Given multiple synthetic datasets that describe hidradenitis suppurativa through different information modalities, each with different levels of coverage, completeness, and selection bias, can data scientists recover the underlying structure of the disease — its latent subpopulations and progression stages — consistently across datasets, and use what each dataset captures well to complement what the others miss, rather than capturing the artefacts of how each dataset was collected?

How might we design a technological solution that reduces the mental burden associated with the everyday management of active paediatric cancer treatment, helping caregivers organise and anticipate tasks while also supporting the progressive autonomy of children and adolescents?

How can we objectively detect and monitor changes in sputum characteristics (such as colour, volume, and viscosity) over time, using accessible technologies such as computer vision, mobile health, or artificial intelligence, to enable earlier identification of bacterial exacerbations and support timely clinical decision-making?