Computer Vision Malaria Diagnostic Systems-Progress and Prospects

Towards digital diagnosis of malaria: How far have we reached?

The need for precise and early diagnosis of malaria and its distinction from other febrile illnesses is no doubt a prerequisite, primarily when standard rapid diagnostic tests (RDTs) cannot be totally relied upon. At the time of disease outbreaks, the pressure on hospital staff remains high and the chances of human error increase. Therefore, in the era of digitalisation of medicine as well as diagnostic approaches, various technologies such as artificial intelligence (AI) and machine learning (ML) should be deployed to further aid the diagnosis, especially in endemic and epidemic situations. Computational techniques are now more at the forefront than ever, and the interest in developing such efficient technologies is continuously increasing. A comprehensive understanding of these digital technologies is needed to maintain the scientific rigour in these attempts. This would enhance the implementation of these novel technologies for malaria diagnosis. This review highlights the progression, strengths, and limitations of various computing techniques so far employed to diagnose malaria.

Field evaluation of the diagnostic performance of EasyScan GO: a digital malaria microscopy device based on machine-learning

BACKGROUND

Microscopic examination of Giemsa-stained blood films remains the reference standard for malaria parasite detection and quantification, but is undermined by difficulties in ensuring high-quality manual reading and inter-reader reliability. Automated parasite detection and quantification may address this issue.

METHODS

A multi-centre, observational study was conducted during 2018 and 2019 at 11 sites to assess the performance of the EasyScan Go, a microscopy device employing machine-learning-based image analysis. Sensitivity, specificity, accuracy of species detection and parasite density estimation were assessed with expert microscopy as the reference. Intra- and inter-device reliability of the device was also evaluated by comparing results from repeat reads on the same and two different devices. This study has been reported in accordance with the Standards for Reporting Diagnostic accuracy studies (STARD) checklist.

RESULTS

In total, 2250 Giemsa-stained blood films were prepared and read independently by expert microscopists and the EasyScan Go device. The diagnostic sensitivity of EasyScan Go was 91.1% (95% CI 88.9-92.7), and specificity 75.6% (95% CI 73.1-78.0). With good quality slides sensitivity was similar (89.1%, 95%CI 86.2-91.5), but specificity increased to 85.1% (95%CI 82.6-87.4). Sensitivity increased with parasitaemia rising from 57% at < 200 parasite/µL, to ≥ 90% at > 200-200,000 parasite/µL. Species were identified accurately in 93% of Plasmodium falciparum samples (kappa = 0.76, 95% CI 0.69-0.83), and in 92% of Plasmodium vivax samples (kappa = 0.73, 95% CI 0.66-0.80). Parasite density estimates by the EasyScan Go were within ± 25% of the microscopic reference counts in 23% of slides.

CONCLUSIONS

The performance of the EasyScan Go in parasite detection and species identification accuracy fulfil WHO-TDR Research Malaria Microscopy competence level 2 criteria. In terms of parasite quantification and false positive rate, it meets the level 4 WHO-TDR Research Malaria Microscopy criteria. All performance parameters were significantly affected by slide quality. Further software improvement is required to improve sensitivity at low parasitaemia and parasite density estimations. Trial registration ClinicalTrials.gov number NCT03512678.

Performance Evaluation of Biozentech Malaria Scanner in Plasmodium knowlesi and P. falciparum as a New Diagnostic Tool

The computer vision diagnostic approach currently generates several malaria diagnostic tools. It enhances the accessible and straightforward diagnostics that necessary for clinics and health centers in malaria-endemic areas. A new computer malaria diagnostics tool called the malaria scanner was used to investigate living malaria parasites with easy sample preparation, fast and user-friendly. The cultured Plasmodium parasites were used to confirm the sensitivity of this technique then compared to fluorescence-activated cell sorting (FACS) analysis and light microscopic examination. The measured percentage of parasitemia by the malaria scanner revealed higher precision than microscopy and was similar to FACS. The coefficients of variation of this technique were 1.2–6.7% for Plasmodium knowlesi and 0.3–4.8% for P. falciparum. It allowed determining parasitemia levels of 0.1% or higher, with coefficient of variation smaller than 10%. In terms of the precision range of parasitemia, both high and low ranges showed similar precision results. Pearson’s correlation test was used to evaluate the correlation data coming from all methods. A strong correlation of measured parasitemia (r²=0.99, P<0.05) was observed between each method. The parasitemia analysis using this new diagnostic tool needs technical improvement, particularly in the differentiation of malaria species.

