Performance Evaluation of State-of-the-Art Texture Feature Extraction Techniques on Medical Imagery Tasks

doi:10.21203/rs.3.rs-2452334/v1

Download PDF

Research Article

Performance Evaluation of State-of-the-Art Texture Feature Extraction Techniques on Medical Imagery Tasks

https://doi.org/10.21203/rs.3.rs-2452334/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Interpreting medical images is certainly a complex task which requires extensive knowledge [1]. According to [2]Computer Aided Diagnosis (CAD) serves as a second opinion that will help radiologists in diagnosis and on the other hand Content-based Image Retrieval uses visual content to help users browse, search and retrieve similar medical images from a database based on the user’s interest [3] [4]. The competency of the CBMIR system depends on feature extraction methods [5]. The textural features are very important to determine the content of a medical image. Textural features provide scenic depth, the spatial distribution of tonal variation, and surface orientation [6]. Therefore, this study seeks to compare and evaluate some of the hand-crafted texture feature extraction techniques in CBMIR. This is to help those concerned in enhancing CBIR systems to make informed decisions concerning the selection of the best textural feature extraction techniques.

Since there is no clear indication of which of the various texture feature extraction techniques is best suited for a given performance metric when considering which of the techniques to choose for a particular study in CBMIR systems. The objective of this work, therefore, is to comparatively evaluate the performance of the following texture feature extraction techniques; Local Binary Pattern (LBP), Gabor Filter, Gray-Level Co-occurrence Matrix (GLCM), Haralick Descriptor, Features from Accelerated Segment Test (FAST) and Features from Accelerated Segment Test and Binary Robust Independent Elementary Features (FAST &BRIEF) using the metrics; precision, recall, f1-score, mean squared error (MSE), accuracy and time. These techniques are coupled with specific similarity measure to obtain results.

The results showed that LBP, Haralick Descriptor, FAST, and GLCM had the best results in terms of (Precision and Accuracy), Time, F1-Score, and Recall respectively. LBP had 82.05% and 88.23% scores for precision and accuracy respectively. The following scores represent the performance of the Haralick descriptor, FAST, and GLCM models respectively; 0.88s, 38.7%, and 44.82%. These test scores are obtained from datasets ranging from 1k-10.5k.

Aside from LBP outperforming the other 5 models mentioned, it still outperformed the following proposed models ' [7]’, ‘Tamura texture feature and wavelet transform combined with Hausdorff distance - [8]’, ‘ [9]’ in terms of (precision, accuracy, and recall) and (precision and recall) respectively and probably f1-score (since f1-score is the weighted average of precision and recall). It is believed that an ensemble of LBP, Haralick descriptors, and Support Vector Machine (SVM) can represent a robust system for both medical image retrieval and classification.

Content-based image retrieval

Medical Image Retrieval

Content-based Medical Image Retrieval

Texture Feature Extraction

Evaluation Metrics

Similarity Measurement

Technology has advanced in numerous ways and most computer scientists are eager to bridge the gap between the capabilities of mankind and machines. This is to say that, these scientists want these machines to perceive and use the knowledge obtained in multitasking. This advancement has led to the development of self-driving cars, delivery drones, etc. and these are realized through a subfield of computer science known as Artificial Intelligence (AI). Computer science’s field of artificial intelligence offers a wide range of applications that have produced some truly remarkable 21^st-century technologies. One of those areas this work focuses on is Computer Vision. The primary aim of Computer vision (CV) is to make it possible for machines to perceive the world as humans do, including the information gained through performing tasks like viewing Media Recreation, Image and Video Recognition, Image Analysis and Classification, etc. some factors that have led to the evolution or rebirth of computer visioning includes;

The need to share moments or events with distant family and friends led to the saturation of photos and videos all over the world using built-in cameras in technology such as smartphones.
Recent enabling design of promising hardware with high computational power which is readily available for tasks such as computer visioning and image analysis.
Availability of algorithms that take advantage of the hardware and software capabilities of recent computers.

[10] Claims that thanks to the work of researchers in the field of computer vision, accuracy rates for object detection and classification have increased from 50 percent to 99 percent in less than ten years. Due to the immense research done in the area of computer vision, sectors such as transportation, healthcare, manufacturing, construction, agriculture, retail, etc. have benefited a lot from the various tasks of computer vision. That is from self-driving cars, medical imaging data processing, defect inspection or reading products barcode, predictive maintenance, crop, and yield monitoring, or plant disease detection to automatic replenishment of goods or items in real-time respectively [11].

