DOI:10.32604/iasc.2022.024361
| Intelligent Automation & Soft Computing DOI:10.32604/iasc.2022.024361 | |
| Article |
An Intelligent Classification System for Trophozoite Stages in Malaria Species
1Faculty of Electrical Engineering Technology, Universiti Malaysia Perlis (UniMAP), Pauh Putra Campus, Arau, 02600, Malaysia
2Department of Microbiology and Parasitology, School of Medical Sciences, Universiti Sains Malaysia, Kubang Kerian, 16150, Malaysia
3Department of Electrical Engineering, Faculty of Engineering, Universitas Muhammadiyah Yogyakarta, Yogyakarta, Indonesia
*Corresponding Author: Siti Nurul Aqmariah Mohd Kanafiah. Email: aqmariah@unimap.edu.my
Received: 14 October 2021; Accepted: 29 November 2021
Abstract: Malaria is categorised as a dangerous disease that can cause fatal in many countries. Therefore, early detection of malaria is essential to get rapid treatment. The malaria detection process is usually carried out with a 100x magnification of thin blood smear using microscope observation. However, the microbiologist required a long time to identify malaria types before applying any proper treatment to the patient. It also has difficulty to differentiate the species in trophozoite stages because of similar characteristics between species. To overcome these problems, a computer-aided diagnosis system is proposed to classify trophozoite stages of Plasmodium Knowlesi (PK), Plasmodium Falciparum (PF) and Plasmodium Vivax (PV) as early species identification. The process begins with image acquisition, image processing and classification. The image processing involved contrast enhancement using histogram equalisation (HE), segmentation procedure using a combination of hue, saturation and value (HSV) color model, Otsu method and range of each red, green and blue (RGB) color selections, and feature extraction. The features consist of the size of infected red blood cell (RBC), brown pigment in the parasite, and texture using Gray Level Co-occurrence Matrix (GLCM) parts. Finally, the classification method using Multilayer Perceptron (MLP) trained by Bayesian Rules (BR) show the highest accuracy of 98.95%, rather than Levenberg Marquardt (LM) and Conjugate Gradient Backpropagation (CGP) training algorithms.
Keywords: Malaria parasite; thin blood smears; image processing; classification
Malaria is an acute (sudden) and chronic disease that can threaten people’s health worldwide. It has been reported that around 229 million cases and 409,000 deaths worldwide were recorded in 2019 due to malaria, which led to an expenditure of US$ 3.0 billion to control and prevent malaria at the global scale, particularly within the African Region, South-East Asia, Eastern Mediterranean, America and Western Pacific [1]. Malaria is caused by Plasmodium parasites that spread to people through the bites of female Anopheles mosquitoes. Malaria parasites can be categorised into five species which are Plasmodium Knowlesi (PK), Plasmodium Falciparum (PF), Plasmodium Vivax (PV), Plasmodium Malariae (PM) and Plasmodium Ovale (PO) [2,3].
PF, PK and PV are commonly discovered in Malaysia [4,5]. Around 4,630 cases are occurring in 2018, where 89% are from PK. The early trophozoite of PK is identical to those of PF and PV due to similar characteristics such as double chromatin dots, multiple infected RBC and applique forms [6,7]. Therefore, early-stage identification of malaria parasites is an essential factor in preventing and controlling the spread of malaria.
Clinically, visual inspection becomes quite challenging if a microscopic image contains blur effects, unwanted artefacts and a lot of noise in the image [8]. In addition, human perception is not only inconclusive but also not error-free [9]. Furthermore, microscopic examination of blood film for malarial parasites also consumes a lot of time and energy of the microbiologist in deciding the positivity of the sample [10]. Thus, computer-aided systems based on image processing have been developed to assist the visual inspection problems [11–13].
In the studies of image processing and automated classification procedure using a computer-aided system, many limitations have been reported by previous researchers. Most of the classification process only focused on the infected RBCs and normal RBCs classification systems [14,15], classification of PF, PM, PV and PO species with their life cycle stages [16], classification of PF, PM, PV and PO species without identifying the stages in malaria parasite [17], classification of the stage at PF [18–20], PV [21], PM and PO [22], and PF and PV species [23–25], respectively. No previous work has been reported on the automated classification of PK, PF and PV since PK is considered a new and regional species are infecting humans.
Thus, this study proposes an intelligent classification system for PK, PF, and PV Malaria species for trophozoite stages. The system is developed using image processing techniques, including contrast enhancement, proposed segmentation procedure, feature extraction and classification process using MLP-BR to classify the image.