Detection and stage classification of Plasmodium falciparum from images of Giemsa stained thin blood films using random forest classifiers

BACKGROUND

The conventional method for the diagnosis of malaria parasites is the microscopic examination of stained blood films, which is time consuming and requires expertise. We introduce computer-based image segmentation and life stage classification with a random forest classifier. Segmentation and stage classification are performed on a large dataset of malaria parasites with ground truth labels provided by experts.

METHODS

We made use of Giemsa stained images obtained from the blood of 16 patients infected with Plasmodium falciparum. Experts labeled the parasite types from each of the images. We applied a two-step approach: image segmentation followed by life stage classification. In segmentation, we classified each pixel as a parasite or non-parasite pixel using a random forest classifier. Performance was evaluated with classification accuracy, Dice coefficient and free-response receiver operating characteristic (FROC) analysis. In life stage classification, we classified each of the segmented objects into one of 8 classes: 6 parasite life stages, early ring, late ring or early trophozoite, mid trophozoite, early schizont, late schizont or segmented, and two other classes, white blood cell or debris.

RESULTS

Our segmentation method gives an average cross-validated Dice coefficient of 0.82 which is a 13% improvement compared to the Otsu method. The Otsu method achieved a True Positive Fraction (TPF) of 0.925 at the expense of a False Positive Rate (FPR) of 2.45. At the same TPF of 0.925, our method achieved an FPR of 0.92, an improvement by more than a factor two. We find that inclusion of average intensity of the whole image as feature for the random forest considerably improves segmentation performance. We obtain an overall accuracy of 58.8% when classifying all life stages. Stages are mostly confused with their neighboring stages. When we reduce the life stages to ring, trophozoite and schizont only, we obtain an accuracy of 82.7%.

CONCLUSION

Pixel classification gives better segmentation performance than the conventional Otsu method. Effects of staining and background variations can be reduced with the inclusion of average intensity features. The proposed method and data set can be used in the development of automatic tools for the detection and stage classification of malaria parasites. The data set is publicly available as a benchmark for future studies.

Cost-effectiveness analysis of malaria rapid diagnostic test in the elimination setting

BACKGROUND

As more and more countries approaching the goal of malaria elimination, malaria rapid diagnostic tests (RDT) was recomendated to be a diagnostic strategy to achieve and maintain the statute of malaria free, as it's less requirments on equipment and experitise than microscopic examination. But there are very few economic evaluations to confirm whether RDT was cost-effective in the setting of malaria elimination. This research aimed to offer evidence for helping decision making on malaria diagnosis strategy.

METHODS

A cost-effectiveness analysis was conducted to compare RDT with microscopy examination for malaria diagnosis, by using a decision tree model. There were three strategies of malaria diagnostic testing evaluated in the model, 1) microscopy, 2) RDT, 3) RDT followed by microscopy. The effect indicator was defined as the number of malaria cases treated appropriately. Based on the joint perspective of health sector and patient, costs data were collected from hospital information systems, key informant interviews, and patient surveys. Data collection was conducted in Jiangsu from September 2018 to January 2019. Epidemiological data were obtained from local malaria surveillance reports. A hypothetical cohort of 300 000 febrile patients were simulated to calculate the total cost and effect of each strategy. One-way, two-way, and probabilistic sensitivity analysis were performed to test the robustness of the result.

