ISSN 2278-3091

Volume 10, No.3, May - June 2021

International
Amani Journal
Al-Ghraibah et al., ofJournal
International Advanced Trends
of Advanced in Computer
Trends in Computer Science andScience
Engineering,and
10(3), Engineering
May – June 2021,
1494 – 1500Available Online at http://www.warse.org/IJATCSE/static/pdf/file/ijatcse011032021.pdf
https://doi.org/10.30534/ijatcse/2021/011032021

Classification of Long Bone X-ray Images using New

features and Support Vector Machine
1
Amani Al-Ghraibah , Mohammad Algharibeh2, Waseem AlMohtasib3, Muneera Altayeb4
1
Al-Ahliyya Amman University, Jordan, a.ghraibah@ammanu.edu.jo
2
Private Orthopedic Surgery, Jordan, mmgh1990@gmail.com
3
Medical Engineering Private Sector, Jordan, was2015eem@gmail.com
4
Al-Ahliyya Amman University, Jordan, m.altayeb@ammanu.edu.jo

ABSTRACT
per year. Several accidents require health care experts
to analyze a huge number of X-ray images and diagnose
Bones are protecting many organs in the human body such
patients with the right decision. So, cases of false diagnosis
as the lungs, brain, heart and other internal organs. Bone
may occur, where a false diagnosis is defined as either
fracture is a common problem in human beings and may
failure to see a significant fading or attaching the incorrect
occur due to the high pressure that is applied to the bones as
diagnosis that is readily seen. A higher false diagnosis rate
a result of an accident or any other reasons. X-ray
will result in weak quality in healthcare and time-delayed
(radiograph) is the noninvasive medical experiment that
treatment [1].
helps doctors diagnose and present medical conditions. X-
rays represent the oldest and most often used kind of
medical imagery. Medical image processing attempts to In 2015, Anu et al [3] extracted features using the Gray-
enhance the bone fracture diagnosis processes by creating an Level Co-occurrence Matrix (GLCM) from X-ray bone
automated system that can go through a large database of the images. Then, the extracted features were used to calculate
X-ray images and identify the required diagnosis faster and the segmented image and based on these features the bone is
with high accuracy than the regular diagnosis processes. In classified whether it has a fracture or not. The presented
this paper, the lower leg bone (Tibia) fracture is studied and methods in [3] work have been tested on a data set with 40
many novel features are extracted using various image X-ray and CT images containing fractured and non-fractured
processing techniques. The purpose of this research is to use (normal) images. In the feature extraction process, the
new investigated features and classify the X-ray bone GLCM features were extracted and the images were
images as a fractured and non-fractured bone and make the classified depending on that into normal and fractured
system more applicable and closer to the user using the images. The performance of their work reached 85%
Graphical User Interface (GUI). The Tibia bone fracture accuracy. The limitation of the presented method was in
detection system was developed in three main steps: the using CT, and some cases of X-ray images, as it was very
preprocessing step, feature extraction using wavelet analysis, problematic to find the area of the fracture [3].
gradient analysis, principal components (PCA), and edge
detection methods and classification using Support Vector Al-Ayyoub and his colleagues (2013) [4] considered the
Machine (SVM). The results were produced using four binary classification problem to investigate the existence of
possible outcomes from the confusion matrix which are TP, the fracture in the hand X-ray images. The dataset consisted
TN, FP, and FN. The classification process was repeated of 98 X-ray images. They focused on Decision Tree (DT),
using different feature groups at a time and the resultant Bayesian Network (BN), Naive Bayes (NB) and Neural
accuracies were ranged between 70%-80%. Networks (NN) methods. Furthermore, as three sets of
features were computed in their work, separate experiments
Key words: Bone fracture detection, Classification of bone were conducted to evaluate which set is more useful by
fracture, Features extraction, Support Vector Machine using each set of features individually in the classification
(SVM), Tibia bone X-ray images. process. But the results were far from perfect and they found
that one way to improve the performance of base classifiers
1. INTRODUCTION is to combine all features and use them in classification, also
they tried to use two-level meta-classifiers as it turned out
Tibia fractures are the most common long bone fracture that it gave the best classification results. In the final stages,
accounting for more than 20% occupancy of hospital wards. different sets of features were used in the classification
On average 26 tibia fractures occur per 100,000 populations process, but the maximum accuracy level was 91.8% which
was obtained by applying boosting and then bagging on the

