international journal of pure and applied mathematics...

8
COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN CV 1 Limshe Pearce. A, 2 Radhika.R 1 PG Student, Department of ECE, S.A Engineering College, Thiruverkadu, Chennai – 600077, Tamil Nadu, India 2 Associate Professor, Department of ECE, S.A Engineering College, Thiruverkadu, Chennai – 600077, Tamil Nadu, India 1 [email protected], 2 [email protected] Abstract: Identification of dental problems is a major issue which is faced by the dentistry worldwide. Computer aided diagnosis has led to development in segmentation techniques. In a survey, it is found that over 90 percent of the humans have dental caries. The dental x-rays are normally subject to low contrast and noisy images. In order to acquire a clear image of the teeth dental cameras are used. These cameras can provide a better image of the teeth. Dental problems are generally detected through visual perception but this is not sufficient because the experience of the doctors play a key role in this method. Hence automatic diagnosis systems are needed to have precise techniques for detection of teeth problems. To develop a system for automatic diagnosis of dental issues an algorithm known as shape detection algorithm to detect the problems associated with the shape of the teeth is used. This system is used for early diagnosis of dental issues which usually occur at the crown of the teeth. The segmentation method used in the system is the Super pixel segmentation without making use of edge algorithms. The Scale invariant feature transform is the method used for extracting the features from the segmented image. The extracted features are compared with the features in the database using Open CV and the result is produced. The experimental results indicate that the number of iterations required for determining the problems of the teeth is less resulting in a computationally fast system. Keywords: Dental camera, Shape detection algorithm, Dental images, Open CV. 1. Introduction Medical imaging techniques have been developing at a very fast rate. Due to the advancements in technology and the reliability in the results obtained modern imaging techniques are widely used [1]. These techniques help the doctors to view the internal parts of the body. In Dental imaging the innermost teeth can also be visualized clearly using a camera [12]. These equipments help in the diagnosis of dental problems at an earlier stage. Early diagnosis of Oral disease helps the doctors to provide a better treatment and prevent them from affecting other of the body. The common problems which affect structure of the teeth are chipped teeth, cracked teeth, dislocation of teeth and dental cavities. The chipping and cracking of teeth may be due to lack of minerals. They are the symptoms which denote that there is a problem in their body. The dental caries affect the crown of the teeth at an earlier stage [3]. Dental cavities are caused by a special type of bacteria that erodes the enamel which acts as a protecting layer of the tooth. Dental Plague is also a condition caused due to a type of bacteria by building up a sticky film over the teeth. Dentin is also a protective layer covering the surface of the tooth. If these conditions are diagnosed at an earlier stage it can be prevented from affecting the other parts of the teeth such as the root. The dental image is composed of two regions: Soft tissue region (low intensity) Teeth region (high intensity) Based upon the difference in intensity values image processing can be done. The low intensity values are removed and only the high intensity values are considered as the region of interest. The further processing is done in the high intensity areas. The image segmentation is a very important step in the image processing. The major part of the image processing technique comprises of image segmentation [4]. The technique used for image segmentation is Super pixel segmentation. This is a method of segmentation which is done without removing the edges of the acquired dental image. The feature extraction involves the main part in this system. The features are extracted from the segmented dental images. The method used for feature extraction is the Scale invariant feature transform which is used in the extraction of important features corresponding to the surface of the image. Intra oral cameras have a International Journal of Pure and Applied Mathematics Volume 118 No. 10 2018, 389-396 ISSN: 1311-8080 (printed version); ISSN: 1314-3395 (on-line version) url: http://www.ijpam.eu doi: 10.12732/ijpam.v118i10.79 Special Issue ijpam.eu 389

Upload: others

Post on 15-Jul-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN CV

1Limshe Pearce. A,

2Radhika.R

1PG Student, Department of ECE, S.A Engineering College,

Thiruverkadu, Chennai – 600077, Tamil Nadu, India 2Associate Professor, Department of ECE, S.A Engineering College,

Thiruverkadu, Chennai – 600077, Tamil Nadu, India [email protected], [email protected]

Abstract: Identification of dental problems is a major issue which is faced by the dentistry worldwide. Computer

aided diagnosis has led to development in segmentation techniques. In a survey, it is found that over 90 percent of the humans have dental caries. The dental x-rays are normally subject to low contrast and noisy images. In order to

acquire a clear image of the teeth dental cameras are used. These cameras can provide a better image of the teeth.