However, in the healthcare sector, areas such as X-rays, Dermatology, Radiology, Cardiology, High-Resolution Computed Tomography (HRCT), Endoscopy, and Magnetic Resonance Imaging (MRI) heavily make use of medical digital images and by so doing managing and accessing these medical images from their repositories has become more complex [12]. Concerningly, content-based or text-based, or even both methods have been put forth by various scholars to work on medical image retrieval systems. Based on a content-based image retrieval system, its relevancy is judged by the kind of feature extraction technique used. Therefore, conducting this comparative analysis on some of the cutting-edge texture feature extraction methods of Content-based Medical Image Retrieval - CBMIR does suffice.

The paper is in six (8) sections; section (1) introduces the content together with the background study which conclusively indicates the objectives of the research. Section (2) talks about the related works. Section (3) talks about the methodology carried out to achieve the said results in section (5). Section (4) provides the code and data source. Section (5) provides the results, analysis and discussion. Section (6) provides summary points of this study. Section (7) discloses any conflict of interests towards carrying out this research. Section (8) acknowledges the bodies that provided support to completing this research successfully together with the referencing materials and their source.

Background: A texture-based image is built on visual patterns that have uniformity characteristics, not on the presence of a single color or intensity [1]. The texture of tissues in medicine is meant to be uniform and constant. As a result, due to the accurate characterization of the texture properties that are seen in organs and other tissue anomalies, computer scientists frequently use texture features in the medical field. [13] [14]. In that texture features, descriptors have been applied in various diagnoses such as brain tumor diagnosis [15] and mass lesion detection in mammogram images [16]. Thus texture-based feature has been the best feature descriptor for easy discrimination amongst organ tissues in the CBIR system [1]. However, certain research on various CBMIR components undertaken by some computer scientists has indicated an improvement in the overall efficiency of the CBIR system. To enhance picture retrieval, some of these works concentrate on the feature extraction method and the combination of various feature descriptors [1]. Some of these researchers used texture and shape features descriptors to represent the content of the images based on the idea that medical images, such as breast cancer [17] [14] [18] and brain tumors except for ophthalmology, pathology and dermatology images that are typically in grayscale channels [19]. Contrary to other feature descriptors like texture and shape, color features are used outside of the medical field since they are the most widely known feature to represent image content [20] [21]. However, some of these works compared a few texture approaches on a limited dataset such as in [9], [7], and [8] making it difficult for practitioners to make an informed choice on which of the various texture-based feature extraction approach to select at some point in time.

Therefore, this study aims to identify which cutting-edge texture feature extraction approaches can ensure the stated level of invariance concerning CBIR in medicine.

The general objective is to comparatively evaluate the performance of the various texture feature extraction techniques for medical imagery retrieval.

The following includes the specific objectives of this study:
What state-of-the-art texture extraction method can be efficiently and effectively used for Content-Based Medical Image Retrieval (CBIR)
Evaluate the strength and weaknesses of the identified texture feature extraction techniques using the computational cost (time and space) as well as retrieval precision ad accuracies

With respect to the increasing amount of medical imagery data has resulted in widespread dissemination of Picture Archiving and Communication Systems (PACS) in hospitals which requires a more efficient and effective retrieval methods for better management of such data. There are currently two ways of retrieving medical images [22]. These retrieval methods includes; in Text-Based Image Retrieval and Content-Based Image Retrieval.

Text-based Medical Image Retrieval System refers to the technique that involves querying an image dataset with a text-based search. This system can be traced back in the 1970s. This system is prevalent in online web search.

Content-based image retrieval system makes use of image indexing based on their visual content i.e. (image features such as color, texture, shape etc.). With respect to medical images, texture features are useful because they reflect the details within an image structure. Initial contributions has it as medical images been included as a subdomain for trials in the content-based image retrieval systems [23] [24].

The table below shows a summary report of some Texture Feature Extraction Techniques used in CBMIR

Experimental Setup

System requirements for the various models implemented require a minimum system specification to be met for a successful study. The specifications are categorized into Hardware and Software specifications.

3.1 Hardware Specification

The following includes the minimum hardware requirement for implementing the various learning models. This is to ensure the smooth execution of each model.

RAM of 8GB and above to ensure the system doesn't crash whiles using a large volume of the dataset
GPU with 2GB and above to enhance the computational power (this requirement is optional)
Storage Capacity of at least 512GB HDD or 256GB SSD

3.2 Software Specification

The software requirements involve programs that support the successful implementation of the various models. The following includes the generic software requirements:

OS: various up to date operating systems such as Windows 10, macOS, Ubuntu, etc. to provide an enabling environment for both the online and offline implementation
Python programming language preferably 3.7 and above

3.2.1 Online Software Specification

Two known online software programs support the implementation of these models namely:

Google Colaboratory also known as Google Colab
Kaggle

3.2.2 Offline Software Requirement

The offline requirements include either of the following known programs:

Pycharm IDE
Jupyter Notebook

3.3. Similarity Measurement

The objective of measuring the distance between two known objects is to obtain a summarized score between the two given objects in a problem domain. Finding or measuring the distance between the query image and the target image in this scenario is the study's primary goal to get an appropriate result (s). The following includes commonly known supervised learning distance measurement methods:

Euclidean Distance
Hamming Distance
Chi-Squared Distance

A summary of each model with the kind of similarity measurement is in the table below.

