The global burden of schistosomiasis and soil-transmitted helminth (STH) infections is substantial, with millions at risk, particularly in sub-Saharan Africa. Traditional diagnostic methods using optical microscopy are labour intensive, operator dependent, and often inaccessible in remote areas. This thesis focuses on the development and validation of an AI-based digital microscope, the Schistoscope, designed to enhance the diagnosis of these infections, particularly in resource-limited settings. The aim is to bridge existing diagnostic gaps by developing an automated system that reduces reliance on skilled personnel, enhances diagnostic accuracy, and improves accessibility.

The research integrates innovations in hardware design, digital imaging, and artificial intelligence (AI) to create an affordable, efficient, and accurate diagnostic tool that addresses the limitations of traditional microscopy. The primary objectives include designing and developing cost-effective digital microscope prototypes, integrating AI algorithms for automated detection and classification of parasite eggs, and validating the diagnostic performance of the system through field studies. The development process began with two low-cost digital microscope prototypes: the Raspberry Pi-based Schistoscope and the smartphone-based Schistoscope. Both designs focused on affordability, portability, and ease of use, with key innovations such as the integration of 3D-printed components and locally sourced materials to ensure sustainability and ease of maintenance in endemic regions.

Subsequent iterations led to the development of the Schistoscope 5.0, which featured significant improvements in imaging quality, automation, and user interface. The device incorporates a whole slide imaging system with an advanced autofocusing algorithm, enhancing image clarity and diagnostic accuracy. A cornerstone of the thesis is the integration of AI for automated detection and quantification of Schistosoma haematobium and intestinal helminth eggs. The diagnostic framework employs deep learning models, particularly convolutional neural networks (CNNs), to perform semantic segmentation and object detection. Key components include the DeepLabV3 with a MobileNetV3 backbone, used for semantic segmentation of Schistosoma haematobium eggs, effectively distinguishing eggs from background artifacts, and the EfficientDet model, applied for the detection of intestinal helminth eggs, including Ascaris lumbricoides, Trichuris trichiura, hookworm, and Schistosoma mansoni.

The models were trained on robust datasets collected from field samples, with extensive image annotation to ensure accuracy. The diagnostic system demonstrated high sensitivity and specificity, meeting the World Health Organisation's target product profiles for schistosomiasis and STH control programs. Comprehensive field validation studies were conducted in Nigeria and Gabon to assess the real-world performance of the Schistoscope. These studies compared the device's diagnostic accuracy with conventional microscopy and composite reference standards incorporating real-time PCR and UCP-LF CAA. In Nigeria, the study also focused on the usability and acceptability of the Schistoscope among healthcare workers, demonstrating high acceptance rate of both the semi- and fully automated modes. In Gabon, the diagnostic performance was assessed on fresh and banked urine samples, with the Schistoscope demonstrating comparable accuracy to traditional microscopy, alongside the added benefits of automation and reduced diagnostic time.

Key findings highlight improved diagnostic accuracy, with the Schistoscope achieving high precision and sensitivity in detecting schistosomiasis and STH infections. The integration of AI reduced the need for skilled personnel, thereby enhancing automation and efficiency. The device's cost-effectiveness is underscored by the use of affordable materials and open-source hardware/software, making it accessible for low-resource settings. Field readiness was confirmed through validation under real-world conditions, attesting to the device's robustness and reliability.

Despite these advancements, the research identified several challenges and limitations. Some images were affected by sub-optimal autofocusing, impacting diagnostic accuracy in certain cases. The AI models require further training on more diverse datasets to improve generalisation across different environmental conditions. Additionally, operational challenges such as mechanical issues in the autofocusing mechanism were noted, necessitating further hardware refinements.

Building on these successes and lessons learned, future research will focus on hardware and software refinement to improve mechanical stability and AI model robustness. The Schistoscope will be adapted for the diagnosis of other parasitic diseases, with enhanced field validation through extensive trials in diverse settings. Sustainable deployment strategies will explore local manufacturing options to reduce costs and improve accessibility. Furthermore, the development of cloud-based data storage and analysis systems will support large-scale public health initiatives.