Dental problems are generally detected through visual perception but this is not sufficient because the experience of

the doctors play a key role in this method. Hence automatic diagnosis systems are needed to have precise techniques

for detection of teeth problems. To develop a system for automatic diagnosis of dental issues an algorithm known as

shape detection algorithm to detect the problems associated with the shape of the teeth is used. This system is used

for early diagnosis of dental issues which usually occur at the crown of the teeth. The segmentation method used in

the system is the Super pixel segmentation without making use of edge algorithms. The Scale invariant feature

transform is the method used for extracting the features from the segmented image. The extracted features are

compared with the features in the database using Open CV and the result is produced. The experimental results

indicate that the number of iterations required for determining the problems of the teeth is less resulting in a

computationally fast system.

Keywords: Dental camera, Shape detection algorithm, Dental images, Open CV.

1. Introduction

Medical imaging techniques have been developing at a very fast rate. Due to the advancements in technology and

the reliability in the results obtained modern imaging techniques are widely used [1]. These techniques help the

doctors to view the internal parts of the body. In Dental imaging the innermost teeth can also be visualized clearly using a camera [12]. These equipments help in the diagnosis of dental problems at an earlier stage. Early diagnosis

of Oral disease helps the doctors to provide a better treatment and prevent them from affecting other of the body.

The common problems which affect structure of the teeth are chipped teeth, cracked teeth, dislocation of teeth and

dental cavities. The chipping and cracking of teeth may be due to lack of minerals. They are the symptoms which

denote that there is a problem in their body. The dental caries affect the crown of the teeth at an earlier stage [3].

Dental cavities are caused by a special type of bacteria that erodes the enamel which acts as a protecting layer of the

tooth. Dental Plague is also a condition caused due to a type of bacteria by building up a sticky film over the teeth. Dentin is also a protective layer covering the surface of the tooth. If these conditions are diagnosed at an earlier

stage it can be prevented from affecting the other parts of the teeth such as the root. The dental image is composed

of two regions:

• Soft tissue region (low intensity)

• Teeth region (high intensity)

Based upon the difference in intensity values image processing can be done. The low intensity values are

removed and only the high intensity values are considered as the region of interest. The further processing is done in

the high intensity areas. The image segmentation is a very important step in the image processing. The major part of the image processing technique comprises of image segmentation [4]. The technique used for image segmentation is

Super pixel segmentation. This is a method of segmentation which is done without removing the edges of the

acquired dental image. The feature extraction involves the main part in this system. The features are extracted from

the segmented dental images. The method used for feature extraction is the Scale invariant feature transform which

is used in the extraction of important features corresponding to the surface of the image. Intra oral cameras have a

International Journal of Pure and Applied MathematicsVolume 118 No. 10 2018, 389-396ISSN: 1311-8080 (printed version); ISSN: 1314-3395 (on-line version)url: http://www.ijpam.eudoi: 10.12732/ijpam.v118i10.79Special Issue ijpam.eu

389

Page 2: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

wide importance as they provide a clear view of the teeth. The panoramic image provides a better dental image

processing [2]. The accessibility of cameras has paved a way for several improved methods. Open CV library is an

open source computer vision library is built to provide a common infrastructure for computer aided applications. It

has several algorithms embedded in it and can be easily programmed using the languages such as C++, Java and

Python. They are used in the detection and tracking applications.

The remaining parts of this paper are as follows. Section II presents the Literature review. The Shape detection

algorithm is explained in Section III. The Section IV presents the Simulation results and finally Section V concludes

the paper.

