Towards a unified rheumatic heart disease imaging dataset.

Abstract

Artificial intelligence (AI) has the potential to improve access to services for underserved populations. Rheumatic heart disease (RHD) is set to be a condition that may benefit significantly from the innovations made possible through AI research [ [1] ]. As with essentially all AI research, developing an AI model for RHD detection will be contingent upon the presence of robust datasets. A unified and collaborative approach to the consolidation of RHD imaging datasets may expedite the development of these algorithms and enable higher algorithmic performance.