https://doi.org/10.31449/inf.v45i2.3556 Informatica 45 (2021) 621–629 621 
 
An Intelligent Information Management System for Retinal Image 
Storage and Recognition in Chronic Disease Using Digital Signal and 
Image Processing  
 
Xiaojun Xu, Zhong Tang and Xijia He 
Guangxi Medical University, Country China 
E-mail: xuxiaojun51@gmail.com, ztang829@gmail.com, xijiahe5@gmail.com 
 
Haixuan Wang 
School of Management, Furen University, Xinbei, Taiwan 242304 
E-mail: haixuanwang@outlook.com 
 
Anil Sharma 
Department of computer science, Faculty of Technology, Debre Tabor University, Ethiopia 
E-mail: anilsharma@dtu.edu.et 
 
Keywords: retinal imaging, retinal vasculature changes, chronic disease management, chronic disease information 
system, retinal blood vessels 
 
Received: May 21, 2021 
Retinal imaging plays a very significant role in the study of retinal vasculature changes indicative of 
chronic disease information related to vision. This study investigates the invulnerability of retinal image 
information in the disease information system and assesses the quantitative method of the morphological 
changes in the retinal vascular network. In this work, the medical digital image transmission protocol 
Digital Imaging & Communications in Medicine (DICOM) version 3.0 and the retinal image Picture 
Archiving & Communication System (PACS) were constructed in the laboratory using browser/server 
mode. Also, the DICOM-SR document was designed in this article using a list or hierarchy, and the 
retinal images to report the information of patients by using the Hypertext Transfer Protocol (HTTP) – 
based Web Access to DICOM Persistent Objects (WADO) approach. The results showed that the retinal 
image PACS system constructed in Browser/Server mode can effectively store and transmit DICOM 
images. When the imaging device is combined with the application program, special adapters are used 
to negotiate the transmission syntax. The message flow is decoded in the communication process, which 
can be connected with the realization to improve the efficiency of information collection. The proposed 
PACS system integrates the quantitative features of retina providing more meaningful research data for 
data mining in comparison to the traditional state of the art methods based on chronic disease 
management system. The diagnostic ability of the retinal imaging procedure using the DICOM images is 
justified by obtaining 98.51%, 98.04%, 99% and 99.01% of accuracy, sensitivity, specificity and 
precision values respectively.  
Povzetek: V prispevku je predstavljen protokol in sistem za shranjevanje slik očesnega ozadja za 
medicinske namene. 
 
1 Introduction 
Clinical retinal digital images are usually obtained by 
digital ophthalmoscopy, which is taken by a special 
camera through the pupil. At present, many ophthalmic 
image management systems mainly use fundus 
workstation equipment for collection and analysis, and 
most of the images are implemented in JPEG and other 
formats, which cannot be shared in hospitals, greatly 
limiting the analysis and utilization of retinal image data 
[1, 2]. The conventional ophthalmic retinal imaging 
system illustrated in Figure 1 shows that the incident 
light from the retinal fundus possesses somewhat 
distorted wavefront which is further subjected to the 
adaptive mirror which is connected to the control system. 
Then the beam-splitter splits the waveform into two: one 
is sent to the waveform sensor and the other one is the 
corrected waveform obtained that results into a high-
resolution retinal image which can be stored for future 
usage [3,4].  
A basic schematic diagram of ophthalmic retinal 
imaging system is presented in Figure 1. 

622 Informatica 45 (2021) 621–629 X. Xu et al. 
 
 
 
Figure 1: Schematic diagram of ophthalmic retinal 
imaging system. 
 