2. Proposed Work

The proposed work consists of the shape detection algorithm. The algorithm which is specially designed to identify

the shapes the teeth is termed as the shape detection algorithm. The Database images serve as the base of the shape

detection algorithm as the initial parameters are extracted from the database images to design the algorithm. The

Block diagram of the proposed work is given in Figure 1. The Blocks represent the entire process involved in the

automatic detection of dental problems in the teeth.

Figure 1. Block diagram of the proposed work

The Input image is the Dental image acquired from a dental camera. The image obtained will be a colour image with

high resolution. The acquired image is initially pre-processed that is the size of the image is adjusted in accordance

with the algorithm requirements. The pre-processed image is then subject to Scale Invariant Transform (SIFT). The

SIFT Transform has a characteristic property of feature extraction in which the main features that determine the

property of the image is extracted. The extracted features are compared with those features that have been extracted

at an earlier stage and stored in the database image. The algorithm works with the predetermined extracted features

and compares them with the new one. The comparison is done with Open CV which has several libraries inbuilt in it

in order to determine object tracking, object detection, object recognition. The algorithm is written in python

language in order to simplify the codes of the program. The system operates and produces the desired result. The

type of problem in the image obtained is displayed as the result of the program.

3. Shape Detection Algorithm

The proposed system is the Shape detection algorithm. An algorithm which is used to detect the problems associated

with the shape of the teeth is shape detection algorithm. Majority of the dental problems can be diagnosed from the

crown of the teeth. In the same way shape detection algorithm helps in automatically detecting the problems of the

teeth related to the structure. The algorithm involves the steps shown in Figure 2

International Journal of Pure and Applied Mathematics Special Issue

390

Page 3: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

Figure 2. Steps involved in Shape detection Algorithm

A. Input Image

An image is generally a two dimensional function given by f(x,y) where x and y are the planar coordinates. The

amplitude of the image at any point can be termed as the intensity of the image [6]. An image is usually available in

an analog format it has to be converted into a digital format to perform processing. A digital image is formed when

an analog coordinates are converted into finite discrete values [5]. The digital image is composed of finite elements

and each finite element is known as the pixel. Pixels have an important role in image processing techniques. The

colour images are stored in a RGB format which is converted to gray scale to perform image processing techniques.

The gray scale image will emit the same amount of light in every channel so in order to differentiate the total amount of emitted light the available pixels are described as dark pixels and bright pixels [9]. The lighter pixels are

termed as bright pixels and the little darker pixels are termed as dark pixels. The input image is the dental image

captured using a Dental camera. The dental camera used here is a high resolution camera to produce a high quality

image. Resolution of the image obtained is high when compared to the other images acquired from different

methods.

B. Pre-processing

Image Scaling and resizing are the processes done in pre-processing. Image Scaling is done in all the digital image

processing techniques whether it may be a Bayer demosaicing or an enlargement technique. Image resizing is done

if there is a need to increase or decrease the total number of pixels present in the image [8]. There are number of

reasons to perform image resizing one of the important reason is each and every camera has its own resolution, so

when a system is designed based on the camera specifications it is important to note that the images acquired meet

the necessary requirements [13]. If not the algorithm will not run and the system would fail to produce an optimized

result. In image processing techniques colour image cannot be processed efficiently. So the input image acquired

which is a colour image is converted into a gray scale image for processing. The captured input image has two

regions the soft tissue region and the teeth region. The soft tissue region is the low contrast region whereas the teeth

region is the high contrast region. On pre-processing, the low contrast region is removed by discrete wavelet

transform techniques [15]. The region of interest is the teeth region which needs to be processed for determination of

the dental problems.