This paper is arranged as follows: In Section 2, the methodology to develop an automated classification system is introduced. Then, Section 3 describes the application of the method and experimental results are presented and analysed. Finally, Section 4 provides the conclusion of this work.
The system architecture is implemented using five main processes: image acquisition, contrast enhancement, image segmentation, feature extraction, classification, and Graphical User Interface (GUI), as shown in Fig. 1. The proposed approach is introduced and validated using Matlab R2018a with Intel® Core™ i7-7500U CPU @ 2.90 GHz and 8 GB RAM.
Figure 1: The system architecture of the intelligent classification system for malaria parasite
Malaria samples were prepared by experts from Hospital Universiti Sains Malaysia, Kelantan, Malaysia. Images were acquired using microscope BX41 connected to Olympus XC50 camera under 100X magnification resolution using immersion oil. The capturing procedure of the sample images has been approved by National Medical Research Register (NMRR), with the register serial number of NMRR-16-1434-31673 (IIR). This study only focuses on differentiating between trophozoite stages of PK, PF, and PV. Moreover, 150 thin blood smear images, including 50 images of each PK, PF and PV, have been captured. Images were saved in the bitmap (BMP) format in 800 × 600 color (RGB) ideas.
The contrast enhancement method aims to enhance the image’s contrast and reduce the color consistency problem of thin blood smear images due to the staining process [26,27]. In this study, the histogram equalisation (HE) method [28] is applied to correct the color intensity of each image by referring to the intensity value of the reference image. Therefore, the microbiologist has validated the reference image as the most suitable image. Fig. 2 shows the reference image using the intensity value for the enhancement process.
Figure 2: Reference image
Using the HE method, the input color image is corrected according to the intensity value of the reference image. The reference image used in the algorithm is fixed and does not change for any color correction process of other photos. Therefore, the reference image contains the best intensity among all samples. Fig. 3 shows the image and histogram, adjusted to the intensity value range as a reference image. Based on Fig. 3, the output image in (i) and histogram in (j), (k) and (l) are similar to the intensity value of the reference image in (e) and histogram in (f), (g) and (h), respectively.
Figure 3: The application of the HE method using the intensity value of the reference image. (a) Input image (b) Input R layer (c) Input G layer (d) Input B layer (e) Reference image (f) Reference R layer (g) Reference G layer (h) Reference B layer (i) Output image (j) Output R layer (k) Output G layer (l) Output B layer
Image segmentation aims to separate the region of interest from the unwanted regions [29]. The proposed segmented procedure has been successfully developed. Eight steps have been implemented in segmentation procedure steps that were successfully done to obtain the region of interest (ROI) by considering the following:
1. Load the original color image from Subsection (2.2).
2. Convert the image in Step 1 to a grayscale image.
3. Segment the grayscale image from Step 2 into a binary image using Otsu’s method [30]. The foreground is represented as white (assigned to 1), and the background is represented as black (assigned to 0).
4. Apply erosion operation by using structuring element ‘disk’ to separate the overlapping cells.
5. Segment image in Step 2 using manual thresholding method according to the following equation:
6. Identify which coordinate in Step 5 overlapped with the cell in Step 4. The rest will be deleted.
7. Infected cells together with an object will be displayed.
8. Apply dilation operation in Step 7 by using the similar structuring element ‘disk’ in Step 4 to obtain the original size of an infected cell.
Steps 1 to 3 are the segmentation process of the cell images, whereas Steps 4 to 8 are the process for obtaining infected cells.
During the segmentation process, the overlapping image occurred after applying Otsu’s method. Therefore, the erosion process has been used to separate the overlapping cells. In Step 5, the manual thresholding method set 50 to 110 intensity value has been used to find the coordinate location of malaria parasite in the image. This intensity value has been selected based on observation from a microbiologist for 150 images during the experiment.
Fig. 4 shows the resultant images after the proposed segmentation steps starting from grayscale image ((a) to (c)), Otsu’s method ((d) to (f)), erosion process ((g) to (i)), finding of coordinate location for malaria parasite ((j) to (l)), and display of infected cell ((m) to (o)).