3.4. Threshold Value

Predicting a class label has always been the main aim of classification predictive modeling. It, therefore, means that before such a model can map to a crisp class label, the machine algorithm has to sort to predict a probabilistic scoring of class membership. It is as important of predicting a particular class label as to identify the best threshold for mapping.

A threshold value of 0.5 can be used as a mapping scheme, according to [25], where all values equal to or greater than the threshold are mapped to one class and all other values are assigned to another class.

He further stated that this default value (0.5) can create a severe classification problem which can result in poor performance using an imbalanced dataset. The researcher should either manually adjust the threshold value to an ideal number or sort using the Receiver Operating Characteristic (ROC) curve and the Precision-Recall Curves, according to his suggestion. The former was carried out and utilized to establish the threshold value for every model. The threshold value for each of the implemented models may be found in the list below.

Local Binary Pattern – 0.015

Gray-Level Co-occurrence Matrix – nine (9) added to the least computed distance value.

Gabor Filter – 0.050

Haralick Texture Descriptor – 0.050

FAST – 25 added to the least computed distance value.

FAST-BRIEF – values greater than or equal to 3

3.5 Conceptual Framework

The conceptual framework is important for the researcher to clearly present and explain ideas and consequently potential relationships pertaining to the study [26]. One way of presenting such concepts is by a visual model that can assists readers by illustrating how the processes or constructs in the research work. This is known as a graphical conceptual framework. The diagram below represents the graphical conceptual framework of this study;

4.1. Code

The code for the six selected texture feature extraction techniques together with the dataset can be freely accessed and executed through Code Ocean via the URL: https://codeocean.com/capsule/975181/tree

The Source code can also be found on GitHub via the URL: https://github.com/SamKDuah/Content-based-Medical-Image-Retrieval

4.2. Data

4.2.1. Dataset Validation

Data processing involves preprocessing. Data preprocessing refers to the technique of preparing (i.e. cleaning and organizing) the raw data to make it suitable for accomplishing a given task. Data preprocessing is a data mining technique that transforms raw data into an understandable and readable format for a given model. This process helps enhance the quality of data to promote the extraction of meaningful insights from the data. The following includes the steps taken in preprocessing the data for the various learning model:

Acquisition of dataset
Uniformity of dataset (i.e. Resizing the data to a specific dimension and Converting the dataset to a grayscale channel if the dataset contains data of different color channels)

4.2.1.1. Dataset Acquisition

A total of 58,954 medical data images were obtained from the web, out of which 1k - 10.5k of the dataset was used. The data below represents the metadata of both the dataset source and the dataset itself.

The dataset license is available to the public domain. Below represents the metadata of the dataset source:

@misc{Medical MNIST Classification,
author = {apolanco3225},
title = {Medical MNIST Classification},
year = {2017},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/apolanco3225/Medical-MNIST-Classification}}
}

License

Public Domain

Below represents a sample image of each class of the dataset.

4.2.1.2. Dataset Uniformity

Python libraries such as "PIL" and cv2 contain functions for resizing images to a preferred dimension. Notwithstanding, the cv2 library also supports the conversion of image color between channels such as RGB, Grayscale, HSV, and Hue.

The table below shows the summary of the results obtained from the experiment using a dataset ranging from 1K – 10.5K.

To significantly tell or evaluate the performance of each model, there should be a general basis for comparison. Therefore, finding the mean average for the aforementioned metrics can serve as the basis for comparison. Since there are four (4) categories of the dataset, which is 1K, 5K, 10K, and 10.5K we find the mean average for the models per the metrics.

The table below represents the mean average result score of each model per the unlisted evaluation metrics. That is to say, the mean of the average scores per metric for each model is calculated to provide the indicator basis for comparison. For easier interpretation, the percentage of the mean average scores is shown in the table below.

The diagrams below illustrate the graphical representation of the results obtained

Analysis

Interpretation of the results above;

Precision: it is noticed that the models did averagely well as the volume of the dataset increased. LBP had the best average precision throughout the various volumes of datasets hence having the best outstanding mean average precision score.

Recall: it is also noticed that some of the models increased proportionally to increasing volumes of dataset yet their recall values in question were below average which means that the number of relevant or similar images to be retrieved from the database are reduced. GLCM performed better in terms of recall. It was the best so far in terms of recall concerning the increasing volumes of the dataset.

F1-score: LBP had a well-balanced score in terms of precision and recall. That is f1-score is the weighted average for both precision and recall. In simpler terms, it is the weighted average for a given precision and recall value.

