Dataset Information

Kankanet: An artificial neural network-based object detection smartphone application and mobile microscope as a point-of-care diagnostic aid for soil-transmitted helminthiases.

ABSTRACT:

Background

Endemic areas for soil-transmitted helminthiases often lack the tools and trained personnel necessary for point-of-care diagnosis. This study pilots the use of smartphone microscopy and an artificial neural network-based (ANN) object detection application named Kankanet to address those two needs.

Methodology/principal findings

A smartphone was equipped with a USB Video Class (UVC) microscope attachment and Kankanet, which was trained to recognize eggs of Ascaris lumbricoides, Trichuris trichiura, and hookworm using a dataset of 2,078 images. It was evaluated for interpretive accuracy based on 185 new images. Fecal samples were processed using Kato-Katz (KK), spontaneous sedimentation technique in tube (SSTT), and Merthiolate-Iodine-Formaldehyde (MIF) techniques. UVC imaging and ANN interpretation of these slides was compared to parasitologist interpretation of standard microscopy.Relative to a gold standard defined as any positive result from parasitologist reading of KK, SSTT, and MIF preparations through standard microscopy, parasitologists reading UVC imaging of SSTT achieved a comparable sensitivity (82.9%) and specificity (97.1%) in A. lumbricoides to standard KK interpretation (97.0% sensitivity, 96.0% specificity). The UVC could not accurately image T. trichiura or hookworm. Though Kankanet interpretation was not quite as sensitive as parasitologist interpretation, it still achieved high sensitivity for A. lumbricoides and hookworm (69.6% and 71.4%, respectively). Kankanet showed high sensitivity for T. trichiura in microscope images (100.0%), but low in UVC images (50.0%).

Conclusions/significance

The UVC achieved comparable sensitivity to standard microscopy with only A. lumbricoides. With further improvement of image resolution and magnification, UVC shows promise as a point-of-care imaging tool. In addition to smartphone microscopy, ANN-based object detection can be developed as a diagnostic aid. Though trained with a limited dataset, Kankanet accurately interprets both standard microscope and low-quality UVC images. Kankanet may achieve sensitivity comparable to parasitologists with continued expansion of the image database and improvement of machine learning technology.

SUBMITTER: Yang A

PROVIDER: S-EPMC6695198 | biostudies-literature | 2019 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Kankanet: An artificial neural network-based object detection smartphone application and mobile microscope as a point-of-care diagnostic aid for soil-transmitted helminthiases.

Yang Ariel A Bakhtari Nahid N Langdon-Embry Liana L Redwood Emile E Grandjean Lapierre Simon S Rakotomanga Patricia P Rafalimanantsoa Armand A De Dios Santos Juan J Vigan-Womas Inès I Knoblauch Astrid M AM Marcos Luis A LA

PLoS neglected tropical diseases 20190805 8

<h4>Background</h4>Endemic areas for soil-transmitted helminthiases often lack the tools and trained personnel necessary for point-of-care diagnosis. This study pilots the use of smartphone microscopy and an artificial neural network-based (ANN) object detection application named Kankanet to address those two needs.<h4>Methodology/principal findings</h4>A smartphone was equipped with a USB Video Class (UVC) microscope attachment and Kankanet, which was trained to recognize eggs of Ascaris lumbri ...[more]

