Classification of Malaria Using Object Detection Models

Abstract

Malaria poses a global health problem every day, as it affects millions of lives all over the world. A traditional diagnosis requires the manual inspection of blood smears from the patient under a microscope to check for the malaria parasite. This is often time consuming and subject to error. Thus, the automated detection and classification of the malaria type and stage of progression can provide a quicker and more accurate diagnosis for patients. In this research, we used two object detection models, YOLOv5 and scaled YOLOv4, to classify the stage of progression and type of malaria parasite. We also used two different datasets for the classification of stage and parasite type while assessing the viability of the dataset for the task. The dataset used is comprised of microscopic images of red blood cells that were either parasitized or uninfected. The infected cells were classified based on two broad categories: the type of malarial parasite causing the infection and the stage of progression of the disease. The dataset was manually annotated using the LabelImg tool. The images were then augmented to enhance model training. Both models YOLOv5 and scaled YOLOv4 proved effective in classifying the type of parasite. Scaled YOLOv4 was in the lead with an accuracy of 83% followed by YOLOv5 with an accuracy of 78.5%. The proposed models may be useful for the medical professionals in the accurate diagnosis of malaria and its stage prediction.

DSCC_Net: Multi-Classification Deep Learning Models for Diagnosing of Skin Cancer Using Dermoscopic Images

“…Our method offers a full view of a particular site, which enables us to identify the disease, as well as interior areas that have been infected with it. Dermoscopy is the most reliable [ 41 ] and time-effective [ 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 ] approach for determining if a lesion is a BCC, MEL, SCC, or MN. A computerized diagnostic approach is required to identify BCC, MEL, SCC, and MN, since the number of confirmed cases of deadly skin cancer is continually growing [ 62 ].…”

“…These images were represented by bkl, mel, and nv. Krishnaraj et al [ 52 ] designed machine learning [ 53 , 54 , 55 , 56 ] classifiers that identified binary classes of cervical cancer, such as adenosquamous carcinoma and SCC. They collected the dataset at the University of California, Irvine (UCI) repository, and the Borderline-SMOTE approach was employed to balance the unbalanced data.…”

Systematic Review on Diagnostic Reference Levels for Computed Tomography Examinations in Radiation Therapy Planning

“…AI algorithms are being developed and trained to identify patterns and features in medical images, such as X-rays, CT scans, and MRI scans, indicative of various diseases, enabling earlier and more accurate diagnoses. There are many studies related to artificial intelligence in medicine including the articles “AI-Assisted Tuberculosis Detection”, “Classification of Malaria Using Object Detection Models”, “Deep Learning Classifier for Detection of Acute Lymphoblastic Leukemia Using Blood Smear Images”, “Uses of AI in Ultrasound Imaging”, and “Machine Learning in Prostate MRI for Prostate Cancer” [ 31 , 32 , 33 , 34 , 35 ]. The integration of AI into medical imaging is an exciting development that has brought about many changes.…”

An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

“…Among the studies that employed deep learning object detectors on thin blood smear images, Yang et al [ 30 ] reported that the YOLOv2 (You Only Look Once v. 2) model achieved an accuracy of 79.22% in detecting the P. vivax -infected cell. In a study conducted by Krishnadas et al [ 31 ], scaled YOLOv4 and YOLOv5 models were used for malaria parasite classification, with the models achieving an accuracy of 83% and 78.5%, respectively. In their study, Sukumarran et al [ 32 ] compared the performances of various object detectors (YOLOv4, Faster R-CNN and SSD-300) for the detection of infected cells and found that the YOLOv4 model outperformed the other object detectors, achieving an accuracy of 93.87%.…”

“…In the study of Koirala et al [ 31 ], thin blood smear images obtained from the Malaria Parasite Image Database (MP-IDB) were utilised for a multi-classification problem, namely to classify the individual malaria parasite species using the scaled YOLOv4 and YOLOv5 models. This is the only study that we are aware of which classifies malaria parasite species using an object detector [ 31 ], but while that authors used the models to classify species on thin blood smear images, it is noteworthy that the study did not include cross-dataset validation or additional model tuning or modification to enhance classification performance. Furthermore, it is still unclear if using deep learning object detectors is ideal for performing multi-class classification of the cell according to the malaria parasite species, as there is no comparison with CNN models in terms of similar classification problems.…”