In conclusion, this thesis demonstrates that the integration of AI with digital microscopy can revolutionise the diagnosis of schistosomiasis and STH infections. The Schistoscope not only matches traditional microscopy in diagnostic accuracy but also offers significant advantages in terms of automation, cost-effectiveness, and field applicability. By addressing key challenges in parasitic disease diagnostics, this research contributes to global efforts in controlling and eliminating neglected tropical diseases (NTD), with the potential to improve health outcomes in some of the world's most vulnerable populations. Therefore, accelerating the achievement of the elimination targets as enshrined in the WHO’s NTD Road Map, 2021-2030 in resource-limited settings. ... Read more

In this work, we developed an automated system for the detection and classification of soil-transmitted helminths (STH) and Schistosoma (S.) mansoni eggs in microscopic images of fecal smears. We assembled an STH and S. mansoni dataset comprising over 3,000 field-of-view (FOV) images containing parasite eggs, extracted from more than 300 fecal smear prepared using the Kato-Katz technique. These images were acquired using Schistoscope—a cost-effective automated digital microscope. After annotating the STH and S. mansoni eggs, we employed a transfer learning approach to train an EfficientDet deep learning model, using 70% of the dataset for training, 20% for validation, and 10% for testing. The developed model successfully identified STH and S. mansoni eggs in the FOV images, achieving weighted average scores of Precision, Sensitivity, Specificity, and F-Score across four classes of helminths (A. lumbricoides, T. trichiura, hookworm, and S. mansoni). Our system highlights the potential of the Schistoscope, enhanced with artificial intelligence, for detecting STH and S. mansoni infections in remote, resource-limited settings and for supporting the monitoring and evaluation of neglected tropical disease (NTD) control programs. ... Read more

Introduction
Schistosomiasis is a significant public health concern, especially in Sub-Saharan Africa. Conventional microscopy is the standard diagnostic method in resource-limited settings, but with limitations, such as the need for expert microscopists. An automated digital microscope with artificial intelligence (Schistoscope), offers a potential solution. This field study aimed to validate the diagnostic performance of the Schistoscope for detecting and quantifying Schistosoma haematobium eggs in urine compared to conventional microscopy and to a composite reference standard (CRS) consisting of real-time PCR and the up-converting particle (UCP) lateral flow (LF) test for the detection of schistosome circulating anodic antigen (CAA).

Methods
Based on a non-inferiority concept, the Schistoscope was evaluated in two parts: study A, consisting of 339 freshly collected urine samples and study B, consisting of 798 fresh urine samples that were also banked as slides for analysis with the Schistoscope. In both studies, the Schistoscope, conventional microscopy, real-time PCR and UCP-LF CAA were performed and samples with all the diagnostic test results were included in the analysis. All diagnostic procedures were performed in a laboratory located in a rural area of Gabon, endemic for S. haematobium.

Results
In study A and B, the Schistoscope demonstrated a sensitivity of 83.1% and 96.3% compared to conventional microscopy, and 62.9% and 78.0% compared to the CRS. The sensitivity of conventional microscopy in study A and B compared to the CRS was 61.9% and 75.2%, respectively, comparable to the Schistoscope. The specificity of the Schistoscope in study A (78.8%) was significantly lower than that of conventional microscopy (96.4%) based on the CRS but comparable in study B (90.9% and 98.0%, respectively).