C. Image Segmentation

Image segmentation is the important phase in the image processing techniques. Segmentation can be given by the

division of the pre-processed image into several segments after which the segments are analyzed [14]. There are a

variety of Image segmentation techniques but still the requirements are not met in the fundamental image

processing. Different methods are developed and used as per the requirements of the system. The segmentation method used in this system is the super pixel segmentation [11]. This is suitable of images acquired using a dental

International Journal of Pure and Applied Mathematics Special Issue

391

Page 4: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

camera. As many of the algorithms use pixel grid as a representation underlying the image in computer vision

techniques the stochastic model such as Markov random fields are defined in a regular field. This pixel grid

representation is not a natural representation method but rather a artefact method of representation in digital image

processing. Super pixels segmentation is an emerging method in computer vision techniques [7]. The segmented

pixels are termed as super pixels. This method is a computationally efficient method and the structure of the image

is conserved. The segmentation process uses normalized cuts which preserve the structure of the image. The Simple

Linear Iterative Clustering groups the similar pixels in a cluster. The pre-processed image is subject to segmentation

process after which features are extracted.

D. Scale Invariant Feature Transform

The interesting points in any image are termed as a feature. These features describe the characteristics of an object.

They can be extracted and grouped into a cluster to perform recognition techniques. It is a important factor that the

features extracted should be clearly viewed even in the presence of certain amount of noise. The edges are the

important features in an object. It should be noted to preserve the edges while processing the image with a variety of

filters. On scaling and resizing there is possibility for occurrence of change in the geometry of the image. The

extraction techniques are mainly used to quantify the area of caries on the surface of the tooth. The method used for feature extraction in this system is the Scale invariant feature transform. Normal feature extraction techniques can

only detect the features in the image if the scaling of the image remains unchanged but using SIFT large number of

features can be extracted enabling them to detected even the scaled and rotated images thus reducing the errors

caused due to local variations [10]. The feature vectors are created for each and every pixel which is segmented

along with the feature vectors of the neighbouring pixel. The features extracted in the dental image correspond to the

faulty portion of the teeth. Depending upon the pixel values of the extracted features the threshold value is assigned.

E. Identification of defects

The Image matching s the final step involved in this system it also corresponds to the final step in the fundamental

steps involved in the digital image processing. The Database images consist of several conditions of the teeth in

which the features extraction is done. The features are embedded in the algorithm so that the result can be

effectively achieved. The Extracted features are subjected to the matching process and the results are obtained. Identification of the defects present in the teeth is the primary aim of the project and this is effectively made possible

by the use of Open CV library. It provides various applications. The programming is done in python language. The

diagnosis of the oral diseases is done using the Open CV in a Linux platform. The effectiveness of the system can be

viewed from the Simulation results. The system is computationally fast and reliable on comparison with the other

methods.

4. Simulation Results

The Shape detection algorithm is implemented in the dental images captured using a dental camera. Open CV is

used to compare the images with the database features and the following results are obtained.

Figure 3. Input image of a cracked tooth

International Journal of Pure and Applied Mathematics Special Issue

392

Page 5: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

Figure 4

The Figure 3 shows the Input image of a cracked tooth. The image is given as the input to the Open CV and the

output is obtained. The Figure 4 shows the output of a cracked tooth. The dental image in which diagnosis should be

done is compared with pixel values in the database. If there is a mismatch or a discontinuity in the pixels values the

output can be observed as “Cracked Tooth”.

Figure 5

Figure 6

Figure 4. Simulation output of a cracked tooth

shows the Input image of a cracked tooth. The image is given as the input to the Open CV and the

shows the output of a cracked tooth. The dental image in which diagnosis should be

done is compared with pixel values in the database. If there is a mismatch or a discontinuity in the pixels values the

output can be observed as “Cracked Tooth”.

Figure 5. Input image of a dental cavity

Figure 6. Simulation output of a dental cavity

shows the Input image of a cracked tooth. The image is given as the input to the Open CV and the

shows the output of a cracked tooth. The dental image in which diagnosis should be

done is compared with pixel values in the database. If there is a mismatch or a discontinuity in the pixels values the

International Journal of Pure and Applied Mathematics Special Issue

393

Page 6: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

The Figure 5 shows the Input image of a dental cavity. The dental cavity can be analysed by comparing the

surface of the input image with the extracted features. Depending on the comparison values the simulation result of a

dental cavity is shown in Figure 6. The structure of the tooth is compared with the input dental image. Based upon

the variation in the values the presence of cavity is determined.