Mean Squared Error/Accuracy: these two move hand-in-hand, meaning an increase in error decreases accuracy and vice versa. From the results above, it is observed that errors increased proportionally to increasing volumes of data. Generally, the models did well in minimizing the errors or otherwise did well in optimizing their accuracies. LBP did very well in optimizing accuracy by minimizing errors hence it had the best score in terms of accuracy.

Time: it is observed that time is proportional to the volume of the dataset, meaning as the dataset increases computational time also increases. Although computational time can be relative in terms of several factors such as coding behavior, device implementation specifications, etc. according to [27] wall time also known as clock time or wall-clock time refers to the total time elapsed during measurement, which includes the difference between the time a program in execution finishes and its start time together with its waiting time for resources if required by the given program. Irrespective of the increasing volumes of the dataset haralick descriptor had the least and best mean average computational time.

Model Strengths

Local Binary Pattern: best in terms of Precision and Accuracy

Haralick Descriptor: best in terms of Time

FAST: according to F1-Score, FAST is the best

Gray-Level Co-occurrence Matrix: best according to Recall

Concerning Table 1: summary report on some texture feature extraction techniques in CBMIR, neglecting or assuming various experiments used the same volume of datasets, from

Table 6: Comparison between the summary report of some existing texture feature extraction techniques used in CBMIR and LBP above we could intuitively say that;

Observably, LBP continues to lead in terms of precision and recall and is likely to lead in terms of f-score as well because it is the weighted average of precision and recall. Lastly, it is the 2^nd to the New Gradient Descriptor from Co-occurrence Matrix in terms of accuracy.

It can be concluded from this study that the Local Binary Pattern is best suited for research that focuses on precision, f1-score, and apparently accuracy for Content-Based Medical Image Retrieval systems.

Haralick Texture Descriptor is best suited for research that focuses more on computational time concerning medical imagery tasks.

Nonetheless, FAST performed better than FAST_BRIEF concerning the precision, recall, and f1-score.

In this study, the models were adapted for CBMIR systems. The resulting values in chapter 4 were obtained after running each model several times over a specific volume and class of dataset. Though the dataset was cleansed during the preparation stage, some images having “Not a Number” (NaN) value whiles extracting its features could not be neglected since it is part of the medical images, and as such the models did quite well when they encountered such challenge, especially the GLCM model. This is why it was able to recall the highest number of images though most were not relevant but had a recall of 44.82% over the range of 1k – 10.5K volume of the dataset.

It should be noted that the results obtained for the basis of comparison represents the mean average scores over the dataset ranging from 1K-10.5K. The idea for using the mean average score was adapted from [28].

The study compares the performance of the six selected texture feature extraction techniques on the medical dataset.

Specifically, the performance metrics we are comparing are on the basis of Precision, Recall, F1-Score (Average of both the Precision and Recall), Mean Squared Error, Accuracy, and Time

The Dataset used is of raw medical image data, therefore there isn't much to do with dealing with outliers and the rest as if the dataset were in the form of an excel file but we rather catered for having a uniform dimension for the image data.

Each Technique had its own way of converting the raw image data to its array of codes. There were a few cases where we encountered a NaN value for the most apparent dark image(s). This issue was raised in the concluding part.

More so, the experiment was conducted separately for each technique using the same conditions, other than that, the records taken for the various execution time won't have reflected their actual time. Therefore, there is no need to apply any correction techniques for this experiment.

For each technique, the experiment was carried out with an increasing volume of dataset i.e. from 1k - 10.5k.

For each technique we took the average score (1k-10.5k dataset) for each performance metric. In the end, we compared the performance metrics of the techniques using the average scores in order to have a general or performance summary of the technique. This is illustrated in the supplementary material.

The Authors declare No conflicts of interest, financial or otherwise for this research.

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

☐ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Acknowledgment

First of all, I will like to use this opportunity to thank the Almighty God, the creator of the universe for giving me the strength and ability to accomplish this task. I will like to also express my profound gratitude to my supervisor – Dr. Obed Appiah and Co-Supervisor – Dr. Peter Appiahene (HoD Computer Science) for their immense contribution towards the completion of this study. I truly appreciate the time and guidelines they offered towards achieving a successful novel project. I say God bless them.