1494
Amani Al-Ghraibah et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May – June 2021, 1494 – 1500

Bayesian Network classifiers with the feature set that In Myint et al (2018) work [8], suggested a Computer-Aided
included: Wavelets, Curvelets and GLCM features [4]. Diagnosis (CAD) technique to automatically recognize and
In Umadevi and Geethalakshmi work [5], twelve features of localize the leg bone fracture. Many image processing
two groups; the shape features and texture features were techniques were used in their paper, they recognized that
used. The texture features that were extracted from long Harris corner detection was an effective tool to catch the
bones X-ray images are Gray Level Co-Occurrence Matrix broken points of the leg bone. Decision Tree (DT) was used
(GLCM) features. While the shape features were extracted as a simple classification for fracture and non-fracture bones.
using a Fast Hough Transformation algorithm. The accuracy KNN was also used as it is suitable for pattern recognition
of a single classification was evaluated using 10-fold cross- and supports to classify fracture types. Fracture types such
validation technique. Three binary classifiers, SVM, BPNN, as Transverse, Oblique, Comminuted and Normal were
and KNN were presented to build ensemble classification classified by the system too. The performance accuracy
models. The classifiers were built with different feature sets concerning fracture and non-fracture were calculated and the
and the presented experiments proved that a group of models accuracy assessment was also evaluated. Kappa accuracy
can significantly improve the quality of fracture assessment was used to consider the error results when
identification [5]. calculating the performance and classifying the types of
fractures. However, the system produces the output results
In 2012, Mahendran and his colleagues [6], focused on their with accurate and reliable performance and less processing
research to build an automated system that detect fractures in time based on the contributed methods. According to the
long bones from diagnostic X-rays using a series of result, best accuracy achieved was 83 % using the Kappa
progressive steps. Three classifiers were considered which accuracy assessment [8].
are: Back Propagation Neural Networks, Support Vector
Machine (SVM) and Naïve Bayes. Also, two feature classes In this paper, new features were extracted from the long
were collected and extracted from X-ray images, namely: bone X-ray images that were discussed in our previous work
texture and shape features, with a total of 11 features. In the [2]. An automated predicting system is built here to predict
classification part, four classifiers were merged and used to the existence of the bone fracture automatically and faster
classify the X-ray images as fractured or non-fractured than the regular diagnosis processes. So, the motivations of
images. The results proved that fusion classifiers were our project are to reduce human errors and reduce the time
efficient for fracture detection and reached a maximum and effort associated during the bone injury diagnostic
accuracy. One difficulty encountered with fusion process which is usually done manually by physicians.
classification was on detecting which classifier produces the Ultimately, this system can be integrated within the software
best accuracy. The researchers considered only simple of the x-ray imaging devices to allow users to produce a
fractures and the experimental results showed that the rapid and highly accurate diagnosis while generating the
performance reduces if the fractures were parallel or image. So, a Graphical User Interface (GUI) is designed
perpendicular to the bone edge [6]. which enables the user to perform interactive tasks.

Mahendran et al [7] used the texture features in bone The novelty of this work covers two main things: using a
fracture detection research. Fused classifiers were proposed large number of Tibia bone X-ray images and classifying the
for fracture detection where some specific classifiers were images based on new features which were investigated in
established and work as a binary classifier, which can report our previous work depends on the physical properties of the
if a bone fracture is detected or not. If detected, the location bone images [2]. Also, Graphical User Interface (GUI) is
of the fracture is highlighted. There are mainly three built to enable the user to perform interactive tasks easily.
classifiers: Feed Forward Back Propagation Neural Many classification processes are presented depending on
Networks (BPNN), Naïve Bayes (NB) and Support Vector the feature group used each time and a comparison between
Machine (SVM) classifier. The fusion classifiers built from the results is also performed.
base classifiers which are (1) Texture features with BPNN,
(2) Texture features with SVM, (3) Texture features with This paper is presented as follows: Section 2 explains briefly
NB and (4) Texture features with BPNN, SVM, and NB. the methodology of feature extraction and image’s
The basic fusion rule used was that if more than 2 classifiers classification. Section 3 shows the results of the
report fracture then the image is said to have a fracture. classification using different feature groups and summarizes
Many experimental works were accompanied to analyze the the graphical user interface (GUI) and Section 4 is the
performance of the proposed fusion classifier-based summary of this work.
detection system concerning its efficiency in terms of correct
detection and speed of the algorithm. After comparing the 2. METHODOLOGY
performance with the traditional single classification system,
the suggested unification of techniques revealed that the
results were improved in terms of accuracy in detecting 2.1 Feature Extraction
fractures and in the speed of the detection [7].
Recent work has analyzed and extracted new features from
Tibia bone X-ray images depending on the physical
1495
Amani Al-Ghraibah et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May – June 2021, 1494 – 1500