The effectual retinal image management and storage 
is effective for the recognition of chronic diseases like 
Glaucoma, Diabetic retinopathy, macular degeneration 
and many more [5,6]. One of the greatest challenges in 
effective detection and diagnosis of various retinal 
diseases is their asymptomatic nature before it 
approaches the severe chronic disease levels [7,8]. 
However, different researches thrived for devising 
various diagnostic and image recognition systems for 
retinal image classification, management and storage. 
The fundus imaging-based methods are widely being 
used by the technological experts for retinal image 
screening [9]. A latest research in the field of 
classification of diabetic retinopathy severity of retinal 
imaging address the challenges of retinal landmark 
segmentation, retinal discrimination and severity grading 
while employing the hierarchal discrimination 
framework [10].  
In order to solve the problem of registration of 
retinal images with different Angle of view of the retinal 
camera, Zou Beji et al. proposed a local feature 
registration method with affine invariability [11]. Then 
the anisotropic image structure is transformed into 
isotropic image structure with different angle factors by 
rotation, sampling and compression [12]. Finally, the 
scale-invariant feature transformation algorithm was used 
to extract feature points from isotropic image structures 
and match the feature point coordinates [13]. The 
experimental results in the private sample set containing 
83 sets of retinal images showed that the root-mean-
square error of this method was 1.247±0.251 pixels, 
which could well solve the problem of perspective and 
scale change [14]. Liu et al. discussed the methods of 
noise reduction in the spatial and frequency domain of 
OCT images, and summarized the methods of precise 
localization and stratification of each layer of tissue in 
the OCT anterior segment and retina images [15]. In this 
paper, the hierarchical methods based on gray value 
search method, active contour model, graph theory and 
pattern recognition are deeply analyzed, and the 
advantages and disadvantages of existing hierarchical 
methods and existing problems are deeply discussed, and 
corresponding solutions and optimization schemes are 
proposed [16]. To analyze and evaluate the clinical 
diagnostic indicators of ophthalmology-related diseases, 
and further discuss and analyze the development trend 
and level of OCT image processing in the future 
according to the clinical demand of ophthalmology and 
the status quo of OCT image processing [17]. 
The motivation behind this article is to present the 
study based on medical information in the image 
transmission protocol (DICOM 3.0) and open source 
DICOM server which were not addressed in the 
previously presented state-of-the-art schemes. The major 
contributions of this work are: 
• This work utilizes the eye-PACS architecture and 
uses the network Access based on HTTP DICOM 
Persistent objects (Web Access to DICOM Persistent 
Object WADO), which is connected [14] by 
developing a WADO network gateway interface.  
• The main contribution of this study is retinal vascular 
analysis indicating chronic diseases, by investigating 
the invulnerability of retinal image information and 
assessment of quantitative method of the 
morphological change of retinal vascular network.  
• The DICOM-SR document was designed using a list 
or hierarchy, and the retinal images and report 
information of patients were accessed using the 
Hypertext Transfer Protocol (HTTP) – based Web 
Access to DICOM Persistent Objects WADO 
approach.  
• This work significantly contributed in the 
interoperation of retinal image information between 
heterogeneous chronic disease information systems, 
which were not addressed in the previously presented 
related work in this particular domain. 
The rest of this article is organized as: Section 2 
presents the related works in various domains of retinal 
imaging. Section 3 consists of material and methods 
comprising the architectures and flowcharts of designed 
modules. Results, analysis and discussion part is covered 
in section 4 followed by concluding remarks in section 5. 
2 Related work 
Various state-of-the-art work in the field of retinal image 
Web-PACS, DICOM-SR designing, segmentation of 
retinal vasculatures, arteriovenous classification is 
presented in this section.  
2.1 Retinal image web-PACS and design of 
DICOM-SR 
Body circulation diseases often occur with retinal 
vascular changes, the retinal image as a vascular image 
can be obtained in vitro; the test in the diagnosis and 
treatment of diseases such as hypertension, diabetes 
evaluation is of great significance. In diabetic patients 
with retinal every once in a while, for six months or a 
year is about to do a fundus examination, hard physical 
labor took an ophthalmologist. Computer aided diagnosis 
is of great importance to solve this problem, including 
fundus retinal blood vessels is accurate and rapid 
automated segmentation technology. As early as in 1989, 

An Intelligent Information Management System for Retinal Image Storage and ... Informatica 45 (2021) 623–629 623 
 