Conclusion
Overall, the performance of the Schistoscope was non-inferior to conventional microscopy with a comparable sensitivity, although the specificity varied. The Schistoscope shows promising diagnostic accuracy, particularly for samples with moderate to higher infection intensities as well as for banked sample slides, highlighting the potential for retrospective analysis in resource-limited settings. ... Read more

Traditionally, automated slide scanning involves capturing a rectangular grid of field-of-view (FoV) images which can be stitched together to create whole slide images, while the autofocusing algorithm captures a focal stack of images
to determine the best in-focus image. However, these methods can be timeconsuming due to the need for X-, Y- and Z-axis movements of the digital microscope while capturing multiple FoV images. In this paper, we propose a solution to minimise these redundancies by presenting an optimal procedure for automated slide scanning of circular membrane filters on a glass slide. We achieve this by following an optimal path in the sample plane, ensuring that only FoVs overlapping the filter membrane are captured. To capture the best infocus FoV image, we utilise a hill-climbing approach that tracks the peak of the mean of Gaussian gradient of the captured FoVs images along the Z-axis. We implemented this procedure to optimise the efficiency of the Schistoscope, an automated digital microscope developed to diagnose urogenital schistosomiasis by imaging Schistosoma haematobium eggs on 13 or 25 mm membrane filters. Our improved method reduces the automated slide scanning time by 63.18%and 72.52% for the respective filter sizes. This advancement greatly supportsthe practicality of the Schistoscope in large-scale schistosomiasis monitoringand evaluation programs in endemic regions. This will save time, resources andalso accelerate generation of data that is critical in achieving the targets for schistosomiasis elimination. ... Read more

Purpose: Automated diagnosis of urogenital schistosomiasis using digital microscopy images of urine slides is an essential step toward the elimination of schistosomiasis as a disease of public health concern in Sub-Saharan African countries. We create a robust image dataset of urine samples obtained from field settings and develop a two-stage diagnosis framework for urogenital schistosomiasis.

Approach: Urine samples obtained from field settings were captured using the Schistoscope device, and S. haematobium eggs present in the images were manually annotated by experts to create the SH dataset. Next, we develop a two-stage diagnosis framework, which consists of semantic segmentation of S. haematobium eggs using the DeepLabv3-MobileNetV3 deep convolutional neural network and a refined segmentation step using ellipse fitting approach to approximate the eggs with an automatically determined number of ellipses. We defined two linear inequality constraints as a function of the overlap coefficient and area of a fitted ellipses. False positive diagnosis resulting from over-segmentation was further minimized using these constraints. We evaluated the performance of our framework on 7605 images from 65 independent urine samples collected from field settings in Nigeria, by deploying our algorithm on an Edge AI system consisting of Raspberry Pi + Coral USB accelerator.

Result: The SH dataset contains 12,051 images from 103 independent urine samples and the developed urogenital schistosomiasis diagnosis framework achieved clinical sensitivity, specificity, and precision of 93.8%, 93.9%, and 93.8%, respectively, using results from an experienced microscopist as reference.

Conclusion: Our detection framework is a promising tool for the diagnosis of urogenital schistosomiasis as our results meet the World Health Organization target product profile requirements for monitoring and evaluation of schistosomiasis control programs.
... Read more

Conventional microscopy is the standard procedure for the diagnosis of schistosomiasis, despite its limited sensitivity, reliance on skilled personnel, and the fact that it is error prone. Here, we report the performance of the innovative (semi-)automated Schistoscope 5.0 for optical digital detection and quantification of Schistosoma haematobium eggs in urine, using conventional microscopy as the reference standard. At baseline, 487 participants in a rural setting in Nigeria were assessed, of which 166 (34.1%) tested S. haematobium positive by conventional microscopy. Captured images from the Schistoscope 5.0 were analyzed manually (semiautomation) and by an artificial intelligence (AI) algorithm (full automation). Semi- and fully automated digital microscopy showed comparable sensitivities of 80.1% (95% confidence interval [CI]: 73.2-86.0) and 87.3% (95%CI: 81.3-92.0), but a significant difference in specificity of 95.3% (95% CI: 92.4-97.4) and 48.9% (95% CI: 43.3-55.0), respectively. Overall, estimated egg counts of semi- and fully automated digital microscopy correlated significantly with the egg counts of conventional microscopy (r50.90 and r50.80, respectively, P < 0.001), although the fully automated procedure generally underestimated the higher egg counts. In 38 egg positive cases, an additional urine sample was examined 10 days after praziquantel treatment, showing a similar cure rate and egg reduction rate when comparing conventional microscopy with semiautomated digital microscopy. In this first extensive field evaluation, we found the semiautomated Schistoscope 5.0 to be a promising tool for the detection and monitoring of S. haematobium infection, although further improvement of the AI algorithm for full automation is required. ... Read more