properties of the bone which detects the changes of these image is filtered (convolved) with the known Sobel filters,
features in the presence and absence of a fracture. Al- Gx and Gy which are given by: Gx = hx*f and Gy=hy*f
Ghraibah et al produced a novel method to examine whether respectively, where * is the two-dimensional convolution
these features are efficient in detecting bone fracture or not. operator, h is the filter and f is the image. To abbreviate this
They used the X-ray images of both legs of the patient (the gradient information into single descriptors for each image,
left and right Tibia bones), taking advantage of human body the following features were computed from the resultant
symmetry, to study the performance of the extracted features gradients from each image: mean, standard deviation,
in detecting bone fracture instead of directly using the maximum, minimum, skewness and kurtosis [10-12].
classification methods [2]. Here, we extend the work in [2]
and use the most significant features to build an automated From the results in [2], a summary was made that the
classification system that classify any X-ray Tibia bone gradient features were efficient in detecting bone fracture
image as a fractured or non-fractured image. and could be used in further bone classification processes.
But it was recommended to exclude: the mean, standard
Four different methods and techniques were used to describe deviation and minimum features, as they were less
the physical properties of the bones. These methods are: 1. significant in detecting the bone fracture. The remaining and
wavelet analysis, 2. gradient analysis, 3. Principle effective features are the maximum, skewness and kurtosis
Component Analysis (PCA) and 4. bone edge detection and will be used in this research.
method. The efficacy of the methods was presented, and the
results showed that, depending on the features changes in the 2.1.3 Bone edge features
presence of the fracture, the most significant features
extracted from each method were summarized. Here, a brief Edge detection defines a set of mathematical procedures that
description of each method and the most significant features aim to recognize points in a digital image where the
are presented. brightness varies sharply or has discontinuities. One of the
second-order derivative operators that is used for edge
2.1.1 Wavelet features detection is the Laplacian edge detector. It is from the zero-
crossing category of the edge detection technique and it
Two-Dimensional Discrete Wavelet Transform (2D DWT) gives better edge detection results than the first-order
was used in image processing as a powerful tool solving to derivative detection techniques, but it is somehow sensitive
image analysis, denoising, image segmentation and other. to noise [13]. In the previous work [2], the Tibia bone X-ray
2D DWT is computed when the original image is convolved images were preprocessed using a smoothing (average) filter
along x and y directions by low pass and high pass filters. to remove the unwanted signals, the image then was
The images obtained are downsampled by half the size of converted to a binary image and the canny edge detection
the original image. The resultant images are convolved again method was used to detect the bone edges. From the
with high pass and low pass filters. The four sub-band resultant image, two features were extracted, which are
images generated contain the approximation coefficient related to the edges of the bone; the sum of the on (white)
(which contain the maximum information of the image), pixels and the number of the 8-connected pixels in the
horizontal, diagonal, and vertical information of the image binary image. It was found that both features are effective in
[9,10]. detecting bone fracture and could be used in further bone
classification processes.
The three detailed images were used to evaluate the energies
of each decomposition level by adding the absolute values of 2.1.4 Principal component features
the wavelet coefficients (the highpass images). Then a sum
of these energies was calculated including the three highpass From a mathematical view, PCA is a statistical process that
images as we are interested in the edge structure. From that, uses an orthogonal transformation to convert a set of
five energy values were extracted which are related to each observations of probably correlated variables into a set of
of the five levels of decomposition namely: Energy level 1, values of linearly uncorrelated variables called Principal
Energy level 2, Energy level 3, Energy level 4, and Energy Components (PCs). This transformation is defined in such a
level 5. The results show that all the wavelet energies can way that the first PC has the largest possible variance, which
detect the existence of bone fracture in a good way and these measures for as much of the variability in the data as
features can be used in further bone classification processes possible. Then, each succeeding component in turn has the
[2]. highest variance possible under the constraint that it
is orthogonal to the preceding components [14]. Six features
2.1.2 Gradient features were extracted from the highest variance PCA which are:
mean, standard deviation, minimum value, maximum value,
From the general image processing science, the spatial skewness, and kurtosis. These six features can detect the
gradient is equivalent to the first derivative of the processed bone fracture effectively depending on the previous work
image. Gradient of the image will highlight fragments and results [2].
edges that may not be noticeable in the original image. The
1496
Amani Al-Ghraibah et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May – June 2021, 1494 – 1500

