Malaria is a life-threatening disease caused by Plasmodium parasites, which are transmitted to humans through the bites of infected mosquitoes. Despite many efforts to control and eliminate the disease, malaria remains a significant public health problem, with an estimated 229 million cases and 409,000 deaths in 2019 (WHO, 2020). Early and accurate diagnosis is crucial for effective treatment and prevention of the disease. In this article, we review recent advances in technology for malaria diagnosis, including microscopy, molecular methods, and machine learning.
Microscopy is the most common method for malaria diagnosis, but it requires trained personnel and quality control measures to ensure accuracy. Conventional light microscopy of Giemsa-stained blood smears is the gold standard for malaria diagnosis, but it has limitations in terms of sensitivity and specificity. One approach to improve microscopy is to use automated image analysis and machine learning algorithms to detect and classify malaria parasites in blood smears (Tek et al., 2009; Das et al., 2015; Jan et al., 2018; Poostchi et al., 2018). These methods can reduce the time and cost of diagnosis, increase accuracy, and enable remote diagnosis in low-resource settings. For example, Torres et al. (2018) developed an automated microscopy system that achieved high sensitivity and specificity for detecting malaria parasites in blood smears from patients in Peru. Chaware et al. (2020) proposed an adaptive illumination system for intelligent microscopy that improves the contrast and resolution of blood smears, enabling better classification of malaria parasites.
Molecular methods such as polymerase chain reaction (PCR) and loop-mediated isothermal amplification (LAMP) are highly sensitive and specific for malaria diagnosis, but they require specialized equipment and expertise. They are useful for detecting low-level parasitemia, monitoring treatment response, and identifying drug-resistant strains of malaria parasites (Mbanefo and Kumar, 2020; Grossenbacher et al., 2020). Picot et al. (2020) conducted a systematic review and meta-analysis of diagnostic accuracy of LAMP methods compared with microscopy, PCR, and rapid diagnostic tests for malaria diagnosis. They found that LAMP has high sensitivity and specificity for detecting Plasmodium falciparum and other species of malaria parasites.
Infrared spectroscopy is a promising technology for malaria diagnosis that uses the unique spectral signature of malaria parasites to distinguish them from normal red blood cells (Heraud et al., 2019). This method is non-invasive, rapid, and requires minimal sample preparation. However, it requires further validation and optimization to improve sensitivity and specificity for malaria diagnosis.
Acridine orange (AO) staining is a fluorescent dye that binds to DNA and RNA in cells, including malaria parasites. AO staining can enhance the sensitivity of microscopy for malaria diagnosis, especially for low parasitemia (Keiser et al., 2002; Kimura et al., 2018). AO staining can also be used in combination with machine learning algorithms for automated detection and classification of malaria parasites in blood smears (Abbas et al., 2019; Mustare, 2020). Other fluorescent dyes such as morin and aluminum can also be used for malaria diagnosis (Malinin and Malinin, 1991; Sodeman and WHO, 1969).
Birefringent hemozoin is a crystalline pigment that is produced by malaria parasites during their life cycle. It can be visualized under polarized light microscopy and used to identify malaria parasites in blood smears (Lawrence and Olson, 1986). However, this method has limited sensitivity and specificity, and it cannot distinguish between different species of malaria parasites.
Machine learning algorithms such as convolutional neural networks (CNNs) and support vector machines (SVMs) have been applied to malaria diagnosis using microscopy images (Vijayalakshmi, 2020; Yu et al., 2020; Molina et al., 2020). These methods can achieve high accuracy and speed for malaria diagnosis, especially when combined with automated image processing and segmentation algorithms (Karthik et al., 2019; Razzak, 2015; Arco et al., 2015; Memeu et al., 2013). However, they require large amounts of annotated training data and may be affected by variations in sample preparation, staining, and imaging conditions.
In conclusion, technology has the potential to improve the accuracy, speed, and accessibility of malaria diagnosis, which is essential for effective control and elimination of the disease. Microscopy, molecular methods, infrared spectroscopy, acridine orange staining, and machine learning algorithms are among the promising approaches for malaria diagnosis that have been developed in recent years. Further research and development are needed to optimize these methods, validate their performance, and integrate them into routine clinical practice in malaria-endemic regions.
- Computer-assisted malaria diagnosis
- Automated malaria diagnosis techniques
- Digital imaging for malaria diagnosis
- Machine learning in malaria diagnosis
- Artificial intelligence in malaria detection
News Source : SpringerLink
Source Link :A Review of Computer-Assisted Techniques Performances in Malaria Diagnosis/