UX Design in AI-based solutions for Mental Health: the FAITH experience

UX Design in AI-based solutions for Mental Health: the FAITH experience

Combining AI and UX Design expertise, Deep Blue members of FAITH Project reveal their approach to enhance digital solutions for cancer patients’ mental health , with a specific focus on explainability and data collection methodologies.


In an ever more digitalised world, the integration of cutting-edge technologies in strategic sectors such as healthcare is both a huge opportunity and an imposing challenge. The FAITH project  seeks to address this, by providing an Artificial Intelligence application that identifies and analyses depression markers in people that have undergone cancer treatment, increasing awareness of their mental health situation. As the FAITH solution will directly engage with both patients and clinicians, smooth and reliable communication between the two parties is crucial. In order to achieve this goal, FAITH partner Deep Blue lends its expertise in Human Factors, with a specific focus on  UX Design aspects. This innovative methodology paves the way for the design of ground-breaking digital supportive solutions, developed around end-users’ requirements and needs.

eHealth UX Design is a particularly hot topic in Human Factors, so much so that the CIEHF – Chartered Institute of Ergonomics and Human Factors published an article on FAITH’s Clinician Interface UX Design in its magazine “The Ergonomist”. In the article, the Deep Blue team collaborating in FAITH, Giuseppe Frau, Tommaso Vendruscolo and Alessandro Tedeschi Gallo, provided a detailed insight on how FAITH’s UX Design was built around healthcare professionals and their work by keeping a close, reiterative relationship with partner hospitals’ doctors.


The two key aspects for a functional HMI: Explainable AI and Data quality control

Through its solution, FAITH would present healthcare providers with advanced warnings to allow timely intervention. To test the FAITH concept, trial sites in Madrid and Lisbon are assessing the solution to ensure its usefulness and scope of action, involving real end users (both clinicians and patients). Therefore, it is crucially important to establish a good communication flow between the machine (gathering data from patients) and the doctors (who monitor the data and interpret them to detect fluctuations in patients’ mental health). Functional Human-Machine Interaction (HMI) ensures that both human and machine actors are working with each other. Based on these simple assumptions, the DBL team started by defining users and their needs.

In FAITH’s specific case, clinicians working with cancer patients need not only easy access to collected data from the trial participants, but also to check data quality and retain awareness about which data is entering the system.That is because it will be on this data that the AI will train to recognise early signs of mental health deterioration. As FAITH gathers a vast amount of data, both passively (i.e. digital tracking) and actively (i.e., medical questionnaires), the FAITH Clinician Dashboard is built to facilitate access and interpretation of data from doctors, in addition to keeping a close eye on user/App interaction. The latter is particularly relevant, as it ensures a continuous flow of data from users, consequently ensuring the data’s good quality. 

Giuseppe Frau acknowledges the importance of HMI in designing digital tools. “In FAITH, Deep Blue is trying to create an AI interface that is explainable, an Artificial Intelligence which shares its decision-making process with its users. This feature would help build trust in humans interacting with the tool, as they can discern the causes behind the machine’s reasoning. It is not an easy task, but we started in an early stage to implement explainability in FAITH’s AI”, says Deep Blue Head of Tech and FAITH WP2 Lead.


The Human Factors role in Digitalisation

In our increasingly digitalised world, with Artificial Intelligence and algorithms being integrated into almost all fields of human activity, it is crucial to look after Human-Machine Interaction, as machines work side by side with humans, making it necessary to establish seamless and fluent communication. 

In this context, Explainable Artificial Intelligence might be the answer to building a relationship of trust with the machines. In fact, as Explainable AI provides the reasoning behind its decision-making process, it seems to be more suitable for eHealth solutions such as the FAITH supportive tool. With its vast expertise on Human Factors and the human role within complex systems, Deep Blue greatly contributes to FAITH’s tool success by keeping the human end-user at the centre of FAITH’s multi-layered ecosystem of doctors, healthcare providers, cancer patients and survivors, data analysts, IT experts, Artificial Intelligence, smart devices, and algorithms.

To know more about the project, visit the website and read the partners’ articles.


Get in touch with us