2.2 Classification or involve repeated rounds of them [16]. A 10-fold cross-

validation method is used in this work where the MATLAB

Table 1: Confusion matrix

Fractured Non-fractured
Fractured TP FP
Non-fractured FN TN

software completes these steps by randomly partition the

data into 10 sets of equal size and train the SVM classifier
on the remaining nine sets. The previous steps were repeated
10 times and at the end, the system combines generalization
Figure 1: General description of SVM statistics from each fold.

The purpose of supervised, machine learning is to build a 2.3 Performance evaluation

model that makes predictions based on evidence in the
presence of uncertainty. The machine learns from the results
when adaptive algorithms classify data patterns. The A connection between our university and King Abdullah
computer improves its prediction performance when University Hospital (KAUH) was settled and the data were
exposed to more observations. Specifically, a supervised collected from the orthopedic department there. The total
learning algorithm uses a known set of input data and known number of Tibia bone images used in this work are 100
responses of the data (classes) and trains a model to generate images for the evaluation purpose of which 50 are with a
reasonable predictions for the response to new data. In this fracture while the rest are normal images. The terms used in
work, the input X-ray image that has a fractured bone the confusion matrix (shown in Table 1) can briefly be
addressed class 1 and the image without a fractured bone described as: True Positive (TP): true decisive system
addressed class 0. Figure 1 shows a simple description of classified as true, True Negative (TN): false event detected
SVM. as false, False Positive (FP): the event is false and
discriminated as true and False Negative (FN): true event
classified as false [17].
The whole set of input data can be called a heterogeneous
matrix as the matrix rows are called observations or
instances and each of them includes several measurements Also, Accuracy (AC) is defined as the probability that the
for a subject. Matrix columns are denoted to as predictors or classification by the system is correct and it is given by (2)
features and each of them represents a measurement of each [20]:
subject [15]. In this research, the observations are the X-ray
images of the Tibia bones where the features of each image +
= ∗ 100 (2)
are set in columns. The data matrix contains one row of + + +
features extracted from each image as given in (1).
The Sensitivity (True Positive Rate (TPR)) and Specificity
(True Negative Rate (TNR)) are also calculated from the
confusion matrix using (3), and (4) respectively [20]:

(1)
= (3)
+
where n is the number of features extracted and m is the
number of images.All supervised learning methods start with
an input data matrix. The data were prepared as each row in = (4)
+
the feature matrix represents one observation and each
column represents one variable or predictor. In this step the
3. RESULTS
features were extracted from each X-ray image in the data
set and arranged in one matrix for each image and called a
data matrix. Each row has ten features related to the ten In this section, the results are presented for the following
features described before. experiments: classification using one feature analysis at a
time, classification using all feature analyses, classification
Cross-validation is a statistical method of calculating and using two or three feature analyses each time. Also, a
comparing learning algorithms by dividing data into two Graphical User Interface (GUI) is presented which was built
segments: the first segment used to train a model and the to classify any new Tibia bone X-ray image into a fractured
other one is to validate the model. The basic form of cross- or non-fractured image automatically using any of the
validation is k-fold cross-validation, while other forms of existence features group of the user choice.
cross-validation are special cases of k-fold cross-validation
1497
Amani Al-Ghraibah et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May – June 2021, 1494 – 1500

3.1 Classification using one feature analysis at a time Table 7: Confusion matrix using all features
Fractured Non-fractured
Table 2: Confusion matrix using gradient features Fractured 38 12
Fractured Non-fractured Non-fractured 15 35
Fractured 45 5
Non-fractured 23 27 3.3 Classification using group of three or two analyses

