In order to facilitate artificial intelligence in the diagnosis of FGS we will use Centre for Bilharzia and Tropical Health Research´s (CBTHR) repository of 15.000 genital images from previous FGS research and demonstrate the proof-of-principle of an “automated visual evaluation” algorithm, (Main objective (MO).1). This can be used in the back-end of the diagnostic support devices. In order to achieve this, we will:

• classify the images for training and testing
• apply a deep learning-based object detection method and the faster region-based convolutional neural network algorithm.

Tasks to be carried out:

• Classification of lesions using images, follow-up data, and laboratory results
• Machine learning, training, validation

The project aims to provide sensitive and specific objective diagnostic support for recognition of FGS (MO.1) amongst health professionals in schistosomiasis endemic areas as well as effective diagnosis of cervical cancer with an enhanced VIA method.

Using spectral signatures from cervical images in real time we will:
Adapt an established doctor’s office innovative spectral imaging colposcope to be used in the field in rural settings by doctors and trained nurses alike.

This point-of-care, multi-spectral, dual diagnostic tool for FGS and cervical cancer will be made economic, usable off-grid (power and network), with fit-for-purpose specifications and easy-to-use operation.

The multi-spectral tool will include high quality LED illumination and utilise a mobile device for control, image acquisition and data analysis. The multi-spectral sensor will be capable of producing real-time video stream and still images at several spectral bands in the visible and infrared parts of the spectrum and can also produce fluorescent images with UV illumination.

The high-definition spectral video stream can be analysed on the fly and classified according to the spectral profile of each pixel, providing a match with patterns of known pathologies. In addition, IR vision can see under the surface of the tissue and discover characteristic features, e.g. eggs of parasites, etc.

Data will be acquired for development of appropriate classifiers and a validation trial will be carried out to evaluate the performance for diagnosis of FGS and cervical cancer using a VIA protocol in the hands of both specialists doctors and trained nurses.

Tasks to be carried out:

• Alpha multi-spectral diagnostic device (M 1-7).
• Initial test of Alpha device (Windows tablet) TRL-7 (M 7-8).
• Training of Field Physicians (M 9-11). Collection of field data with Alpha multi-spectral, dual diagnostic tool for FGS and cervical cancer (M 12-15).
• Field data analysis, software update (M 14-20).
• Field validation of alpha device (M 21-32).
• Development of low-cost field device (Beta) (M 12-29).
• Development of mobile version of software (also using machine learning from WP1) (M 14-29).
• Internal technical/clinical verification/validation, optimization and documentation of the multi-spectral, dual diagnostic tool for FGS and cervical cancer (M 30-48)

We seek to facilitate innovation on the clinical management and prevention of Female Genital Schistosomiasis (MO.2) through the following work package objectives:

(1) Make a scalable online/App-based training course for frontline primary health professionals and gynaecologists (those will meet patients exposed to schistosomiasis and who use speculums in their work)

(2) Evaluate and certify their investigation techniques and diagnostic skills

(3) Evaluate the online training course.
The training will contain basic knowledge on FGS (i.e. manifestations, cycle, distribution), diagnosis (all-vaginal-walls’ inspection, visual match-making, analogue diagnosis training), treatment, problems (side effects, ethics, regulations).

Training courses will be tailormade (online, offline, in-classroom) by health professionals specialised in FGS, ICT eLearning experts, pedagogical experts, multi-media, and users.

The evaluation of the training will address the professional competence and clinical efficacy with baseline and post-training assessments.

Co-creation workshops will inform new curricula for nursing and medical doctors, and post-graduate training (inclusion into the programme for Continued Medical Education (CME)).

Tasks to be carried out

Task 1 Decisions on FGS clinical content.

Identification of symptoms (FGS can mimic various gynaecological ailments) the typical lesions, follow a decision tree on management and treatment.

• Acceptability and validation studies will be performed by the social scientists (WP5).
• A workshop of gynaecologists and subject experts will take place to advise on content to be included in the course and ensure co-creation.
• A rural health professionals’ reference group will evaluate the appropriateness of all aspects of the training.