PMID: 31381573

Similar Datasets

Project description:IntroductionSchistosomiasis (SCH) and soil transmitted helminthiases (STH) have been historically recognized as a major public health problem in Angola. However, lack of reliable, country wide prevalence data on these diseases has been a major hurdle to plan and implement programme actions to target these diseases. This study aimed to characterize SCH and STH prevalence and distribution in Angola.MethodsA country wide mapping was conducted in October 2018 (1 province) and from July to December 2019 (14 provinces) in school aged (SAC) children in 15 (of 18) provinces in Angola, using WHO protocols and procedures. A total of 640 schools and an average of 50 students per school (N = 31,938 children) were sampled. Stool and urine samples were collected and processed using the Kato-Katz method and Urine Filtration. Prevalence estimates for SCH and STH infections were calculated for each province and district with 95% confidence intervals. Factors associated with SCH and STH infection, respectively, were explored using multivariable logistic regression accounting for clustering by school.ResultsOf the 131 districts surveyed, 112 (85.5%) are endemic for STH, 30 (22.9%) have a prevalence above 50%, 24 (18.3%) are at moderate risk (prevalence 20%-50%), and 58 (44.3%) are at low risk (<20% prevalence); similarly, 118 (90,1%) of surveyed districts are endemic for any SCH, 2 (1.5%) are at high risk (>50% prevalence), 59 (45.0%) are at moderate risk (10%-50% prevalence), and 57 (43.5%) are at low risk (<10% prevalence). There were higher STH infection rates in the northern provinces of Malanje and Lunda Norte, and higher SCH infection rates in the southern provinces of Benguela and Huila.ConclusionsThis mapping exercise provides essential information to Ministry of Health in Angola to accurately plan and implement SCH and STH control activities in the upcoming years. Data also provides a useful baseline contribution for Angola to track its progress towards the 2030 NTD roadmap targets set by WHO.

Project description:BackgroundPrograms for control of soil-transmitted helminth (STH) infections are increasingly evaluating national mass drug administration (MDA) interventions. However, "unprogrammed deworming" (receipt of deworming drugs outside of nationally-run STH control programs) occurs frequently. Failure to account for these activities may compromise evaluations of MDA effectiveness.MethodsWe used a cross-sectional study design to evaluate STH infection and unprogrammed deworming among infants (aged 6-11 months), preschool-aged children (PSAC, aged 1-4 years), and school-aged children (SAC, aged 5-14 years) in Kibera, Kenya, an informal settlement not currently receiving nationally-run MDA for STH. STH infection was assessed by triplicate Kato-Katz. We asked heads of households with randomly-selected children about past-year receipt and source(s) of deworming drugs. Local non-governmental organizations (NGOs) and school staff participating in school-based deworming were interviewed to collect information on drug coverage.ResultsOf 679 children (18 infants, 184 PSAC, and 477 SAC) evaluated, 377 (55%) reported receiving at least one unprogrammed deworming treatment during the past year. PSAC primarily received treatments from chemists (48.3%) or healthcare centers (37.7%); SAC most commonly received treatments at school (55.0%). Four NGOs reported past-year deworming activities at 47 of >150 schools attended by children in our study area. Past-year deworming was negatively associated with any-STH infection (34.8% vs 45.4%, p = 0.005). SAC whose most recent deworming medication was sourced from a chemist were more often infected with Trichuris (38.0%) than those who received their most recent treatment from a health center (17.3%) or school (23.1%) (p = 0.05).ConclusionUnprogrammed deworming was received by more than half of children in our study area, from multiple sources. Both individual-level treatment and unprogrammed preventive chemotherapy may serve an important public health function, particularly in the absence of programmed deworming; however, they may also lead to an overestimation of programmed MDA effectiveness. A standardized, validated tool is needed to assess unprogrammed deworming.