Chaudhuri et al. proposed a kind of retinal blood vessels 
detection method based on matched filter [18]. Wang and 
others are using multi-resolution Hemiite model to 
modeling and tag vessels [19]. Lam, and others use the 
gradient vector detect and extract the centerline vascular 
structures, further eliminate noise segmentation results 
using the centerline pruning technology. Boykov et al., in 
2004 an interactive image segmentation method based on 
graph cut is proposed in [20].  
The rapid development in artificial intelligence-
based disease management and diagnosis assistant 
methods for standardizing and monitoring the real time 
health for immediate treatment of retinal disorders. 
Various feature-based image registration techniques like 
feature-based image registration utilizing global-local 
matching, model refinement using the hierarchal methods 
and dual- bootstrapping are particularly effective in the 
identification of sufficient number of vasculature features 
for disease discrimination. Various optimized image 
features were estimated for the registration of retinal 
fundus. Powell [21] derived the retinal vessel centre line 
distance for finding out the vasculature branching points 
as a landmark for image correspondence. The approach 
presented in [22] was tested for 462 pairs of the green 
channelled fundus images and reliable outcomes are 
achieved for the retinal image registration based on the 
identified vascular features for 2D vessel segmentation.  
The related work section provides the detailing of 
various retinal image registration, management and 
storage approached for the recognition of chronic 
diseases like Glaucoma, Diabetic retinopathy, macular 
degeneration. The challenges posed by the state-of-the-
art methods reported in the literature lies in effective 
detection and diagnosis of various retinal diseases due to 
the asymptomatic nature of severe chronic disease levels. 
There are various researches being carried out in the field 
of retinal imaging for effectual analysis and timely 
treatment. To solve the problem of registration of retinal 
images with different Angle of view of the retinal 
camera, this work exploits the medical information in 
medical digital image transmission protocol 
(DICOM3.0). This experimentation adopts the model of 
browser/server (B/S) architecture of the retinal image 
PACS system in the laboratory. For retinal images of 
DICOM format, use DICOM gateway after processing 
integrated into PACS. Also, this experiment is to the 
design of DICOM SR document using a list or hierarchy, 
the basic unit is the “item” (the content item). Each item 
includes name/value and relationship using WADO 
manner based on HTTP connections, through the 
development of network gateway with WADO interface. 
WADO requests will be made through the HTTP GET 
method implementation and the users in the image list 
page by clicking on the link. Then send WADO requests 
to the server and the server returned response directly on 
the page. When the user wants to enter the patient list 
JSP view, they just ask the DICOM archive server sends 
the request form, which provides access to the patients 
with retinal image and report information. 
2.2 Retinal vascular segmentation and 
arteriovenous classification 
Retinal blood vessels segmentation is the basis of 
accurate quantitative analysis of retinal vascular. Uneven 
illumination in retinal fundus photography network of 
blood vessels by arteriovenous GuanShu cross makes the 
characteristics of automatic segmentation of retinal 
arteriovenous became very difficult. According to the 
above characteristics, in this topic, we obtain the retinal 
image of uneven background correction, image green 
channel extraction, and the image on the retina of 
discrete wavelet transform to enhance the sampling. By 
extracting pixel gray value, retinal vascular 
neighborhood two-dimensional gray entropy texture 
feature, and by using fuzzy c-means clustering algorithm 
for clustering, extract retinal blood vessels. The 
framework first use morphological image processing on 
the optic disc structure on the retinal image automatic 
segmentation and extraction. Then, different labeled seed 
points were used as the initial points to track the 
segmented vessels of different types, and the retinal 
vascular network was re-labeled to provide a basis for the 
subsequent extraction and analysis of retinal 
arteriovenous morphological features [21]. 
2.3 Quantitative analysis of retinal 
vascular tortuosity 
In clinical ophthalmology work, usually the first change 
of vascular patterns is vascular contorts change. Eye 
doctors commonly used qualitative index to describe the 
change, such as mild circuitry, moderate, severe and 
extreme circuitry gives an index of the severity of 
diabetic retinopathy. To give the characterization, the 
curvature is the main mathematical models of retinal 
blood vessels. This study will retinal vascular centerline 
as a random process, first of all, according to the image 
neighborhood information to extract the blood center line 
and then the vascular segment between two adjacent 
feature points is automatically separated to extract the 
pixel of each pixel on the Central Line of the segment 
coordinate information, and by using the second order 
moments of the random process metrics such as the 
variation of statistical features to reflect the degree of 
change curve, through the simulation curve data and 
actual test on the retinal image, the test after the 
experiment methods for different scale, rotation of 
vascular give stability, and compared with other indices 
like chord is reported in the literature [22]. 
2.4 Quantitative analysis of retinal vessel 
width 
Retinal vessel width and Artery-to-Vein diameter Ratio 
(AVR) for the diagnosis of chronic diseases, especially 
the change of blood pressure diseases such as 
hypertension is of great significance. The retinal images 
can be directly through the noninvasive fundus 
examination, thus convenient vascular morphology of 
them is analyzed. This paper measures the blood vessels 
cross section of the pixel gray value distribution curve 