Task 2 Development of a course for post-graduate doctors and nurses
Development of (online/offline/in-classroom) digital/in-situ FGS training system for and with health professionals based on the content produced in this WP.

Task 3 Development of medical and nursing student curricula

In order to facilitate the diagnosis of disease (MO.3), training of health professionals (MO.4), and steps for inclusion of FGS into health policy documents and management protocols (MO.4).

This work package aims to help us understand the perspectives of the healthcare providers and patients – the enablers and constraints – towards the diagnostic procedures for FGS and cervical cancer.

Work package objectives:
1. Explore healthcare providers’ views on the usability of the modern technological innovation for the screening of FGS and cervical cancer

2. Analyse the practical implications of the use of modern technological innovation to conduct co-screening for FGS and CC

3. Explore patients’ views on the use of the device to provide the FGS and CC screening services

4. Understand social norms and attitudes towards examination for FGS and CC.

5. Implement nominal group technique as a consensus-building mechanism in order to inform the policy, based on study findings

6. Investigate the documents of the involved departments, regions, and countries (Institutional Ethnography) in order to be able to identify possible points for improvement, based on the results of the project.

A qualitative approach will be applied, informed by Institutional Ethnography and Lipsky’s theory of street-level bureaucracy theory of change. We also aim to generate evidence on FGS suffering and how patients perceive the diagnostic procedures, alongside the experiences of healthcare providers on if the new multi-spectral tool improves or inhibits the diagnostic procedures.

In-depth interview (IDI) guide approach will be used to generate data. Furthermore, we will explore health policy documents that may influence health professionals’ thinking and behaviour, identifying how these policy documents could be improved.

Building on evidence generated through the review of policy documents, experiences of patients, alongside healthcare providers, nominal group technique (NGT), as a consensus-building mechanism, will be used as a build-up to policy reform and/or propositions. Institutional ethnography is another key approach to achieving our objective, and this is premised on the view that how genital disease is defined and perceived through public documents and policy will influence local practice.

Therefore, health workers will also be interviewed as part of generating insight about the influence of public policy and policy documents on their everyday interaction with patients and how they make implementation decisions based on professional discretion, available resources, costs and practical arrangements; a phenomenon supported by Lipsky’s theory of street-level bureaucrats.

The use of qualitative research in parallel will explore opportunities for mentorship and interventions e.g. on water contact or dissemination of knowledge.

The study is anticipated to improve our understanding of people’s concerns, behaviours and other related contextual factors, thereby generating evidence which will be communicated back into the local community, both through discussions with the women, but also through interaction with local health professionals, educators, and policymakers.

A focus on health workers practices as well as policy as ruling actions and approaches to health work will generate increased knowledge about the relationship between policy and health work, as well as how this would influence and interacts with people’s daily lives.

Such insight is significant when discussing the revision of health policies and practice among health staff.

Tasks to be carried out

Task 1: Institutional ethnography – data collection and analysis (MO.3 and MO.5)
Document reviews, ethnographic exploration of policy documents, and staff interviews.

Task 2: In-depth interviews with selected health professionals (MO.3 and MO.5)
A feasible number of health professionals in the selected clinics will be interviewed following an interview guide with focus on the FGS diagnostic and management procedures, in line with the existing policies.

Task 3: Nominal group technique (NGT) with key (6-8) members of interviewed health professionals will be held (MO.3 and MO.5). Summary of the findings from the interviews will be presented and key issues will be raised for brainstorming and for proposing areas of improvements.

Task 4: In-depth interviews with selected FGS patients (MO.5)

Women suspected to have FGS and subjected to the diagnostic procedures using the novel multi-spectral device, will be interviewed through an interview guide, in order to explore their experiences of the diagnostic intervention.

Task 5: Nominal group technique with key (6-8) members of interviewed patients will be held (MO.5) Summary of the findings from the interviews will be presented and key issues will be raised for brainstorming and for identifying areas of improvements regarding the use a point-of-care, multi-spectral, dual diagnostic tool for FGS and cervical cancer.

