-22 19:13:46