624 Informatica 45 (2021) 621–629 X. Xu et al. 
 
can appear due to retinal artery reflective phenomenon, 
so this experiment using the Gaussian mixture model of 
grey value distribution curve fitting. First, the vascular 
tree is utilized according to the bifurcation point is 
broken, and then from the patrol vessels branches 
automatic calculation area and main shaft length, then 
further obtain vascular branches. The average width of 
the value we will compare different width, the 
consistency of the measurement results. 
2.5 Quantitative analysis of retinal blood 
vessels 
Retinal image is a kind of biological image, has certain 
anatomical characteristics of retinal vein embarks from 
the optic disc, gradually outward radiation, and the lower 
branches, blood vessel diameter decreases, the tree 
structure, the small blood vessels in general is a branch 
of the great vessels that branch of the tree structure is 
rich, the entire network of blood vessels in complex 
structures, and topology variety, not easy words to 
describe. In clinical work, general ophthalmic clinicians 
with macular and optic disc regional center as the center, 
to observe the retinal blood vessel image of the up and 
down or so four quadrant vascular lesions of the 
abnormal lesions. At present, many different studies have 
shown that no retinal image lesions with area have 
different diagnostic significance. In addition, fractal 
dimension of retinal vascular network is analyzed to 
know the changes occurring due to chronic diseases [15]. 
3 Materials and methods 
This section includes the architecture of various modules 
like; Web-PACS for Retinal Images based on 
browser/server mode, designing of DICOM-SR for 
Retinal Vascular Images.  
3.1 Architecture of Web-PACS for retinal 
images based on B/S 
In this experiment, the Browser/Server (B/S) mode was 
adopted to construct the retinal image PACS system, 
which mainly included the client layer, the Server layer 
and the data layer presented in Figure 2. 
 
 
 
Figure 2: System Design Diagram for Web-Pals. 
 
The client layer mainly includes a variety of Web 
browser software to support requests to access the middle 
tier services. In this experiment, Microsoft's Internet 
explorer browser, 360 browser and Google Chrome 
browser were installed on the desktop computer in the 
laboratory, safari was installed on the ipad tablet 
computer, and UC browser was installed on the android 
operating system mobile phone. For the server layer, the 
data layer is primarily connected through the Open Data 
Base Connectivity (ODBC) protocol and is serviced by 
the client layer over the HTTP/HTTPs protocol. In this 
experiment, we used dcm4chee open-source server 
software to provide WADO access resolution service for 
network customers, processed service requests from the 
customer layer and made corresponding responses, and 
provided various information needed to access customers 
through the designed interface. 
 
In this experiment, a Lenovo TS230 (Windows7 
operating system, Intel Core i3 processor, 2GB of 
memory, and 5OOGB of hard disk) was installed as the 
server in the laboratory, and the J2EE enterprise version 
servlet engine support JSP running environment was 
installed on it. Web cache and JBOSS server program 
were also installed to support WADO application in the 
Web environment. The data layer includes DICOM 
storage to store DICOM inspection, image series and 
other data. The database and services for this experiment 
are installed on a workstation, and the database 
management system used is an open-source MySQL 
program. The conversion of retinal JPEG images to 
DICOM images is realized by using DICOM gateway 
and transmitted to PACS. 
3.2 Design of DICOM-SR for retinal 
vascular images 
Fundus image diagnosis report contains a large amount 
of medical information reflecting the fundus 
characteristics of patients, but the current image 
diagnosis report USES the narrative natural language or 
text data of medical personnel, making it difficult for the 
computer to extract, analyze study and classify them. 
With the popularity of PACS system and the continuous 
upgrade of DICOM standard, DICOM3.0 standard also 
formulated an image diagnosis result management 
service class for the management of diagnostic report. In 
2000, Information Object Definitions (IODs) of 
Structured Report (SR) was added to DICOM3.0 
standard, providing a basis for computer to understand 
and express image diagnostic reports as presented in 
Figure 3. 
Another important purpose of the design of retinal 
image DICOM-SR is to provide a basis for Computer 
aided diagnosis (CAD). In the diagnosis of retinal 
images, it is of great value to provide reasonable 
templates for clinicians. In this experiment, the CAD-SR 
template for eye foundation disease was designed 
according to the tree structure. Starting from the root of 
the file, it was divided into four parent nodes, and then 
each parent node had several child nodes presented in 
Figure 4. 
 

An Intelligent Information Management System for Retinal Image Storage and ... Informatica 45 (2021) 625–629 625 
 
 
 
Figure 3: e-r model of retinal DICOM-SR. 
 
In terms of function, the process designed in this 
experiment (Figure 3-5) is the blood vessel quantitative 
analysis module provided by the system after the system 
receives the image, which can be used for quantitative 
analysis of retinal blood vessels. For complex fundus 
lesions, doctors can detect and extract the lesions with 
the help of retinal image processing algorithm provided 
by the system, and measure the lesions. 
 
 
 
Figure 4: Design framework of retinal image CAD-SR 
document. 
 
 
 
Figure 5: DICOM model business process diagram of 
CAD-SR. 
 
