Dermatology Image Quality Assessment (DIQA): Artificial Intelligence to ensure the clinical utility of images for remote consultations and clinical trials

Authors

Ignacio Hernández Montilla, Taig Mac Carthy, Andy Aguilar, Alfonso Medela

Summary

Dermatological imaging is extensively used both in clinical practice and clinical trials, especially for diagnosis and severity assessment. The problem is that images require a minimum visual quality in order to be analyzed, be it by a doctor or an algorithm. Images that lack visual quality can derail the clinical process, disrupt clinical trials and pose a risk to patient safety.

Clinical images are widely used in dermatology to capture the state of conditions in a non-invasive way. However, these images are subject to huge variability in lightning, distance to the lesion, focus and other factors, which impact the semantic content and therefore the clinical usefulness of an image.

To solve this problem, we developed an AI-based tool that gauges the dermatological image quality and ensures the quality and clinical utility of images during remote consultations and clinical trials.

DIQA shows promise as a quality-check tool that improves remote consultation and clinical trials, especially those in which patients with skin pathologies report their condition remotely. Furthermore, artificial intelligence can be implemented into CADx systems to ensure that only high-quality images are uploaded.

Read full text

Available at https://doi.org/10.1016/j.jaad.2022.11.002.