For many parasitic diseases, the microscopic examination of clinical samples such as urine and stool still serves as the diagnostic reference standard, primarily because microscopes are accessible and cost-effective. However, conventional microscopy is laborious, requires highly skilled personnel, and is highly subjective. Requirements for skilled operators, coupled with the cost and maintenance needs of the microscopes, which is hardly done in endemic countries, presents grossly limited access to the diagnosis of parasitic diseases in resource-limited settings. The urgent requirement for the management of tropical diseases such as schistosomiasis, which is now focused on elimination, has underscored the critical need for the creation of access to easy-to-use diagnosis for case detection, community mapping, and surveillance. In this paper, we present a low-cost automated digital microscope—the Schistoscope—which is capable of automatic focusing and scanning regions of interest in prepared microscope slides, and automatic detection of Schistosoma haematobium eggs in captured images. The device was developed using widely accessible distributed manufacturing methods and off-the-shelf components to enable local manufacturability and ease of maintenance. For proof of principle, we created a Schistosoma haematobium egg dataset of over 5000 images captured from spiked and clinical urine samples from field settings and demonstrated the automatic detection of Schistosoma haematobium eggs using a trained deep neural network model. The experiments and results presented in this paper collectively illustrate the robustness, stability, and optical performance of the device, making it suitable for use in the monitoring and evaluation of schistosomiasis control programs in endemic settings. ... Read more

Schistosomiasis is a neglected tropical disease thatis predominantly diagnosed by conventional microscopy in Sub-Saharan Africa. However, effective diagnosis by conventional microscopy is limited by multiple technical and logistic barriers.Alternative diagnostic techniques are needed. The Schistoscope is a digital optical device that has been designed to support microscopists for the detection of schistosomiasis in endemic resource-limited settings. Aim: A user-centered design approachwas used to assess the usability and user-acceptance of the Schistoscope compared to conventional microscopy in the Federal Capital Territory, Abuja, Nigeria. In this study, usability and acceptance are defined as being easy-to-use, efficient, and suitable in the daily workflow by end-users. Methods: Using a qualitative conventional context analysis approach, a mixedmethods questionnaire was used to elucidate themes related to the usability and user-acceptance of the device. Participants included trained microscopists and university students (n=17). Results: Participants answered both ranked and open questions. Overall the device’s use was considered to be easy and acceptable
in the routine workflow of a microscopist. The auto-scan feature was considered to have added value. Critical feedback regarding aesthetics of the device, particularly related to size, was noted by the participants. Conclusion: The usability approach used in this study elucidated valuable insights of end-users. The Schistoscope was very well perceived by both medical students and trained microscopists. Critical feedback will be used to further improve the next iterative design of the device. ... Read more

Schistosomiasis is a neglected tropical disease of Public Health importance affecting over 252 million people worldwide with Nigeria having a very high number of cases. It is caused by blood flukes of the genus Schistosoma and transmitted by freshwater snails. To achieve the current global elimination objectives, low-cost and easy-to-use diagnostic tools are critically needed. Recent innovations in optical and computer technologies have made handheld digital and smartphone-based microscopes a viable diagnostic approach. Development, validation and deployment of these diagnostic devices for field use, however, require the optimisation of its optical train for the registration of high-resolution images and the realisation of a robust system design that can be locally produced in low-income countries. Field research conducted in Nigeria with active involvement of key stakeholders in research and development (RD) led to the design of an initial prototype device for the diagnosis of urinary schistosomiasis, called Schistoscope 1.0. In this paper, we present further development of the Schistoscope 1.0 along two parallel design trajectories: A Raspberry Pi and a Smartphone-based Schistoscope. Specifically, we focused on the optimization of the optics, embodiment design and the electronics systems of the devices so as to produce a robust design with potential for local production. ... Read more