Figure 5 depicts the DICOM model business process 
diagram of CAD-SR. In addition, doctors can also use 
the fundus lesions and fundus structure for text 
description. The above text and numerical information 
can be stored into PACS data in DICOM-SR format for 
easy query and transmission. For retinal image report 
files, graphic reports are usually required. To this end, 
this experiment designed a structured report document 
model with report content and image. The text 
information in the file is populated into the XSLT style 
sheet. The experimental design of retinal image type is 
contained in the SR, as a result, the XSLT part of the 
image is generated from the WADO image requests.  It is 
seen in the form of XML to the browser side, in order to 
access the users to browse information. 
3.3 Fundus image prepossessing 
A. Normalization of image color 
Because images of retinal blood vessels produced by 
different imaging devices can be affected by light and 
shadow, this will affect the image feature extraction 
algorithm based on seeds point resulting in heavy play 
effect. Therefore, by the experiment of the retinal image 
of the need to be normalized RGB color space. The 
retinal image depicting different R, G, and B channels 
after image color normalization are shown in Figure 6 
along with their respective histograms. 
B. Normalization of image scale 
Due to different imaging equipment of retinal blood 
vessel image resolution may be different, on the back of 
the feature extraction algorithm based on seeds point 
based on different image width results have important 
influence. Therefore, need to experiment by the scale of 
the retinal image space normalized processing, the 
experiment of all processing image pixel is adjusted for 
unified 539 x539 pixel size. 
 
 
 
Figure 6: Image color normalization into R, G, B 
channels (a) Red channel (b) Green Channel (c) Blue 
Channel (d) Histogram of Red channel (e) Histogram of 
Green channel (f) Histogram of Blue channel. 
3.4 Extraction of optic disc in retinal 
images 
Optic disc (OD) is located at the oracular area nasal side 
about 3 mm and 1.5 mm in diameter, is the start of a 
central retinal blood vessels into the fundus. Under 
normal circumstances, 0 d on the fundus image boundary 
clear, white, disc. This experiment by use of the 
morphological characteristics of 0 d first to 0 d, which 
can identify the extraction of intravenous seed point lays 
the foundation for the next step. This method could leak 
will retinal image threshold segmentation of seepage 
clearly separated, the experiment of image processing, 
when the threshold T is equal to 150, has better 
segmentation effect, the result of the finalization image 

626 Informatica 45 (2021) 621–629 X. Xu et al. 
 
recycling used earlier in this paper. The mathematical 
shape of the image processing method to obtain the final 
OD segmentation results. The complete process of OD 
center identification, Boundary localization and OD 
segmentation is depicted in Figure 7. 
 
 
 
Figure 7: (a) Original Retinal image with located OD 
centre, (b) Cropped retinal area with OD boundary 
identified (c) OD segmented portion (d) Retinal image 
after OD segmentation. 
3.5 Retinal image report transmission 
based on WADO protocol 
WADO standard is a technical standard released by 
international DICOM standard organization in 2004, 
which is fully known as Web Access to DICOM 
Persistent Objects. As the 18th part of DICOM3.0 
standard, WADO standard mainly defines a mechanism 
for storing and displaying DICOM Objects from HTML 
or XML documents based on Web through HTTP or 
HTTPS protocol. In addition to single frame images, the 
server can also respond to client requests with different 
types, sending multiple frames of images and text 
documents. Therefore, this experiment carried out 
relevant experiments of retinal DICOM image and 
structured report through WADO standard. Our 
framework based on IFIE ophthalmic technology 
(EYECARE TF, version 3.7) to confirm the function 
modules of the medical information enterprises subset, 
and made clear a series of mutual coordination between 
the medical enterprises, based on the standard of the 
transaction. IHE defines the DICOM queue from the 
service users (Supreme Court of the United States) and 
services (sarcoplasmic calcium-binding proteins) 
matching and comes back key, eligible keys used to 
select the query service instance of a class (instance) to 
users of the service industry, and then the key value 
returned back only specific data is not used for matching. 
Figure 8 illustrates the flowchart of retinal image/report 
under IHE specification.  
 