RESULTS

The results showed that RDT strategy was the most effective (245 cases) but also the most costly (United States Dollar [USD] 4.47 million) compared to using microscopy alone (238 cases, USD 3.63 million), and RDT followed by microscopy (221 cases, USD 2.75 million). There was no strategy dominated. One-way sensitivity analysis reflected that the result was sensitive to the change in labor cost and two-way sensitivity analysis indicated that the result was not sensitive to the proportion of falciparum malaria. The result of Monte Carlo simulation showed that RDT strategy had higher effects and higher cost than other strategies with a high probability.

CONCLUSIONS

Compared to microscopy and RDT followed by microscopy, RDT strategy had higher effects and higher cost in the setting of malaria elimination.

Trichoscopy of Alopecia Areata: Hair Loss Feature Extraction and Computation Using Grid Line Selection and Eigenvalue

Recently, the hair loss population, alopecia areata patients, is increasing due to various unconfirmed reasons such as environmental pollution and irregular eating habits. In this paper, we introduce an algorithm for preventing hair loss and scalp self-diagnosis by extracting HLF (hair loss feature) based on the scalp image using a microscope that can be mounted on a smart device. We extract the HLF by combining a scalp image taken from the microscope using grid line selection and eigenvalue. First, we preprocess the photographed scalp images using image processing to adjust the contrast of microscopy input and minimize the light reflection. Second, HLF is extracted through each distinct algorithm to determine the progress degree of hair loss based on the preprocessed scalp image. We define HLF as the number of hair, hair follicles, and thickness of hair that integrate broken hairs, short vellus hairs, and tapering hairs.

Lab-on-a-Chip Technologies for the Single Cell Level: Separation, Analysis, and Diagnostics

In the last three decades, microfluidics and its applications have been on an exponential rise, including approaches to isolate rare cells and diagnose diseases on the single-cell level. The techniques mentioned herein have already had significant impacts in our lives, from in-the-field diagnosis of disease and parasitic infections, through home fertility tests, to uncovering the interactions between SARS-CoV-2 and their host cells. This review gives an overview of the field in general and the most notable developments of the last five years, in three parts: 1. What can we detect? 2. Which detection technologies are used in which setting? 3. How do these techniques work? Finally, this review discusses potentials, shortfalls, and an outlook on future developments, especially in respect to the funding landscape and the field-application of these chips.

Diagnosis of Malaria Parasites Plasmodium spp. in Endemic Areas: Current Strategies for an Ancient Disease

Fast and effective detection of the causative agent of malaria in humans, protozoan Plasmodium parasites, is of crucial importance for increasing the effectiveness of treatment and to control a devastating disease that affects millions of people living in endemic areas. The microscopic examination of Giemsa-stained blood films still remains the gold-standard in Plasmodium detection today. However, there is a high demand for alternative diagnostic methods that are simple, fast, highly sensitive, ideally do not rely on blood-drawing and can potentially be conducted by the patients themselves. Here, the history of Plasmodium detection is discussed, and advantages and disadvantages of diagnostic methods that are currently being applied are assessed.

Asymptomatic malaria in the clinical and public health context

Introduction: Historically, the global community has focused on the control of symptomatic malaria. However, interest in asymptomatic malaria has been growing, particularly in the context of malaria elimination.Areas covered: We undertook a comprehensive PubMed literature review on asymptomatic malaria as it relates to detection and elimination with emphasis between 2014 and 2019. Diagnostic tools with a low limit of detection (LOD) have allowed us to develop a more detailed understanding of asymptomatic malaria and its impact. These highly sensitive diagnostics have demonstrated that the prevalence of asymptomatic malaria is greater than previously thought. In addition, it is now possible to detect the malaria reservoir in the community, something that was previously not feasible. Asymptomatic malaria has previously not been treated, but research has begun to examine whether treating individuals with asymptomatic malaria may lead to health benefits. Finally, we have begun to understand the importance of asymptomatic malaria in ongoing transmission.Expert opinion: Therefore, with malaria elimination back on the agenda, asymptomatic malaria can no longer be ignored, especially in light of new ultra-sensitive diagnostic tools.