The researchers will transcribe data from interviews and analyse interview data and policy data.

The analysis will be fed into a) scientific publications b) policy briefs c) popular dissemination. Insights from research will also be applied into various relevant courses in UiAs master programme, e.g. the course “Gender, culture and everyday life”.

Sociocultural research

There is a growing realization that boundary walls separating different disciplines in their engagement with research and/or health interventions compromise their potential value.

Work Package 5 (WP5), which is a social science component of the Dual screening by Spectral Artificial Visual Examination for Female Genital Schistosomiasis and Cervical Cancer (DUALSAVE-FGS), embodies how different disciplines can come together and work towards a common goal.

For example, while this project is about trying a new technology to efficiently diagnose female genital schistosomiasis (FGS) and cervical cancer (CC), we are equally concerned about how will this technology be received by both health professionals and patients, alike, let alone the cultural appropriateness and acceptance by the local people.

Therefore, the contribution of WP5 is to explore the potential barriers and facilitators to the uptake of this new intervention, so that corrective actions can be designed and implemented. We will study the behaviour of treatment seeking women from the moment they experience discomfort though the development of the illness.

This project also provides an opportunity to build research capacity while potentially transforming the lives of women who fall prey to FGS and CC.

Consequently, students are being recruited to use this project to enroll for their Masters and Doctoral qualifications. These students will also benefit from a host of varying research expertise nested in the team of principal investigators, co-investigators and investigators.

It is anticipated that this project will create a rich platform for cross-pollination and peer-mentoring by students from participating countries. Students’ research will cover a range of thematic areas, including barriers and facilitators to FGS care, from healthcare worker, patient and policy maker’ perspectives.

They will also explore how messages for FGS and CC are packaged, including materials, wording, and the channels used. Furthermore, local knowledge systems in the understanding and management of FGS and CC will be explored. The research will be conducted in Eswatini, Mozambique and  South Africa.    

To facilitate cooperation and have a smoothly running research consortium with optimal impact (MO.4).

Tasks to be carried out:

The OUS Research Administration Office (Norway), CBTHR (South Africa), a project administrator and three field coordinators will support the work. The Project Steering Committee, headed by the coordinator, will be the decision- making body. The Steering Committee is further assisted by the External Advisory Board, the Privacy Advisory Group and Institutional Representatives/ Academic and Industry Board, which will consist of the members from each participating institution. The management will:

Task 1. Administer and report Grant management and quality assurance will jointly be done by OUS, UKZN and CBTHR. CBTHR will help with accommodation, streamline induction into country issues (culture, security, constraints), information dress code, make rosters for staff and scientists, talk to next of kin, check staff work, assist in permissions when necessary.

Task 2. Keep overview of all ethical issues: Ensure that the entire team undergoes a yearly course in ethical and privacy issues, Confirm sub- protocols, ethics applications, reports and scientific papers. Liaise with referral institutions in preparation for cases.

Task 3. Recruit and increase competency of local staff and scientists. Potential staff and students will be screened by at least two people from the senior team; references will be contacted; high stress-tolerability and team working abilities are important in addition to the academic and practical merits. Provide hands-on research experience in multi-disciplinary, inter-sectoral clinical and field situations. Facilitate communication in research and negotiation settings (present, discuss with community, vet and train in the use of interpreters)

Task 4. Organise meetings: Virtual data management plan meetings. Virtual FGS diagnosis consensus meeting. Annual “Summer-Winter School” where study scientists and staff will meet and present. Meetings to review analyses. Provide meet-the-scholar sessions, facilitate tutorship, and lecture. Organise conference participation to present results. Organise workshops that will be held back-to-back with conferences, thus enabling the partners to present to and liaise with conference participants and to discuss with other donors. Researchers with abstracts and findings will participate.

The objective is to ensure compliance with the ‘ethics requirements set out in this work package.

Tasks to be carried out:

This work package sets out the ‘ethics requirements’ that the project must comply with.