Figure 4: The segmentation procedure for trophozoite stages in PF, PV and PK malaria species. (a) Grayscale PF image (b) Grayscale PV image (c) Grayscale PK image (d) Result of Otsu’s method for PF (e) Result of Otsu’s method for PV (f) Result of Otsu’s method for PK (g) Result of erosion for PF (h) Result of erosion for PV (i) Result of erosion for PK (j) Result of PF coordinate (k) Result of PV coordinate (l) Result of PK coordinate (m) Result of PF infected cell (n) Result of PV infected cell (o) Result of PK infected cell
Three features have been extracted, namely the size of infected red blood cell (RBC), brown pigment in the parasite and texture feature using the Gray Level Co-occurrence Matrix (GLCM) technique [31].
The size of infected RBC is based on the sum of white pixel values obtained from the segmented image shown in Figs. 5a–5c.
Figure 5: Feature extraction for the size of infected RBC (a) Binary image of PF (b) Binary image of PV (c) Binary image of PK
2.4.2 Brown Pigment in Parasite
The brown pigment in the trophozoite stage of the malaria parasite only exists in PK species. The hue, saturation and value (HSV) color model [25] is applied to the segmented RBC images as in Figs. 6a–6c. Hence, HSV threshold value is set as (0.00, 0.59, 0.51). The brown pigment detection in the HSV color model is shown in Figs. 6d–6f. On the other hand, Fig. 7 shows the detail of brown pigment detected in PK.
Figure 6: The brown pigment detection in HSV color model (a) Color Segmented PF image (b) Color Segmented PV image (c) Color Segmented PK image (d) HSV PF image (e) HSV PV image (f) HSV PK image
Figure 7: Detection of brown pigment in PK
2.4.3 Texture Feature Using GLCM Technique of Infected RBC
The GLCM method [31] was used to extract the texture of grayscale infected RBC for PF, PV, and PK. GLCM characteristics such as contrast, correlation and entropy were used to extract the texture of grayscale infected RBC image. Fig. 8 shows the grayscale images that have been used to extract the texture features of infected RBC.
Figure 8: Feature extraction for texture analysis (a) Grayscale PF image (b) Grayscale PV image (c) Grayscale PK image
All data from the feature extraction process has been analysed using Multilayer Perceptron (MLP) for the classification process. In this study, the Levenberg-Marquardt (LM), Bayesian Regulation (BR) and Conjugate Gradient Backpropagation (CGP) had been used as a learning algorithm to train and test for the accuracy and robustness of the result [32]. A total of 150 images were divided into 90% for training and 10% for testing datasets.
Tab. 1 shows the comparison performances of the three classifiers. BR provided the highest classification accuracy of 99.81% for training and 98.95% for testing. Therefore, the proposed procedure can effectively differentiate the trophozoite stages between PF, PV and PK images and has produced promising results in trophozoite stages of malaria.
2.6 Graphical User Interface for Malaria Classification System
Fig. 9 shows the graphical user interface (GUI) implemented inside the proposed intelligent classification system for trophozoite stages in PK, PF and PV. To perform the malaria classification process, the user must first select the blood sample image by clicking the ‘LOAD IMAGE’ button. The image will display in the ‘Input Image’ layout, as shown in Fig. 9. Then, the user starts the automatic image processing and classification by clicking the ‘PROCESS IMAGE’ button. The processed image will display in the ‘Processed Image’ layout (Fig. 9). Finally, the classification result will show in the status box labelled as ‘Result’.
Figure 9: The GUI of an Intelligent classification for Malaria parasite
An Intelligent Classification System for Trophozoite Stages in PK, PF and PV Malaria Species has been successfully analysed using image processing and classified using MLP to get the performance of the data. In the image processing method, the process of contrast enhancement, proposed segmentation procedure and feature extraction is used to obtain image characteristics. The experimental results have been performed using MLP-BR to obtain the good accuracy of classification. The proposed system helps classify trophozoite stages of PK, PF and PV with an accuracy of 98.95%.
Acknowledgement: The authors would like to thank the technologist in Centre for Research Laboratory (CRL), Hospital University Sains Malaysia (HUSM), Kubang Kerian, Kota Bharu, Kelantan, Malaysia its facilities and providing useful information during this research study.
Funding Statement: This work was supported by the Ministry of Health, Malaysia (MOH) through the Fundamental Research Grant Scheme (FRGS), Phase 1/2016 (grant number 9003-00586) and Mentorship Research grant 2020 (9001-00606).
Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.