Figure 7

Figure 8

The Figure 7 shows the Input image of chipped teeth. The input image is compared with the features available

in the database by processing the shape detection algorithm.

The Figure 8 shows the simulation output of a chipped tooth. The image values and the output is produced. The threshold values are compared with the input dental image and based

upon the level of variation the results are obtained. If the variation of the threshold is negligible

be “NO CHIPPED”. If the variation of threshold value is minimum then the output will be displayed as “chipped

teeth” and if the difference in the threshold value is maximum then the output will be given as “Tooth chipped

maximum”.

The simulation results show the results of problems of the teeth. The effectiveness of the system can be viewed

from the results as accurate results are obtained.

In this paper, an automatic dental diagnostic system has been implemented to identify the problems in the teeth. The

dental images which are captured by an

algorithm that is used to detect the shape of the teeth also recognizes the type the teeth. The algorithm has good

ability to detect most of the defective shapes in the image. The image to be pro

an improved analysis. A threshold is assigned to the images in the database which helps in matching the image

The Figure 5 shows the Input image of a dental cavity. The dental cavity can be analysed by comparing the

surface of the input image with the extracted features. Depending on the comparison values the simulation result of a

6. The structure of the tooth is compared with the input dental image. Based upon

the variation in the values the presence of cavity is determined.

Figure 7. Input image of a chipped teeth

Figure 8. Simulation output of a chipped teeth

shows the Input image of chipped teeth. The input image is compared with the features available

in the database by processing the shape detection algorithm.

shows the simulation output of a chipped tooth. The pixels values are compared with the input image values and the output is produced. The threshold values are compared with the input dental image and based

upon the level of variation the results are obtained. If the variation of the threshold is negligible

be “NO CHIPPED”. If the variation of threshold value is minimum then the output will be displayed as “chipped

teeth” and if the difference in the threshold value is maximum then the output will be given as “Tooth chipped

The simulation results show the results of problems of the teeth. The effectiveness of the system can be viewed

from the results as accurate results are obtained.

5. Conclusion

In this paper, an automatic dental diagnostic system has been implemented to identify the problems in the teeth. The

dental images which are captured by an dental camera are processed using the Shape detection algorithm. The

algorithm that is used to detect the shape of the teeth also recognizes the type the teeth. The algorithm has good

ability to detect most of the defective shapes in the image. The image to be processed is converted into gray scale for

an improved analysis. A threshold is assigned to the images in the database which helps in matching the image

The Figure 5 shows the Input image of a dental cavity. The dental cavity can be analysed by comparing the

surface of the input image with the extracted features. Depending on the comparison values the simulation result of a

6. The structure of the tooth is compared with the input dental image. Based upon

shows the Input image of chipped teeth. The input image is compared with the features available

pixels values are compared with the input image values and the output is produced. The threshold values are compared with the input dental image and based

upon the level of variation the results are obtained. If the variation of the threshold is negligible then the result will

be “NO CHIPPED”. If the variation of threshold value is minimum then the output will be displayed as “chipped

teeth” and if the difference in the threshold value is maximum then the output will be given as “Tooth chipped

The simulation results show the results of problems of the teeth. The effectiveness of the system can be viewed

In this paper, an automatic dental diagnostic system has been implemented to identify the problems in the teeth. The

amera are processed using the Shape detection algorithm. The

algorithm that is used to detect the shape of the teeth also recognizes the type the teeth. The algorithm has good

cessed is converted into gray scale for

an improved analysis. A threshold is assigned to the images in the database which helps in matching the image

International Journal of Pure and Applied Mathematics Special Issue

394

Page 7: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

based upon the threshold values of the processed dental image. The comparison of dental images is done using Open

Cv and the results are obtained. The effectiveness of the system can be known from the results obtained on

processing the image.

References

[1] Anupama Bhan, Garima Vyas, Sourav Mishra and Pulkit Pandey (2016), “Detection and Grading Severity of

Caries in Dental X-Ray Images”, International Conference on Micro-Electronics and Telecommunication

Engineering (ICMETE), pp.375-378.

