Cardiac related disease is the number one cause of death in the Western world, including Norway. Echocardiography is the most important imaging tool for the cardiologist to assess cardiac function. An echo examination of the heart is real time, cost effective and can be performed without discomfort to the patient and without harmful radiation. These are great advantages compared to other medical imaging modalities. The key performance indicators for an echo lab are diagnostic accuracy and productivity. For the patient a fast and accurate diagnosis is important for receiving the right treatment. For the hospital the productivity of the echolab has direct economic implications, and early accurate diagnosis alleviates the need for involving more expensive imaging modalities. A typical cardiac ultrasound examination takes 30-40 minutes; while only about half of this time is spent efficiently acquiring diagnostic quality images. The wasted time is spent manually adjusting image acquisition parameters, searching for optimal views and performing manual measurements. The vision for this project is to create an INtelligent Cardiovascular Ultrasound Scanner - INCUS - that is capable of assisting the user to increased productivity and accelerated decision making. This will be achieved by introducing intelligent algorithms in the scanner that can exploit knowledge from expert users and previously acquired data and learn from this. We will build on recent research in machine learning that has led to a set of approaches referred to as Deep Learning. Deep Learning has boosted the performance substantially for many applications like speech recognition and image classification. The ambition is to offer a new ultrasound scanner providing increased diagnostic confidence as well as significantly enhanced productivity compared to ultrasound scanners that exist on the market today.
Project leader: Erik Normann Steen
Institution: GE VINGMED ULTRASOUND AS