 
Figure 8: Flow chart of retinal image/report under IHE 
specification. 
When the patient comes for fundus examination, the 
type of examination is first set, and the user queries the 
department or operation arrangement through the 
DICOM modal work list. The work list is downloaded 
into the imaging mode, and the user confirms the 
information. When a shoot transaction is executed, this 
information is written to the header section of the 
resulting object. Get the modal or get the modal import 
operation sends the retrieved modal Image to the Image 
Archive (< Image Archive) and appears in the header file 
information that produces the Image. After receiving the 
image display demand of retrieval operation, the DICOM 
image and measurement required are transferred from the 
image file to the image display module for display. 
Results and Analysis 
This section illustrates the analysis of results 
obtained for retinal image Web-PACS, access and 
display of retinal image on DICOM-SR and finally 
presents its discussion and summary. 
3.6 Retinal image Web-PACS 
The functions of the retinal image PACS system in this 
experiment include: first, the function of storage and 
browsing. Open the PACS system and you can enter the 
work list of fundus images. After selecting the desired 
image catalog, you can display the pictures and reports of 
patients on the Web page. In addition, third-party 
DICOM browser can also be used to look up the patient 
picture by name, check mode and other means. The 
patient image is stored in the same way as the SR report 
and can be read at the same time. Second, the basic 
processing functions of retinal image realizes browser-
based image processing functions, including window 
width and window level adjustment, manual 
measurement, zoom size, rotation and translation, 
negative display and other basic image processing 
functions. Third function is the communication which 
can realize remote communication function, image 
compression and transmission, download patient 
information and image retrieval and query function. In 
this study, the website is logged on through computer 
system with different operating system which can 
provide the fundus images and related reports 
conveniently. The Web-PACS system connected in the 
laboratory for the experimentation purpose is shown in 
Figure 9. 
 

An Intelligent Information Management System for Retinal Image Storage and ... Informatica 45 (2021) 627–629 627 
 
 
 
Figure 9: Web-PACS system connected in the laboratory. 
3.7 Access and display of retinal image 
DICOM-SR 
Retinal image transmission based on WADO standard 
can better solve the problem of heterogeneous medical 
information system accessing retinal image on the 
Internet. Based on the traditional PACS system, this 
standard defines how the Web client obtains the format 
and method of retinal image URL from the Web enabled 
DICOM server. In this experiment, retinal image 
DICOM browser and WADO Application server are both 
Application entities (AE) conforming to DICOM 
standard, which can realize the interaction function of 
heterogeneous medical information integration platform 
through the standard DICOM communication service. 
The basic structure of WADO is http://< server IP 
address >< path >? < query parameter >, where the study 
UID parameter is the UID of the object's check, and the 
series UID is the UID of the object's series.  
For comparing the diagnostic ability of the retinal 
imaging procedure using the DICOM images, various 
evaluation parameters are taken into consideration like 
Accuracy, Sensitivity, Specificity and Precision. These 
values are derived from true positives (TP), true 
negatives (TN), false positives (FP) and false negatives 
(FN) values. The outcomes of the diagnostic ability in 
terms of recognition performance are depicted in Table 1. 
 
Evaluation Parameters  Recognition Value 
Accuracy  98.51% 
Sensitivity 98.04% 
Specificity 99% 
Precision 99.01% 
Table 1: Outcomes of diagnostic ability in terms of 
recognition performance. 
 
It is seen from Table 1 that the accuracy value of 
98.51% is obtained for the retinal image recognition 
along with the sensitivity of 98.04%, Specificity of 99% 
and precision value of 99.01%. The viability of the 
accurate recognition rate and other parameters is justified 
by the outcomes obtained.  
3.8 Discussion and summary 
Different from the general medical imaging equipment, 
which directly integrates the images into the hospital 
PACS system, the object of this experiment is retinal 
images, which are usually produced in the 
ophthalmology and endocrinology departments of the 
hospital and generally stored on the workstation of the 
inspection equipment. Arvin et al. converted pathological 
images into DICOM images and uploaded them to the 
PACS system, realizing the information system 
integration of pathological images. Similarly, retinal 
image is integrated into PACS system by dicomization of 
retinal image. In addition, it is a better solution to use the 
image device with DICOM standard digital interface for 
direct transmission. The information exchange between 
application entities follows the Client/Server (C/S) 
pattern, with Service Class Consumer (SCU) and Service 
Class Provider (SCP) as Client Server roles. When the 
imaging device is combined with the application 
program, special adapters are used to negotiate the 
transmission syntax and how the message flow is 
decoded in the communication process, which can be 
connected with the realization to improve the efficiency 
of information collection.  
In this article, we discussed the slow disease 
information management in the retinal image information 
and its significance, and then puts forward a framework 
of retinal image transmission based on WADO standards, 
and through the architecture of retinal image PACS 
system, realized the browser to access the retinal image 
and structured report, for retinal image information in a 
heterogeneous invulnerability laid a solid foundation in 
the health care system. The WADO standard adopted in 
this experiment is based on the Web and can store and 
display DICOM objects from HTML or XML documents 
through HTTP or HTTPS protocols. DICOM server set 
DICOM parser, can achieve DICOM to JPEG format 
conversion, so that the browser client can access the 
retina DICOM image. In order to enable browsers to 
access DICOM data through the WADO standard, 
WADO server components need to be designed and 
deployed so that users can access retinal images and 
report data using the Web on top of the existing 
information system. In this experiment, through the 
architecture of retinal image WADO application server 
conforming to DICOM standard, the interactive function 
of retinal DICOM image communication service based 
on Web was realized. In addition to the parameters used 
in the URL in this experiment, the URL can also be used 
to define the transfer syntax of the image 
(lossless/lossless compression), to define the luminosity 
and contrast of the image, the portion of the image that 
will be displayed, and to add annotations to the image. 
4 Conclusions 
The proposed WADO based retinal image transmission 
technology and structured numerical report in DICOM-
SR can better solve the invulnerability problem of retinal 
image in different systems. The analysis done in this 

