Multi-input 2-dimensional deep belief network: diabetic retinopathy grading as case study
- PDF / 1,262,119 Bytes
- 16 Pages / 439.37 x 666.142 pts Page_size
- 43 Downloads / 163 Views
Multi-input 2-dimensional deep belief network: diabetic retinopathy grading as case study Amirali Amini Tehrani 1 & Ali Mohammad Nickfarjam 1 & Hossein Ebrahimpour-komleh 1 & Dawood Aghadoost 2 Received: 9 August 2019 / Revised: 27 August 2020 / Accepted: 6 October 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
The most important action in treating diabetic retinopathy is early diagnosis and its progression degree. This paper presents a two-dimensional Deep Belief Network based on Mixed-restricted Boltzmann Machine capable of receiving multiple two-dimensional inputs. Using multiple inputs provides more appropriate prior information for learning. In this proposed method, the image is transferred to the HSV color space and then the 3D color image is converted to a 2D matrix using a weighted mean. This weighted mean is calculated based on the entropy criterion. The resulting two-dimensional matrix is not in pixel and is merely a raw description of the image. The local, regional and global descriptions are extracted from this matrix and provided for the network. The proposed deep network automatically extracts the appropriate features to determine the progression degree of diabetic retinopathy by the network. Window by window image processing can overcome one of the basic problems of image classification, i.e. the small number of labeled data. Experiments showed that the proposed method is superior when compared to other methods. Keywords Diabetic retinopathy . Mixed-restricted Boltzmann machine . Retinal image . Deep networks . Multi-input 2-dimensional deep belief network
* Hossein Ebrahimpour-komleh [email protected] Amirali Amini Tehrani [email protected] Ali Mohammad Nickfarjam [email protected] Dawood Aghadoost [email protected] Extended author information available on the last page of the article
Multimedia Tools and Applications
1 Introduction During the last century, diabetes has been developing and affecting not only the eyes but also other organs of the body. Generally, eye lesions can be detected because of observing the inner space of the eye by using appropriate devices. The small veins of the body are the effect point of diabetes in such a way that the wall of the small veins becomes thin and loses its resistance, resulting in destructing veins. The high level of blood sugar in diabetes causes the sugar to deposit in the retinal vein and the lens to swell, affecting the patient’s ability to observe things [28]. Blurry vision can be considered as a sign of a serious ocular problem caused by diabetes, in which cataracts, glaucoma, and diabetic retinopathy are among the three major ocular problems of people with diabetes [15]. Retinal damage caused by diabetes is called diabetic retinopathy, which is a general term used to express vascular problems in the retina of diabetic patients [28]. Developing this complication in diabetic patients can lead to blindness in the case of no timely diagnosis. The treatment process of diab
Data Loading...