M. Z. Nor Asma, Y. Rozianiwati, A. L. Saima, M. Aida, S. Norhalina and I. Rosziati, "Content-Based Image Retrieval in Medical Domain: A Review," Journal of Physics: Conference Series, p. 13, 2018.
K. Belattar, S. Mostefai and A. Draa, "Intelligent Content-Based Dermoscopic Image Retrieval with Relevance Feedback for Computer-Aided Melanoma Diagnosis.," Journal of Information Technology Research (JITR), pp. 85-108, 2017.
C. B. Akgul, D. L. Rubin, S. Napel, C. F. Beaulieu, H. Greenspan and B. Acar, "Content-based image retrieval in radiology: current status and future directions.," Journal of Digital Imaging, pp. 208-222, 2011.
K. Bunte, M. Biehl, M. F. Jonkman and N. Petkov, "Learning effective color features for content based image retrieval in dermatology.," Pattern Recognition, pp. 1892-1902, 2011.
S. Amita, R. Amol and P. Chetankumar, "Content based medical image retrieval based on new efficient local neighborhood wavelet feature decriptor," Biomedical Engineering Letters, p. 8, 2019.
E. Burhan and B. Muhammet, "Texture based feature extraction methods for content based medical image retrieval systems," Bio-Medical Materials and Engineering, pp. 3055-3062, 2014.
R. I. Minu, B. Muthukumar, V. Vedanarayanan, S. D. Sundarsingh, G. Nagarajan and B. Muthukumar, "Hybrid Genetic Algorithm for Medical Image Feature Extraction and Selection," in International Conference on Computational Modeling and Security (CMS 2016), 2016.
X. Sun, N. Zhang, H. Wu, X. Yu, X. Wu and S. Yu, "Medical Image Retrieval Approach by Texture Features Fusion Based on Hausdorff Distance," Hindawi: Mathematical Problems in Engineering, p. 12. https://doi.org/10.1155/2018/7308328, 27 August 2018.
C. F. Joaquim, J. M. T. Agma and T. J. Caetano, "Retrieval by Content of Medical Images Using Texture for Tissue Identification," in Proceedings of the 16th IEEE Symposium on Comuter-Based Medical Systems (CBMS '03), 2003.
L. Simpson-Kent, "Computer Vision: An Overview," 03 03 2022. [Online]. Available: https://www.harnham.com/post/2022-3/computer-vision-an-overview.
R. Alberto, "Computer vision," 1 august 2022. [Online]. Available: https://www.v7labs.com/blog/computer-vision-applications#h2.
K. Palwinder and K. S. Dr. Rajesh, "Content-Based Imae Retrieval Using Machine Learning and Soft Computing Techniques," Int'l Journal of Scientific & Technology Research, p. 6, 2020.
C. H. Wei, Y. Li and P. J. Huang, "Mammogram retrieval through machine learning within BI-RADS standards.," Journal of biomedical informatics., pp. 607-614, 2011.
S. A. Lashari, I. T. R. Yanto, T. Herawan, R. Ibrahim and N. Senan, "Application of Wavelet De-noising Filters in Mammogram Images Classification Using Fuzzy Soft Set," in In International Conference on Soft Computing and Data Mining, 2016.
C. Traina-Jr, J. D. E. B. Neto, A. J. Traina, S. F. Da Silva and M. X. Ribeiro, "Improving the ranking quality of medical image retrieval using a genetic feature selection method.," Decision Support Systems, pp. 810-820, 2011.
B. Zhang, "Computer-aided diagnosis in mammography using content-based image retrieval approaches: current status and future perspectives.," Algorithms, pp. 828-849, 2009.
J. E. De Oliveira, A. M. Machado, G. C. Chavez, A. P. B. Lopes, T. M. Deserno and A. D. A. Araujo, "MammoSys: A content-based image retrieval system using breast density patterns.," Computer methods and programs in biomedicine, pp. 289-297, 2010.
D. A. Chandy, J. S. Johnson and S. E. Selvan, " Texture feature extraction using gray level statistical matrix for content-based mammogram retrieval," Multimedia tools and applications, pp. 2011-2024, 2014.
A. Depeursinge and H. Miller, "Medical visual information retrieval based on multidimensional texture modelling.," Procedia Computer Science, pp. 127-129, 2011.
A. Talib, M. Mahmuddin, H. Husni and L. E. George, "A weighted dominant color descriptor for content-based image retrieval," Journal of Visual Communication and Image Representation, pp. 345-360, 2013.
Z. C. Huang, P. P. Chan, W. W. Ng and D. S. Yeung, "Content-based image retrieval using color moment and gabor texture feature.," in In: International Conference on Machine Learning and Cybernetics (ICMLC), 2010.
Y. Liu, D. Zhang, G. Lu and W. Ma, "A Survey of Content Based Image Retrieval with high-level semantics.," Pattern Recognition, pp. 40:262-282, 2007.
P. Kellu and M. Cannon, "Query by image example: the comparison algorithm for navigating digital image databases (CANDID) approach," in In: Proceedings of the Storage & Retrieval for Image and Video Databases, 1995.
S. Orphanoudakis, C. Chronaki and S. Kostomanolakis, "A system for the indexing, Storage and retrieval of medical images by content.," Med Inform, pp. 19: 109-122, 1994.
J. Brownlee, "Machine Learning Mastery," 10 February 2020. [Online]. Available: https://machinelearningmastery.com/threshold-moving-for-imbalanced-classification/.
D. Waller, "how to develop graphical framework chart your research," 31 March 2022. [Online]. Available: https://www.timeshighereducation.com/campus/how-develop-graphical-framework-chart-your-research.
T.-T. Itamar, "python execution time of program," 1 October 2021. [Online]. Available: https://pynative.com/python-get-execution-time-of-program/.
Z. Wengang, L. Houqiang and T. Qi, "Recent Advance in Content-based Image retrieval: A Literature Survey," IEEE, pp. 14-22. http://arxiv.org/abs/1706.06064v2, 2017.
C. Wei, L. Yu, W. Weiping, M. B. Erwin, G. Theodoros, F. Paul, L. Li, M. I. Senior and S. L. Michael, "Deep Image Retrieval: A Survey," arXiv:2101.11282v2, 2021.
L. R. Nair, K. Subramaniam and G. Prasannavenkatesan, "A review on multiple approaches to medical image retrieval system," Intelligent Computing in Engineering, vol. 1125, pp. 501-509, 2020.
Y. Liu, D. Zhang, G. Lu and W. -. Y. Ma, "A Survey of content-based image retrieval with high-level semantics. Pattern Recognition," pp. 262-282, 2007.
K. Jenni, S. Mandala and M. S. Sunar, "Content based image retrieval using colour strings comparison.," Procedia Computer Science, pp. 374-379. https://doi.org/10.1016/j.procs.2015.04.032, 2015.
F. All and A. Hashem, "Content Based Image Retrieval (CBIR) by statistical methods.," Baghdad Science Journal, p. 694. https://doi.org/10.21123/bsj.2020.17.2(SI).0694, 2020.
C. Bal, J. Chen, L. Huang, K. Kpalma and S. Chen, "Sallency-based multi-feature modelling for semantic image retrieval.," Journal of Visual Communication and Image Representation, pp. 199-204. https://doi.org/10.1016/j.vcir.2017.11.021, 2018.
N. Shrivastava and V. Tyagi, "Corrigendum to "Content based Image retrieval based on relative locations of multiple regions of interest using selective regions matching".," [Information Sciences 259 (2014) 212-224] Information Sciences, 421, p. 273. https://doi.org/10.1016/j.ins.2017.09.017, 2017.
Z. Dengsheng, L. Guojun and L. Ying, "Region-based image retrieval with high-level semantics using decision tree learning.," Pattern Recognition, 41(8), pp. 2554-2570, 2008.
Z. Dengsheng, M. I. Md, L. Guojun and H. Jin, "Semantic image retrieval using region based inverted file," in In Proceedings of the 2009 Digital Image Computing: techniques and Applications, Washington DC, USA, 2009.
K. Alex, S. Ilya and E. H. Geoffery, "Imagenet classification with deep convolutional neural networks.," In Advances in neural information processing systems, pp. 1097-1105, 2012.
R. Datta, D. Joshi, J. Li and J. Z. Wang, "Image Retrieval," ACM Computing Surveys, 40(2), pp. 1-60. https://doi.org/10.1145/1348246.1348248, 2008.
D. P. Tian, "A review on Image feature extraction and representation techniques.," International Journal Multimedia Ubiquitous Engineering, pp. 385-395. ttps://gvpress.com/journals/IJMUE/vol8_no4/39.pdf, 2013.
B. S. Manjunath, J.-R. Ohm, V. V. Vasudevan and A. Yamada, "Color and Texture descriptors.," IEEE Transactions on circuits and systems for video technology, pp. 703-715. https://doi.org/10.1109/76.927424, 2001.
G. R. Cross and A. K. Jain, "Markov random field texture models.," IEEE Transactions on Pattern Analysis and machine Intelligence, PAMI-5(1), p. 25–39. https://doi.org/10.1109/TPAMI.1983.4767341, 1983.
R. M. Hawlick, "Statistical and Structural Approaches to Texture.," Advances in Intelligent Systems and Computing, 459 (5), pp. 1-644. doi: 10.1109/ PROC.1979.11328, 2017.
C. S. W. Won, D. K. P. Park and S.-J. P. Park, "Efficient use of MPEG-7 edge histogram descriptor," ETRI Journal, 24(1), p. 23–30. https://doi.org/10.4218/etrij.02.0102.0103, 2002.
T. Ojala, M. Pietikainen and T. Maenpaa, "Multiresolution gray-scale and rotation invariant texture classification with local binary patterns.," IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(7), pp. 971-987. https://doi.org/10.1109/TPAMI.2002.1017623, 2002.
M. Braschler and C. Peters, "Cross-language evaluation forum: Objectives, results, achievements Information Retrieval," pp. 7-31, 2004.
E. M. Voorhees and D. K. Harman, "TREC: Experiment and evaluation in information retrieval," Cambridge, MA, 2005.
C. Buckley and E. M. Voorhees, "Retrieval System Evaluation," in Voorhees EM, Harman DK Eds. TREC: Experiment and Evaluation in Information Retrieval., Cambridge, MA, 2005.
W. R. Hersh, "Advancing biomedical image retrieval: Development and analysis of a test collection.," J. Am Medical Information Association, pp. 488-496, 2006.
H. P. Chan, C. Vyborny, R. A. Schmidt and e. al., "Improvements in radiologists detection of clustered microcalcifications on mammograms.," Invest. Radiol. vol. 25, pp. 1102-1110, 1990.
F. Pinciroli and g. Pozzi, "ARCADIA: a system for the integration of Angiographic Data and images by Object-Oriented DBMS," Computer and Biomedical Research vol. 28, pp. 5-23, 1995.
H. MacMahon, H.-P. Chan and a. et, "Computer-aided diagnosis in chest radiology.," J. Thoracic Imaging vol. 5, pp. 67-74, 1990.
G. Mandloi, "A Survey on Feature Extraction Techiques for Color Images," International Journal of Computer Science and Information technologies, Vol. 5(3), pp. 4615-4620, 2014.
G. Dany and A. Reda, "Improving image Retrieval by Clustering," IGI Global, 2009.
"Confusion Matrix," 19 May 2022. [Online]. Available: https://en.wikipedia.org/wiki/Confusion_matrix.
L. Zheng, L. Shen, L. Tian, S. Wang, J. Wang and Q. Tian, "Scalable person re-identification: A benchmark," ICCV, pp. 1116-1124, 2015.
U. Chaudhuri, B. Banerjee and A. Bhattacharya, "Siamese graph convultional network for content based remote sensing image retrieval," Comput. Vis. Image Understanding, vol. 184, pp. 22-30, 2019.
Z. Liu, P. Luo, S. Qiu, X. Wang and X. Tang, "Deepfashion: Powering robust clothes recognition and retrieval with rich annotations," CVPR, pp. 1096-1104, 2016.
J. Bright, "BRIEF and FAST Feasture extraction," [Online]. Available: https://valueml.com/brief-and-fast-feature-extraction-python/.
C. Junyi, Z. Hao, L. Anming and W. N. Eric, "Deep learning in Computer Vision: A critical review of emerging techniques and application scenarios," Elsevier, p. 2, 2021.
Y. Guo, Y. Liu, A. Oerlemans, S. Lao, S. Wu and M. S. Lew, "Deep learning for visual understanding: A review.," Neurocomputing, pp. 27-48. https://dx.doi.org/10.1016/j.neucom.2015.09.116, 2016.
Y. Hang, Y. T. Laurence, Z. Qingchen, A. David and D. M. Jamal, "Convolutional neural networks for medical image analysis: State-of-the-art, comparisons, improvement and perspectives," Neurocomputing, p. 105, 2021.
C. Shyu, C. Brodley, A. Kak, A. Kosaka, A. Aisen and L. Broderick, "ASSERT: a physician-in-the-loop content-based image retrieval system for HR CT image databases," Computer Vis Image Inderstand, pp. 75: 111-132, 1999.
D. Keysers, J. Dahmen, H. Ney, B. Wein and T. Lehmann, "Statistical framework for model-based image retrieval in medical applications.," J Electron Imaging , pp. 12: 59-68, 2013.
T. Deselaers, "Features for Image Retrieval," Aachen, Germany: Rheinisch-Westfalische Technische, 2003.

Table 1

summary report on some texture feature extraction techniques in CBMIR
Source	Proposed Method	Results
[9]	A new gradient descriptor is created by creating a gradient vector from the co-occurrence matrix, whose indexes are values of \|i-j\| I and j are also co-occurrence matrix indexes), and whose content is the total number of occurrences in the elements with the same values of \|i-j\| in the co-occurrence matrix.	For queries looking for comparable photos up to 20% of the database, the retrieval accuracy for medical images was consistently over 90%.
[7]	A new and more effective feature selection approach in medical image retrieval system forming a hybrid technique. The "branch and bound algorithm", "Artificial bee colony algorithm" and relevance feedback make up this hybrid technique.	The final retrieval result showed that while the vast majority of pertinent photos were located, the diversified density relevance feedback system failed to locate certain glaringly pertinent photographs. They proposed that more sophisticated algorithms may be used to increase relevancy.
[8]	A method for retrieving medical images that use the wavelet transform algorithm, Tamura texture feature, and Hausdorff distance.	The proposed method outperforms a Tamura feature texture and wavelet transform technique paired with Euclidean distance, according to the experiments. In the brain MRI trials, the average retrieval accuracy of the suggested strategy is greater than that of the other five approaches by up to 19.33%, 17.33%, 18%, 11.33%, and 6%. The average retrieval accuracy of the suggested strategy is greater than the other three ways for lung CT studies, coming in at more than 17.33%, 16%, 16.66%, 12%, and 6.66%, respectively.

