Cardiac ultrasound is the most common way to look at and measure heart function. It is used to detect many different types of heart disease, and to guide treatment once heart disease is detected. More than 75,000 cardiac ultrasounds are performed per year in New Zealand. Cardiac MRI is a more expensive and time consuming test, but gives better images of the heart. In this study, we aim to develop a method of generating new images from routinely acquired cardiac ultrasound, by training artificial intelligence models to produce MRI-like images from standard ultrasound images. If successful, the method would be applicable to any newly acquired cardiac ultrasound, and could potentially transform our methods of assessing heart disease.