Table 3: Confusion matrix using wavelet features Other groups of features are used to build the classification
Fractured Non-fractured model and classify the images as a fractured and non-
Fractured 46 4 fractured image. Table 8 shows the features groups which
Non-fractured 16 34 are used in each process which contains a collection of the
features that are extracted using three analyses out of the
The result confusion matrix of gradient analysis is shown in four presented analyses. The reason behind using different
table 2, where the accuracy (AC) is equal to 72%, TPR is features groups is to find the best collection of the extracted
66.2% and TNR is 84.4%. While the confusion matrix of the features that could give better classification accuracy. From
wavelet analysis is presented in table 3 and the resultant AC table 8 the results show that the best feature group in
is 80%, TPR is 74.2% and TNR is 89.5%. From these detecting the bone fracture is the group of Wavelet, Edge
results, the wavelet features are more effective than the and PCA features, which means all features except the
gradient features in detecting the bone fracture. gradient features. While the group who gives a balance
accuracy (AC) with TPR and TNR is the group of Gradient,
The confusion matrix results after using the edge features is Wavelet and PCA features (all features except the edge
presented in table 4, where the calculated accuracy AC is features). So, the later feature group can detect the fractured
equal to 72%, TPR is 73.8% and TNR is 67.2%. Also, the and non-fractured images with the same accuracy.
confusion matrix of using the PCA features is shown in table
5; the AC is 71%, TPR is 68.3% and TNR is 77.5%. From Table 9 presents the results while using a group of two
the previous results, the accuracy of detecting the bone feature analyses one of them is the wavelet analysis. In
fracture using the wavelet analysis is higher than using the general, the resultant accuracies are somehow close to each
other features: gradient, edge and PCA features. Table 6 other with maximum accuracy (AC) is reached using
presents a summary of the classification results while using wavelet and edge features together with high TNR result ~
each analysis alone in the classification process. As 90%.
mentioned above the wavelet analysis gives the best
classification accuracy (80%) and it shows the best TPR and 3.4 Graphical User Interface (GUI)
TNR results too.
The classification system can be integrated within the
3.2 Classification using all features software of the X-ray imaging devices so the users can
diagnose the images quickly and accurately during image
The result confusion matrix using all features is shown in generation. Graphical User Interface (GUI) is a graphical
table 7. Where the accuracy (AC) is equal to 73%. While the display in one or more windows containing components, that
TPR and TNR are equal to 71.7% and 74.5%, respectively. enable the user to perform interactive tasks. The reason
behind designing an application using a graphical interface
Table 4: Confusion matrix using edge features is to make the system more applicable and friendly interface
Fractured Non-fractured [18]. In this work, the designed GUI lets physicians to choo-
Fractured 41 9
Non-fractured 19 31 Table 8: Summary of classification results using three analysis
each time
Table 5: Confusion matrix using PCA features Gradient, Gradient, Gradient, Wavelet,
Wavelet Wavelet Edge & Edge &
Fractured Non-fractured
& Edge & PCA PCA PCA
Fractured 35 15
Non-fractured 14 36 Accuracy 73% 76% 71% 77%
(AC)
Table 6: Summary of classification results using each analysis
TPR 68.3 % 76% 69.8% 74.5%
alone TNR 81.1% 76% 72.3% 80%
Gradient Wavelet Edge PCA
Accuracy (AC) 72% 80% 72% 71%

TPR 66.2% 74.2% 73.8% 68.3%

TNR 84.4% 89.5% 67.2% 77.5%
1498
Amani Al-Ghraibah et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May – June 2021, 1494 – 1500

Table 9: Summary of classification results using two analysis each

time

Gradient & Edge & PCA &

Wavelet Wavelet Wavelet
Accuracy 77% 80% 77%
(AC)
TPR 70.7% 74.2% 73.7%
TNR 88.5% 89.5% 81.4%

