Cervical cancer poses a huge challenge to global health (Sung et al., 2021). To provide access to cervical-cancer care to women in Low Resource Settings (LRS), the startup GIC Space is developing the C-Spec and GICMED application. With these innovations the company aims to offer remote Cervical Cancer Screening (CCS) and diagnosis, and a more comfortable experience than with a regular speculum. With the implementation of Artificial Intelligence (AI) algorithms, diagnoses will partly be automated and this requires a new workflow for the app.

Literature revealed that the deployment of information systems (such as smartphones) is needed to offer professional support to medical personnel in LRS in performing and managing CCS. The objective of this master thesis was to redesign the existing GICMED application to enhance the interaction between patient and healthcare provider, and assist in keeping track of patient data more efficiently. GIC Space set the challenge to design for both highly trained medical professionals, like general practitioners and gynaecologists, and less trained healthcare providers such as nurses and midwives. By responding to medical practitioners' lack of expertise and experience to accurately diagnose cervical cancer, the GICMED application could increase screening reliability in LRS. In addition, allowing healthcare providers to have more time for patients, will enable them to better empathise with their client and will lead to better care.

The usability of the current application was inspected with a cognitive walkthrough. Insights about desirable features were gathered by means of a benchmark with competitors. Four User Journey Maps were created to explore design opportunities. To determine whether the concept would be viable, the Lean Startup Model was deployed to rapidly develop a Minimum Viable Product (MVP). The MVP was assessed through user testing in order to determine its desirability.

Since the redesign combines paper work and implements EMRs, patient data can be stored and accessed digitally, allowing personnel to work in a efficient manner and devote more time on providing care. EMRs enable exchange of patient data between healthcare facilities, which is particularly beneficial for reducing loss of follow-up. In order to establish an accurate diagnosis and management plan, AI should solely serve as a second opinion and still rely on the user's own judgement. To avoid AI’s diagnosis from simply being adopted, users should be requested to share their findings and diagnosis first, before AI hands over its conclusion.

During user tests, the redesigned application received an average score of 82.8 out of 100. Participants unanimously agreed that the app helps healthcare providers to keep track of patient data more efficiently and two out of three subjects agreed the app allows both user groups to devote more attention to patients and to express empathy.

This master thesis highlights that potential users are open to trust the result given by AI. Users will first provide their own judgment and then utilise AI’s findings to come to a final conclusion. This thesis thereby affirms the potential for the integration of AI to partly automate diagnosis for cervical cancer.