628 Informatica 45 (2021) 621–629 X. Xu et al. 
 
work for the investigation of invulnerable retinal imaging 
information can be used for quantitative analysis of 
morphological change of retinal vascular network. This 
work is mainly focused on the medical digital image 
transmission protocol Digital Imaging and 
Communications in Medicine (DICOM) version 3.0 and 
the retinal image Picture Archiving and Communication 
System (PACS) was constructed in the laboratory. The 
retinal image PACS system constructed in B/S mode can 
effectively store and transmit DICOM images when 
combined with the application program. This project will 
integrate quantitative features of retina in future research, 
providing more meaningful research data for data mining 
based on chronic disease management system. In 
addition, a study will be conducted on the conversion of 
retinal images and reports based on DICOM 3.0 standard 
and HL7 CDA documents. Therefore, in order to provide 
a technical basis for the integration of retinal images and 
existing resident health records with HL7 interfaces. The 
feasibility of the recognition rate and other evaluation 
parameters is justified by obtaining the 98.51% accuracy 
rate with comparatively better values of sensitivity, 
specificity and precision. 
References  
[1] Tu, Y., Huang, Y., & Yi, F. (2015, April). Chronic 
disease management system design based on cloud 
storage architecture. In 2015 2nd International 
Conference on Information Science and Control 
Engineering, pp. 654-658. IEEE. 
https://doi.org/10.1109/ICISCE.2015.151.  
[2] Sharma, A., Ansari, M. D., & Kumar, R. (2017, 
September). A comparative study of edge detectors 
in digital image processing. In 2017 4th International 
Conference on Signal Processing, Computing and 
Control (ISPCC), pp. 246-250.  
https://doi.org/10.1109/ISPCC.2017.8269683. 
[3] Bhardwaj, C., Jain, S., & Sood, M. (2020). Retinal 
blood vessel localization to expedite PDR 
diagnosis. Periodicals of Engineering and Natural 
Sciences (PEN), 8(3), 1233-1246.  
http://dx.doi.org/10.21533/pen.v8i3.208. 
[4] Bhardwaj, C., Jain, S., & Sood, M. (2019). 
Automatic blood vessel extraction of fundus images 
employing fuzzy approach. Indonesian Journal of 
Electrical Engineering and Informatics (IJEEI), 7(4), 
757-771.  
https://doi.org/10.11591/ijeei.v7i4.991. 
[5] Shah, A. R., & Gardner, T. W. (2017). Diabetic 
retinopathy: research to clinical practice. Clinical 
diabetes and endocrinology, 3(1), 1-7.  
https://doi.org/10.1186/s40842-017-0047-y. 
[6] Dogra, J., Jain, S., Sharma, A., Kumar, R., & Sood, 
M. (2020). Brain tumor detection from MR images 
employing fuzzy graph cut technique. Recent 
Advances in Computer Science and 
Communications (Formerly: Recent Patents on 
Computer Science), 13(3), 362-369.  
https://doi.org/10.2174/22132759126661812071526
33. 
[7] Bhardwaj, C., Jain, S., & Sood, M. (2020). Diabetic 
Retinopathy Lesion Discriminative Diagnostic 
System for Retinal Fundus Images. Advanced 
Biomedical Engineering, 9, 71-82.  
https://doi.org/10.14326/abe.9.71. 
[8] Bhardwaj, C., Jain, S., & Sood, M.(2020). 
Automated Diagnostic Hybrid Lesion Detection 
System for Diabetic Retinopathy 
Abnormalities. International Journal of Sensors 
Wireless Communications and Control, 10(4), 494-
507. 
https://doi.org/10.2174/22103279096661911260924
11.  
[9] Sharma, A., Tomar, R., Chilamkurti, N., & Kim, B. 
G. (2020). Blockchain based smart contracts for 
internet of medical things in e-
healthcare. Electronics, 9(10), 1609.  
https://doi.org/10.3390/electronics9101609.  
[10] Bhardwaj, C., Jain, S., & Sood, M. (2020). 
Hierarchical severity grade classification of non-
proliferative diabetic retinopathy. Journal of 
Ambient Intelligence and Humanized Computing, 1-
22.  
https://doi.org/ 10.1007/s12652-020-02426-9. 
[11] Mishra, A. R., Garg, A. K., Purwar, H., Rana, P., 
Liao, H., & Mardani, A. (2021). An extended 
intuitionistic fuzzy multi-attributive border 
approximation area comparison approach for 
smartphone selection using discrimination 
measures. Informatica, 32(1), 119-143,  
DOI: https://doi.org/ 10.15388/20-INFOR430. 
[12] Chung, K., & Park, R. C. (2019). Cloud based u-
healthcare network with QoS guarantee for mobile 
health service. Cluster Computing, 22(1), 2001-
2015.  
https://doi.org/10.1007/s10586-017-1120-0. 
[13] Zhang, S., Tang, J., Meng, F., & Yuan, R. (2021). A 
group decision making method with interval-valued 
intuitionistic fuzzy preference relations and its 
application in the selection of cloud computing 
vendors for SMEs. Informatica, 32(1), 163-193,  
DOI: https://doi.org/ 10.15388/20-INFOR416 
[14] Xiaosong, L., Gang, S., & Yong, C. (2015, 
November). Design and development of the web-
based health and chronic disease assessment 
management system. In 2015 7th International 
Conference on Information Technology in Medicine 
and Education (ITME), pp. 72-75. IEEE.  
https://doi.org/10.1109/ITME.2015.167. 
[15] Liu, X., Zeng, H., Chand, N., & Effenberger, F. 
(2015). Efficient mobile fronthaul via DSP-based 
channel aggregation. Journal of Lightwave 
Technology, 34(6), 1556-1564.  
https://doi.org/10.1109/JLT.2015.2508451. 
[16] Gusyev, M. A., Morgenstern, U., Nishihara, T., 
Hayashi, T., Akata, N., Ichiyanagi, K., ... & Stewart, 
M. K. (2019). Evaluating anthropogenic and 
environmental tritium effects using precipitation and 
Hokkaido snowpack at selected coastal locations in 
Asia. Science of The Total Environment, 659, 1307-
1321.  