Project description:BACKGROUND:The goal of soil-transmitted helminthiases (STH) control programmes is to eliminate STH-associated morbidity in the target population by reducing the prevalence of moderate- and heavy-intensity infections and the overall STH infection prevalence mainly through preventive chemotherapy (PC) with either albendazole or mebendazole. Endemic countries should measure the success of their control programmes through regular epidemiological assessments. We evaluated changes in STH prevalence in countries that conducted effective PC coverage for STH to guide changes in the frequency of PC rounds and the number of tablets needed. METHODS:We selected countries from World Health Organization (WHO)'s Preventive Chemotherapy and Transmission control (PCT) databank that conducted ?5?years of PC with effective coverage for school-age children (SAC) and extracted STH baseline and impact assessment data using the WHO Epidemiological Data Reporting Form, Ministry of Health reports and/or peer-reviewed publications. We used pooled and weighted means to plot the prevalence of infection with any STH and with each STH species at baseline and after ?5?years of PC with effective coverage. Finally, using the WHO STH decision tree, we estimated the reduction in the number of tablets needed. RESULTS:Fifteen countries in four WHO regions conducted annual or semi-annual rounds of PC for STH for 5?years or more and collected data before and after interventions. At baseline, the pooled prevalence was 48.9% (33.1-64.7%) for any STH, 23.2% (13.7-32.7%) for Ascaris lumbricoides, 21.01% (9.7-32.3%) for Trichuris trichiura and 18.2% (10.9-25.5%) for hookworm infections, while after ?5?years of PC for STH, the prevalence was 14.3% (7.3-21.3%) for any STH, 6.9% (1.3-12.5%) for A. lumbricoides, 5.3% (1.06-9.6%) for T. trichiura and 8.1% (4.0-12.2%) for hookworm infections. CONCLUSIONS:Countries endemic for STH have made tremendous progress in reducing STH-associated morbidity, but very few countries have data to demonstrate that progress. In this study, the data show that nine countries should adapt their PC strategies and the frequency of PC rounds to yield a 36% reduction in drug needs. The study also highlights the importance of impact assessment surveys to adapt control strategies according to STH prevalence.

Project description:BackgroundInfections caused by soil-transmitted helminths (STHs) are the most prevalent neglected tropical diseases and result in a major disease burden in low- and middle-income countries, especially in school-aged children. Improved diagnostic methods, especially for light intensity infections, are needed for efficient, control and elimination of STHs as a public health problem, as well as STH management. Image-based artificial intelligence (AI) has shown promise for STH detection in digitized stool samples. However, the diagnostic accuracy of AI-based analysis of entire microscope slides, so called whole-slide images (WSI), has previously not been evaluated on a sample-level in primary healthcare settings in STH endemic countries.Methodology/principal findingsStool samples (n = 1,335) were collected during 2020 from children attending primary schools in Kwale County, Kenya, prepared according to the Kato-Katz method at a local primary healthcare laboratory and digitized with a portable whole-slide microscopy scanner and uploaded via mobile networks to a cloud environment. The digital samples of adequate quality (n = 1,180) were split into a training (n = 388) and test set (n = 792) and a deep-learning system (DLS) developed for detection of STHs. The DLS findings were compared with expert manual microscopy and additional visual assessment of the digital samples in slides with discordant results between the methods. Manual microscopy detected 15 (1.9%) Ascaris lumbricoides, 172 (21.7%) Tricuris trichiura and 140 (17.7%) hookworm (Ancylostoma duodenale or Necator americanus) infections in the test set. Importantly, more than 90% of all STH positive cases represented light intensity infections. With manual microscopy as the reference standard, the sensitivity of the DLS as the index test for detection of A. lumbricoides, T. trichiura and hookworm was 80%, 92% and 76%, respectively. The corresponding specificity was 98%, 90% and 95%. Notably, in 79 samples (10%) classified as negative by manual microscopy for a specific species, STH eggs were detected by the DLS and confirmed correct by visual inspection of the digital samples.Conclusions/significanceAnalysis of digitally scanned stool samples with the DLS provided high diagnostic accuracy for detection of STHs. Importantly, a substantial number of light intensity infections were missed by manual microscopy but detected by the DLS. Thus, analysis of WSIs with image-based AI may provide a future tool for improved detection of STHs in a primary healthcare setting, which in turn could facilitate monitoring and evaluation of control programs.

Dataset Information

Kankanet: An artificial neural network-based object detection smartphone application and mobile microscope as a point-of-care diagnostic aid for soil-transmitted helminthiases.

Background

Methodology/principal findings

Conclusions/significance

Publications

Kankanet: An artificial neural network-based object detection smartphone application and mobile microscope as a point-of-care diagnostic aid for soil-transmitted helminthiases.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets