Dataset Information

Affordable artificial intelligence-based digital pathology for neglected tropical diseases: A proof-of-concept for the detection of soil-transmitted helminths and Schistosoma mansoni eggs in Kato-Katz stool thick smears.

ABSTRACT:

Background

With the World Health Organization's (WHO) publication of the 2021-2030 neglected tropical diseases (NTDs) roadmap, the current gap in global diagnostics became painfully apparent. Improving existing diagnostic standards with state-of-the-art technology and artificial intelligence has the potential to close this gap.

Methodology/principal findings

We prototyped an artificial intelligence-based digital pathology (AI-DP) device to explore automated scanning and detection of helminth eggs in stool prepared with the Kato-Katz (KK) technique, the current diagnostic standard for diagnosing soil-transmitted helminths (STHs; Ascaris lumbricoides, Trichuris trichiura and hookworms) and Schistosoma mansoni (SCH) infections. First, we embedded a prototype whole slide imaging scanner into field studies in Cambodia, Ethiopia, Kenya and Tanzania. With the scanner, over 300 KK stool thick smears were scanned, resulting in total of 7,780 field-of-view (FOV) images containing 16,990 annotated helminth eggs (Ascaris: 8,600; Trichuris: 4,083; hookworms: 3,623; SCH: 684). Around 90% of the annotated eggs were used to train a deep learning-based object detection model. From an unseen test set of 752 FOV images containing 1,671 manually verified STH and SCH eggs (the remaining 10% of annotated eggs), our trained object detection model extracted and classified helminth eggs from co-infected FOV images in KK stool thick smears, achieving a weighted average precision (± standard deviation) of 94.9% ± 0.8% and a weighted average recall of 96.1% ± 2.1% across all four helminth egg species.

Conclusions/significance

We present a proof-of-concept for an AI-DP device for automated scanning and detection of helminth eggs in KK stool thick smears. We identified obstacles that need to be addressed before the diagnostic performance can be evaluated against the target product profiles for both STH and SCH. Given that these obstacles are primarily associated with the required hardware and scanning methodology, opposed to the feasibility of AI-based results, we are hopeful that this research can support the 2030 NTDs road map and eventually other poverty-related diseases for which microscopy is the diagnostic standard.

SUBMITTER: Ward P

PROVIDER: S-EPMC9258839 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundTwo Kato-Katz thick smears (Kato-Katzs) from a single stool are currently recommended for diagnosing Schistosoma mansoni infections to map areas for intervention. This 'gold standard' has low sensitivity at low infection intensities. The urine point-of-care circulating cathodic antigen test (POC-CCA) is potentially more sensitive but how accurately they detect S. mansoni after repeated praziquantel treatments, their suitability for measuring drug efficacy and their correlation with egg counts remain to be fully understood. We compared the accuracies of one to six Kato-Katzs and one POC-CCA for the diagnosis of S. mansoni in primary-school children who have received zero to ten praziquantel treatments. We determined the impact each diagnostic approach may have on monitoring and evaluation (M&E) and drug-efficacy findings.Method/principle findingsIn a high S. mansoni endemic area of Uganda, three days of consecutive stool samples were collected from primary school-aged children (six - 12 years) at five time-points in year one: baseline, one-week-post-, four-weeks-post-, six-months-post-, and six-months-one-week-post-praziquantel and three time-points in years two and three: pre-, one-week-post- and four-weeks-post-praziquantel-treatment/retreatment (n = 1065). Two Kato-Katzs were performed on each stool. In parallel, one urine sample was collected and a single POC-CCA evaluated per child at each time-point in year one (n = 367). At baseline, diagnosis by two Kato-Katzs (sensitivity = 98.6%) or one POC-CCA (sensitivity = 91.7%, specificity = 75.0%) accurately predicted S. mansoni infections. However, one year later, a minimum of three Kato-Katzs, and two years later, five Kato-Katzs were required for accurate diagnosis (sensitivity >90%) and drug-efficacy evaluation. The POC-CCA was as sensitive as six Kato-Katzs four-weeks-post and six-months-post-treatment, if trace readings were classified as positive.Conclusions/significanceSix Kato-Katzs (two/stool from three stools) and/or one POC-CCA are required for M&E or drug-efficacy studies. Although unable to measure egg reduction rates, one POC-CCA appears to be more sensitive than six Kato-Katzs at four-weeks-post-praziquantel (drug efficacy) and six-months-post-praziquantel (M&E).

Project description:BackgroundMonitoring the success of soil-transmitted helminth (STH) control programs relies on accurate diagnosis and quantitative assessment of infection prevalence and intensity. As preventative chemotherapeutic program coverage for STH expands, the necessity of gaining insights into the relative or comparative sensitivities, in terms of limits of detection (LOD) and egg-recovery-rates (ERR) for microscopy and quantitative polymerase chain reaction qPCR-based diagnostic techniques becomes imperative to inform suitability for their intended use for large scale STH monitoring and treatment efficacy studies.Methodology/principal findingsThe diagnostic performance in terms of ERR and LOD of the Kato-Katz (KK) thick smear technique, sodium nitrate (NaNO3) faecal floatation (FF) and qPCR for the accurate detection and enumeration of STH eggs were calculated and expressed in eggs per gram (EPG), by experimentally seeding parasite-free human faeces with Ascaris spp., Trichuris spp. and Necator americanus eggs representing low, medium and high intensity infections. The efficiency of NaNO3 flotation was also calculated over a range of specific gravities (SpGr) for the optimum recovery of STH eggs. FF of SpGr 1.30 recovered 62.7%, 11% and 8.7% more Trichuris spp., Necator americanus and Ascaris spp. eggs respectively, than the recommended SpGr of 1.20. All diagnostic methods demonstrated strong direct correlation to the intensity of seeded EPG. KK and FF (SpGr 1.30) resulted in significant lower ERRs compared to qPCR (p <0.05). qPCR demonstrated significantly (p <0.05) greater sensitivity with an ability to detect as little as 5 EPG for all three STH, compared to 50 EPG by KK and FF (SpGr 1.30).Conclusions/significanceThis study compares the diagnostic parameters in terms of LOD and ERRs of STHs for the KK, FF and qPCR. These results indicate that the diagnostic performance of qPCR assays should be considered by control programs in the phase that aims to seek confirmation of transmission break and cessation of preventive chemotherapy in low-transmission settings, in line with the control targets of the WHO neglected tropical diseases 2030 Roadmap.

Project description:BackgroundThe Kato-Katz thick smear (Kato-Katz) is the diagnostic method recommended for monitoring large-scale treatment programs implemented for the control of soil-transmitted helminths (STH) in public health, yet it is difficult to standardize. A promising alternative is the McMaster egg counting method (McMaster), commonly used in veterinary parasitology, but rarely so for the detection of STH in human stool.Methodology/principal findingsThe Kato-Katz and McMaster methods were compared for the detection of STH in 1,543 subjects resident in five countries across Africa, Asia and South America. The consistency of the performance of both methods in different trials, the validity of the fixed multiplication factor employed in the Kato-Katz method and the accuracy of these methods for estimating 'true' drug efficacies were assessed. The Kato-Katz method detected significantly more Ascaris lumbricoides infections (88.1% vs. 75.6%, p<0.001), whereas the difference in sensitivity between the two methods was non-significant for hookworm (78.3% vs. 72.4%) and Trichuris trichiura (82.6% vs. 80.3%). The sensitivity of the methods varied significantly across trials and magnitude of fecal egg counts (FEC). Quantitative comparison revealed a significant correlation (Rs >0.32) in FEC between both methods, and indicated no significant difference in FEC, except for A. lumbricoides, where the Kato-Katz resulted in significantly higher FEC (14,197 eggs per gram of stool (EPG) vs. 5,982 EPG). For the Kato-Katz, the fixed multiplication factor resulted in significantly higher FEC than the multiplication factor adjusted for mass of feces examined for A. lumbricoides (16,538 EPG vs. 15,396 EPG) and T. trichiura (1,490 EPG vs. 1,363 EPG), but not for hookworm. The McMaster provided more accurate efficacy results (absolute difference to 'true' drug efficacy: 1.7% vs. 4.5%).Conclusions/significanceThe McMaster is an alternative method for monitoring large-scale treatment programs. It is a robust (accurate multiplication factor) and accurate (reliable efficacy results) method, which can be easily standardized.

Project description:BackgroundSoil-transmitted helminths infect about one fifth of the world's population and have a negative impact on health. The Kato-Katz technique is the recommended method to detect soil-transmitted helminth eggs in stool samples, particularly in programmatic settings. However, some questions in its procedure remain. Our study aimed to investigate the effect of storage time, storage temperature and stirring of stool samples on fecal egg counts (FECs).Methodology/principal findingsIn the framework of a clinical trial on Pemba Island, United Republic of Tanzania, 488 stool samples were collected from schoolchildren. These samples were evaluated in three experiments. In the first experiment (n = 92), two Kato-Katz slides were prepared from the same stool sample, one was stored at room temperature, the other in a refrigerator for 50 hours, and each slide was analyzed at nine time points (20, 50, 80, 110, 140 minutes, 18, 26, 42 and 50 hours). In the second experiment (n = 340), whole stool samples were split into two, one part was stored at room temperature, and the other part was put in a refrigerator for 48 hours. From each part one Kato-Katz slide was prepared and analyzed at three time points over two days (0, 24 and 48 hours). In the third experiment (n = 56), whole stool samples where stirred for 15 seconds six times and at each time point a Kato-Katz slide was prepared and analyzed. Mean hookworm FECs of Kato-Katz slides stored at room temperature steadily decreased following slide preparation. After two hours, mean hookworm FECs decreased from 22 to 16, whereas no reduction was observed if Kato-Katz slides were stored in the refrigerator (19 vs 21). The time x storage interaction effect was statistically significant (coefficient 0.26, 95% CI: 0.17 to 0.35, p < 0.0001). After 24 hours mean hookworm FECs dropped close to zero, irrespective of the storage condition. Whole stool samples stored at room temperature for one day resulted in a mean hookworm FEC decrease of 23% (p < 0.0001), compared to a 13% reduction (p < 0.0001) if samples were stored in the refrigerator. Fecal egg counts of A. lumbricoides and T. trichiura remained stable over time regardless of storage temperature of whole stool samples. Finally, we found a significant reduction of the variation of hookworm and T. trichiura eggs with increasing rounds of stirring the sample, but not for A. lumbricoides. For hookworm we observed a simultaneous decrease in mean FECs, making it difficult to draw recommendations on stirring samples.Conclusions/significanceOur findings suggest that stool samples (i) should be analyzed on the day of collection and (ii) should be analyzed between 20-30 minutes after slide preparation; if that is not possible, Kato-Katz slides can be stored in a refrigerator for a maximum of 110 minutes.

Project description:BackgroundThe prevalence of Schistosoma mansoni infection is usually assessed by the Kato-Katz diagnostic technique. However, Kato-Katz thick smears have low sensitivity, especially for light infections. Egg count models fitted on individual level data can adjust for the infection intensity-dependent sensitivity and estimate the 'true' prevalence in a population. However, application of these models is complex and there is a need for adjustments that can be done without modeling expertise. This study provides estimates of the 'true' S. mansoni prevalence from population summary measures of observed prevalence and infection intensity using extensive simulations parametrized with data from different settings in sub-Saharan Africa.MethodologyAn individual-level egg count model was applied to Kato-Katz data to determine the S. mansoni infection intensity-dependent sensitivity for various sampling schemes. Observations in populations with varying forces of transmission were simulated, using standard assumptions about the distribution of worms and their mating behavior. Summary measures such as the geometric mean infection, arithmetic mean infection, and the observed prevalence of the simulations were calculated, and parametric statistical models fitted to the summary measures for each sampling scheme. For validation, the simulation-based estimates are compared with an observational dataset not used to inform the simulation.Principal findingsOverall, the sensitivity of Kato-Katz in a population varies according to the mean infection intensity. Using a parametric model, which takes into account different sampling schemes varying from single Kato-Katz to triplicate slides over three days, both geometric and arithmetic mean infection intensities improve estimation of sensitivity. The relation between observed and 'true' prevalence is remarkably linear and triplicate slides per day on three consecutive days ensure close to perfect sensitivity.Conclusions/significanceEstimation of 'true' S. mansoni prevalence is improved when taking into account geometric or arithmetic mean infection intensity in a population. We supply parametric functions and corresponding estimates of their parameters to calculate the 'true' prevalence for sampling schemes up to 3 days with triplicate Kato-Katz thick smears per day that allow estimation of the 'true' prevalence.