1. World Health Organization, Malaria Report, 2020. [Google Scholar]
2. M. De Niz, E. Meibalan, P. Mejia, S. Ma, N. M. B. Brancucci et al., “Plasmodium gametocytes display homing and vascular transmigration in the host bone marrow,” Science Advances, vol. 4, no. 5, pp. 1–15, 2018. [Google Scholar]
3. S. Tirakarn, P. Riangrungroj, P. Kongsaeree, M. Imwong, Y. Yuthavong et al., “Cloning and heterologous expression of Plasmodium ovale dihydrofolate reductase-thymidylate synthase gene,” Parasitology International, vol. 61, no. 2, pp. 324–332, 2012. [Google Scholar]
4. A. Z. Chin, M. C. M. Maluda, J. Jelip, M. S. Jeffree, R. Culleton et al., “Malaria elimination in Malaysia and the rising threat of Plasmodium knowlesi,” Journal of Physiological Anthropology, vol. 39, no. 36, pp. 1–9, 2020. [Google Scholar]
5. S. M. Alasil and A. A. Khairul, “An epidemiological review on emerging and Re-emerging parasitic infectious diseases in Malaysia,” The Open Microbiology Journal, vol. 13, pp. 112–120, 2019. [Google Scholar]
6. D. W. Nyamai and O. T. Bishop, “Aminoacyl TRNA synthetases as malarial drug targets: A comparative bioinformatics study 11 medical and health sciences 1108 medical microbiology 06 biological sciences 0601 biochemistry and cell biology,” Malaria Journal, vol. 18, no. 1, pp. 1–27, 2019. [Google Scholar]
7. B. Singh and C. Daneshvar, “Human infections and detection of Plasmodium knowlesi,” Clinical Microbiology Reviews, vol. 26, no. 2, pp. 165–184, 2013. [Google Scholar]
8. I. N. D. Lubis, “Molecular epidemiology of Plasmodium spp. in North Sumatera, Indonesia, and the efficacy of dihydroartemisinin-piperaquine and artemether-lumefantrine for treatment of clinical malaria,” Ph.D. thesis, School of Hygiene & Tropical Medicine, London, 2018. [Google Scholar]
9. M. Poostchi, K. Silamut, R. J. Maude, S. Jaeger and G. Thoma, “Image analysis and machine learning for detecting malaria,” Journal of Translational Research, vol. 194, pp. 36–55, 2016. [Google Scholar]
10. A. Mbanefo and N. Kumar, “Evaluation of malaria diagnostic methods as a Key for successful control and elimination programs,” Journal of Tropical Medicine and Infectious Disease, vol. 5, no. 2, pp. 1–15, 2020. [Google Scholar]
11. J. Somasekar, G. Ramesh, G. Ramu, P. D. K. Reddy, B. E. Reddy et al., “A dataset for automatic contrast enhancement of microscopic malaria infected blood RGB images,” Data in Brief, vol. 27, pp. 1–7, 2019. [Google Scholar]
12. K. M. Sulur, A. S. Abdul Nasir, W. A. Mustafa, H. Jaafar and Z. Mohamed, “Analysis of colour constancy algorithms for improving segmentation of malaria images,” Journal of Telecommunication, Electronic and Computer Engineering, vol. 10, no. 1–16, pp. 43–49, 2018. [Google Scholar]
13. D. Usha, “Detection of malaria based on the blood smear images using image processing techniques,” International Journal of Engineering Research & Technology (IJERT), vol. 5, no. 21, pp. 1–4, 2017. [Google Scholar]
14. A. Bashir, Z. A. Mustafa, I. AbdelHameid and R. Ibrahim, “Detection and classification of malaria parasites using digital image processing,” in Proc. of the 2017 Int. Conf. on Communication, Control, Computing and Electronics Engineering (ICCCCEE), Khartoum, Sudan, vol. 04, no. 05, pp. 87–89, 2017. [Google Scholar]
15. M. J. Sadiq and V. V. S. S. S. Balaram, “DTBC: Decision tree based binary classification using with feature selection and optimisation for malaria infected erythrocyte detection,” International Journal of Applied Engineering Research, vol. 12, no. 24, pp. 15923–15934, 2017. [Google Scholar]
16. A. Nanoti, S. Jain, C. Gupta and G. Vyas, “Detection of malaria parasite species and life cycle stages using microscopic images of thin blood smear,” in Proc. of the Int. Conf. on Inventive Computation Technologies, (ICICT 2016), Coimbatore, India, 2016. [Google Scholar]
17. L. Malihi, K. Ansari-Asl and A. Behbahani, “Malaria parasite detection in giemsa-stained blood cell images,” in Proc. of the Iranian Conf. on Machine Vision and Image Processing, MVIP-IEEE, Zanjan, Iran, pp. 360–365, 2013. [Google Scholar]