An Intelligent Information Management System for Retinal Image Storage and ... Informatica 45 (2021) 629–629 629 
 
https://doi.org/10.1016/j.scitotenv.2018.12.342. 
[17] International Society for Optics and Photonics. 
https://doi.org/10.1117/12.843714. Kalibatienė, D., 
& Miliauskaitė, J. (2021). A Hybrid Systematic 
Review Approach on Complexity Issues in Data-
Driven Fuzzy Inference Systems 
Development. Informatica, 32(1), 85-118,  
DOI: https://doi.org/ 10.15388/21-INFOR444. 
[18] Koutelakis, G. V., & Lymberopoulos, D. K. (2008). 
WADA service: an extension of DICOM WADO 
service. IEEE Transactions on Information 
Technology in Biomedicine, 13(1), 121-130.  
https://doi.org/10.1109/TITB.2008.2007197. 
[19] Al-Rawi, M., & Karajeh, H. (2007). Genetic 
algorithm matched filter optimization for automated 
detection of blood vessels from digital retinal 
images. Computer methods and programs in 
biomedicine, 87(3), 248-253.  
https://doi.org/10.1016/j.cmpb.2007.05.012. 
[20] Xu, Xin, Li Li, and Ashutosh Sharma. "Controlling 
messy errors in virtual reconstruction of random 
sports image capture points for complex systems." 
International journal of system assurance 
engineering and management (2021): 1-8.  
https://doi.org/10.1007/s13198-021-01094-y 
[21] Powell, M. J. (1964). An efficient method for 
finding the minimum of a function of several 
variables without calculating derivatives. The 
computer journal, 7(2), 155-162.  
https://doi.org/10.1093/comjnl/7.2.155. 
[22] Boncea, R., PETRE, I., SMADA, D. M., & 
ZAMFIROIU ANAGRAMA, A. (2017). A Maturity 
Analysis of Big Data Technologies. Informatica 
Economica, 21(1),DOI:  
https://doi.org/10.12948/issn14531305/21.1.2017.05 
  

630 Informatica 45 (2021) 621–629 X. Xu et al.