Örjan Smedby

In modern healthcare, imagery is increasingly used to make diagnoses and choose the most appropriate treatment. One example is when judging whether a constriction in the heart’s coronary artery is so pronounced that it should be treated with a balloon catheter or not. Another situation involves evaluating the effects of chemotherapy on a tumour. With an MRI exam before and after treatment, it is decided whether the effect is enough to justify the side-effects and costs associated with the treatment. To solve these problems, computers are preferably used to get numerical measurements of the degree of illness from the images, known as image biomarkers. This often involves measuring volume, thickness or length of an anatomical structure in the image. This presupposes that the computer can distinguish the object from surrounding tissues, which is called segmentation.
Örjan Smeby’s research team works to develop new methods for automatic segmentation that are so precise that they can be used to make medical decisions and at the same time, so fast that they meet the demands of the stressful workflows in healthcare.
Another area of research involves using CT scans to make a quantitative description of the spicules’ structure to assess the degree of bone brittleness.