18. W. Kudisthalert, K. Pasupa and S. Tongsima, “Counting and classification of malarial parasite from giemsa-stained thin film images,” IEEE Access, vol. 8, pp. 78663–78682, 2020. [Google Scholar]
19. H. A. Nugroho, S. A. Akbar and E. Murhandarwati, “Feature extraction and classification for detection malaria parasites in thin blood smear,” in Proc. of the 2nd Int. Conf. on Information Technology, Computer, and Electrical Engineering: Green Technology Strengthening in Information Technology, Electrical and Computer Engineering Implementation, Proc., (ICITACEE 2015), Semarang, Indonesia, 2015. [Google Scholar]
20. A. Ravendran, D. S. Roshali and R. Senanayake, “Moment invariant features for automatic identification of critical malaria parasites,” in Proc. of the 2015 IEEE 10th Int. Conf. on Industrial and Information Systems, (ICIIS 2015), Peradeniya, Sri Lanka, 2015. [Google Scholar]
21. D. Das, M. Ghosh, C. Chakraborty, A. Maiti and M. Pal, “Probabilistic prediction of malaria using morphological and textural information,” in Proc. of the Int. Conf. on Image Information Processing, (ICIIP 2011), Shimla, India, vol. 1, no. 6, 2011. [Google Scholar]
22. H. A. Nugroho, A. Darojatun, I. Ardiyanto and R. L. B. Buana, “Classification of Plasmodium malaria and Plasmodium ovale in microscopic thin blood smear digital images,” International Journal on Advanced Science, Engineering and Information Technology, vol. 8, no. 6, pp. 2301–2307, 2018. [Google Scholar]
23. W. Lumchanow and S. Udomsiri, “Image classification of malaria using hybrid algorithms: Convolutional neural network and method to find appropriate K for K-nearest neighbor,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 16, no. 1, pp. 382–388, 2019. [Google Scholar]
24. H. A. Nugroho, A. F. D. Marsiano, K. Xaphakdy, P. Sihakhom, E. L. Frannita et al., “Multithresholding approach for segmenting Plasmodium parasites,” in Proc. of the 2019 11th Int. Conf. on Information Technology and Electrical Engineering, (ICITEE 2019), Pattaya, Thailand, 2019. [Google Scholar]
25. S. S. Savkare and S. P. Narote, “Blood cell segmentation from microscopic blood images,” in Proc. of the IEEE Int. Conf. on Information Processing, (ICIP 2015), Pune, India, pp. 502–505, 2015. [Google Scholar]
26. A. Chourasiya and N. Khare, “A comprehensive review of image enhancement techniques,” International Journal of Innovative Research and Growth, vol. 8, no. 6, pp. 60–71, 2019. [Google Scholar]
27. S. Gupta and Y. Kaur, “Review of different local and global contrast enhancement techniques for a digital image,” International Journal of Computer Applications, vol. 100, no. 18, pp. 18–23, 2014. [Google Scholar]
28. B. Gupta, M. Tiwari and S. S. Lamba, “Visibility improvement and mass segmentation of mammogram images using quantile separated histogram equalisation with local contrast enhancement,” CAAI Transactions on Intelligence Technology, vol. 4, no. 2, pp. 2–8, 2019. [Google Scholar]
29. M. Sridevi, “Image segmentation based on multilevel thresholding using firefly algorithm,” in Proc. of the 2017 Int. Conf. on Inventive Computing and Informatics (ICICI), Coimbatore, India, pp. 750–753, 2017. [Google Scholar]
30. S. S. Savkare and S. P. Narote, “Automated system for malaria parasite identification,” in Proc. of 2015 Int. Conf. on Communication, Information & Computing Technology (ICCICT), IEEE, Mumbai, India, pp. 1–4, 2015. [Google Scholar]
31. A. H. Setianingrum, L. K. Wardhani, A. F. Ridwan and S. F. Nasution, “Identification of Plasmodium falciparum stages using support vector machine method,” in Proc. of the 2019 7th Int. Conf. on Cyber and IT Service Management, (CITSM 2019), Jakarta, Indonesia, pp. 147–151, 2019. [Google Scholar]
32. P. Finne, R. Finne, A. Auvinen, H. Juusela, J. Aro et al., “Predicting the outcome of prostate biopsy in screen-positive Men by a multiplayer perceptron network,” Urology, ELSEVIER, vol. 56, no. 3, pp. 418–422, 2000. [Google Scholar]
 | This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. |