-se the feature extraction method and then show the resultant
features after using the selected feature method and finally
classify the image into a fractured or non-fractured. Figure 2
Figure 2: GUI example; classifying the input image using wavelet
shows an example of classifying an X-ray Tibia bone image
analysis.
using the wavelet analysis method, the image on the left is
the original image selected randomly from the data, the bar
REFERENCES
next to the image shows an options for the available feature
extraction methods where the user can select the type of the 1. N. Umadeviand S. N. Geethalakshmi. Multiple
method by click on the method name. classification system for fracture detection in human
bone x-ray images. In 2012 Third International
The images next to the feature methods bar represent the Conference on Computing, Communication and
image at the last step of image processing which then used to Networking Technologies (ICCCNT'12), IEEE, 2012,
extract the wanted features. The last bar on the right is the pp. 1-8.
features that were extracted and used in the classification 2. A. Al-Ghraibah, M. Algharaibeh, W. Al-Muhtaseb, F.
process. The image class (fractured or non-fractured) is Al-Khateeb, and I. Al-Anis. Investigating the
presented at the bottom. Significance of New Features Extracted From Long
Bones X-ray Images. 2020IEEE 5th Middle East and
Africa Conference on Biomedical Engineering
(MECBME) rank. IEEE, 2020.
4. CONCLUSION 3. T. C. Anu and R. Raman. Detection of bone fracture
using image processing methods. International
Recent work of Al-Ghraibah et al has analyzed new features Journal of computer applications, 975, pp. 8887, 2015.
from Tibia bone X-ray images depending on the physical 4. M. Al-Ayyoub, I. Hmeidi and H. Rababah. Detecting
properties of the bone which detects the changes of these Hand Bone Fractures in X-Ray Images. J. Multim.
features in the presence or absence of a fracture. They Process. Technol., vol. 4, no. 3, pp. 155-168, 2013.
produced a novel method to examine whether the extracted 5. N. Umadevi, and S. N. Geethalakshmi. .Multiple
features are efficient in detecting bone fracture or not. In this classification system for fracture detection in human
work, an extension to the work in [2] is presented and is bone x-ray images. In 2012 Third International
used the most significant features to build an automated Conference on Computing, Communication and
classification system that classifies any X-ray Tibia bone Networking Technologies (ICCCNT'12), IEEE, July
image as a fractured or non-fractured image. Four different 2012, pp. 1-8.
methods were used, namely: wavelet analysis, gradient 6. S. K. Mahendran and S. S. Baboo. Ensemble Systems
analysis, Principle Component Analysis (PCA) and bone for Automatic Fracture Detection. International
edge detection method. Classification process was repeated Journal of Engineering and Technology, vol. 4, no.1,
using each feature analysis group alone, using all feature pp. 7, 2012.
analyses groups together and using two or three feature 7. S. K. Mahendran and S. S.Baboo. An enhanced tibia
analyses groups at a time. Also, a Graphical User Interface fracture detection tool using image processing and
(GUI) was presented to classify any new Tibia bone X-ray classification fusion techniques in X-ray
image into a fractured or non-fractured image automatically images. Global Journal of Computer Science and
using any of the existence features group of the user choice. Technology, vol. 11, no.14, pp. 22-28, 2011.
The maximum accuracy result was reached when the 8. W. W. Myint, K. S. Tun and H. M. Tun. Analysis on
wavelet features are used alone in the classification process Leg Bone fracture detection and classification using
or if they were used along with the edge features too. X-Ray images. Machine Learning Research, vol. 3, no.
3, pp. 49-59, 2018.
9. R. C. Gonzalez and R. E. Woods. Image
processing. Digital image processing, 2nd ed.,2007, ch.
7.

1499
Amani Al-Ghraibah et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May – June 2021, 1494 – 1500

10. A. Al-Ghraibah, L. E.Boucheron and R. T. J.McAteer.

An automated classification approach to ranking
photospheric proxies of magnetic energy build-
up. Astronomy & Astrophysics, vol. 579, no. A64, 2015.
11. R. C. Gonzalez, S. L. Eddins and R. E. Woods. Digital
image publishing using MATLAB. Prentice Hall.
2004.
12. S. Gupta and S. G. Mazumdar. Sobel edge detection
algorithm. International journal of computer science
and management Research, vol. 2, no. 2, pp. 1578-
1583, 2013.
13. R. Anas, H. A. Elhadi and E. S. Ali. Impact of Edge
Detection Algorithms in Medical Image
Processing. World Scientific News, vol. 118, pp. 129-
143, 2019.
14. I. T. Jolliffe and J. Cadima. Principal component
analysis: a review and recent
developments. Philosophical Transactions of the Royal
Society A: Mathematical, Physical and Engineering
Sciences, vol. 374, no. 2065, pp. 20150202, 2016.
15. W. L. Chao. Machine learning tutorial. Digital Image
and Signal Processing, 2011.
16. P. Refaeilzadeh, L. Tang and H. Liu. Cross-
Validation. Encyclopedia of database systems, vol. 5,
pp. 532-538, 2009.
17. A. Tharwat. Classification assessment
methods. Applied Computing and Informatics. Vol.
ahead-of-print, No. ahead-of-print, 2020.
18. M. Kalechman.. Practical MATLAB basics for
engineers. Crc Press, 2018.

1500