Table 2

Models and their adapted similarity measure
Model	Similarity Measurement
LBP	Euclidean Distance Measure
Gabor Filter	Chi-Squared
Gray-Level Co-occurrence Matrix	Euclidean Distance Measure
Haralick Descriptor	Euclidean Distance Measure
Fast	Euclidean Distance Measure
Fast & Brief	Hamming Distance Measure

Table 3

Dataset Metadata
SOURCE	NAME OF DATASET	CLASS	VOLUME OF DATA
https://www.kaggle.com/datasets/andrewmvd/medical-mnist?resource=download	Medical MNIST	AbdomenCT	10,000
https://www.kaggle.com/datasets/andrewmvd/medical-mnist?resource=download	Medical MNIST	BreastMRI	8,954
https://www.kaggle.com/datasets/andrewmvd/medical-mnist?resource=download	Medical MNIST	ChestCT	10,000
https://www.kaggle.com/datasets/andrewmvd/medical-mnist?resource=download	Medical MNIST	CXR	10,000
https://www.kaggle.com/datasets/andrewmvd/medical-mnist?resource=download	Medical MNIST	Hand	10,000
https://www.kaggle.com/datasets/andrewmvd/medical-mnist?resource=download	Medical MNIST	Head CT	10,000

Table 4

Summary of Overall Results
MODELS	PERFORMANCE EVALUATION METRICS
MODELS	PRECISION	RECALL	F-SCORE	ACCURACY	TIME/s
LOCAL BINARY PATTERN	0.8205	0.2448	0.3504	0.8823	10.87
GABOR FILTER	0.3666	0.0299	0.0492	0.8524	2.75
GRAY-LEVEL CO-OCCURRENCE MATRIX	0.5536	0.4482	0.2975	0.785	387.75
HARALICK	0.6507	0.042	0.0877	0.8613	0.88
FAST_BRIEF	0.405	0.0796	0.1244	0.8532	9.83
FAST	0.5536	0.3836	0.387	0.8008	15.54

Table 5: Percentiles of Mean Average Scores

		TEXTURE FEATURE EXTRACTION TECHNIQUES
		LBP	GABOR FILTER	GLCM	HARALICK	FAST	FAST_BRIEF
M E T R I C S	PRECISION %	*82.05*	*36.66*	*55.36*	*65.07*	*55.36*	*40.5*
	RECALL %	*24.48*	*2.99*	*44.82*	*4.2*	*38.36*	*7.96*
	F1-SCORE %	*35.04*	*4.92*	*29.75*	*8.77*	*38.7*	*12.44*
	MSE %	*11.78*	*14.76*	*21.5*	*13.87*	*19.93*	*14.68*
	ACCURACY %	*88.23*	*85.24*	*78.5*	*86.13*	*80.08*	*85.32*
	TIME/s	*10.87*	*2.75*	*387.75*	*0.88*	*15.54*	*9.83*

Table 6: Comparison between the summary report of some existing texture feature extraction techniques used in CBMIR and LBP

	PERFORMANCE EVALUATION METRICS
TEXTURE FEATURE EXTRACTION MODELS	Precision	Recall	Accuracy
LBP	82.05%	24.48%	88.23%
New Gradient Descriptor from Co-occurrence Matrix	-	20%	90%
"Artificial bee colony algorithm", "branch and bound algorithm" and relevance feedback.	Good precision	Poor recall	-
An approach Combining the wavelet transform technique, Tamura texture feature, and Hausdorff distance.	-	-	Average of 17.67% for experiments on brain and CT

Download PDF

Version 1

posted

You are reading this latest preprint version

Performance Evaluation of State-of-the-Art Texture Feature Extraction Techniques on Medical Imagery Tasks

Status:

Version 1

Abstract

Figures

Introduction

Related Works

Materials And Method

3.1 Hardware Specification

3.2 Software Specification

3.2.1 Online Software Specification

3.2.2 Offline Software Requirement

3.3. Similarity Measurement

3.4. Threshold Value

3.5 Conceptual Framework

4.1. Code

4.2. Data

4.2.1. Dataset Validation

4.2.1.1. Dataset Acquisition

License

4.2.1.2. Dataset Uniformity

Results Analysis And Discussion

Analysis

Conclusion

Summary

Declarations

Acknowledgment

References

Tables

Supplementary Files

Status:

Version 1