[2] Brllmann D, Schmidtmann I, Warzecha K, and d’Hoedt B (2011), “Recognition of root canal orifices at a

distance - a preliminary study of teledentistry”, Journal of Telemedicine and Telecare, vol. 17, no. 3, pp. 154-157.

[3] Grace F. Olsen, Susan S. Brilliant and David Primeaux (2009), “An Image-Processing Enabled Dental Caries

Detection System”, in International Conference on Complex Medical Engineering (ICME), pp.1-8.

[4] H. A. R Nasimento, A. C. A Ramos, S. Neves, S. L. de- Azeuedo-vaz and D. Q. Feritas (2014), “The Sharpen Filters Improves the Radiographic Detection of Vertical Root Fractures”, International Endodontic Journal, vol.48,

no.5, pp. 428–434.

[5] Kapur, J.N., P.K. Sahoo, and A.K. Wong (2011), “A new method for graylevel picture thresholding using the

entropy of the histogram”, Computer vision, graphics, and image processing, vol.29, no.3, pp. 273-285.

[6] Keith Angelino, David A. Edlund, and Pratik Shah (2017), “Near-infrared imaging for detecting caries and

structural deformities in teeth”, IEEE Journal of Translational Engineering in Health and Medicine, vol. 5.

[7] Leo Grady (2006), “Random walks for image segmentation”, IEEE Transaction on Pattern Analysis and

Machine Intelligence, vol. 28, no. 11, pp. 1768–1783.

[8] MaiKasai, Yuka Iijima and Hiroshi Takemura (2016), “Dental Plaque Assessment Lifelogging System Using Commercial Camera for Oral Health care”, Annual International conference of IEEE engineering in medicine and

biology society, pp. 2566-2569.

[9] Orazio Gambino, Fausto Lima Roberto Pirrone and Edoardo Ardizzone (2014), “Second Opinion System for

Intraoral Lesions”, IEEE 27th International Symposium on Computer Based Medical Systems, pp. 495-496.

[10] Raghav Agarwal and Abhay Kumar Agrawal (2016), “A Review Paper on Diagnosis of Approximal and Occlusal Dental Caries using Digital Processing of Medical Images”, International Conference on Emerging Trends

in Electrical, Electronics and Sustainable Energy Systems, pp.383–385.

[11] Ramzi Ben Ali, Ridha Ejbali and Mourad Zaied (2015), “GPU-based Segmentation of Dental X-ray Images

using Active Contours Without Edges”, 15th International Conference on Intelligent Systems Design and

Applications (ISDA), pp. 505-510.

[12] Sameh M. Yamany and Aly A.Farag (2011), “A System for Human Jaw Modeling Using Intra-Oral Images”,

Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology

Society, vol. 2, pp. 562-566.

[13] Tran Manh Tuan, Nguyen Hai Minh and Nguyen Van Tao (2016), “Medical Diagnosis from Dental X-Ray

Images: A Novel Approach Using Clustering Combined with Fuzzy Rule based Systems”, Annual Conference of

the North American Fuzzy Information Processing Society (NAFIPS), pp. 1–6.

[14] Truong Quang Vinh, Bui Minh Thanh and Nguyen Ngoc Tai (2013), “Dental Intraoral System Supporting

Tooth Segmentation”, International Conference on Computing Management and Telecommunications, pp. 326-329.

[15] Yongjia Xiang, Jiayong Yan, and Xiaohua Jian (2011), “Automated Detection and Quantification of Early

Caries Lesions on Images Captured by Intraoral Camera”, International Symposium on Bioelectronics and Bioinformations, pp.251-254.

International Journal of Pure and Applied Mathematics Special Issue

395

Page 8: International Journal of Pure and Applied Mathematics ...acadpubl.eu/jsi/2018-118-10-11/articles/10/79.pdf · COMPUTER AIDED DIAGNOSIS OF ORAL DISEASE IN DENTAL IMAGES USING OPEN

396