Diagnostische echografie en artificiële intelligentie bij de karakterisering van adnexiële massa’s: naar een 3d classificatiemodel

The objective of the thesis is to develop knowledge on gynecologic ultrasound. Specifically I will concentrate myself on the fields of artificial intelligence and three- dimensional ultrasound applied to the ovarian cancer first and later also focus on the three- dimensional ultrasound to endometrial cancer and uterine malformations. I will participate in the already ongoing research projects and take the chance to start new projects based on them. Once completely developed and well-established, artificial intelligence will permit the automatic pattern recognition of basic sonographic features of ovarian masses, and this will probably help sonographers in the selection of malignant masses, according to what is nowadays known on this area, and sort the cases that need particular follow up or secondary care. On the other hand, three- dimensional ultrasound is a challenging skill that could eventually help gynecologists to better understand the nature of ovarian masses, endometrial cancers and uterine malformations compared to the two-dimensional ultrasound, and therefore it may improve the specificity of the diagnosis. However further studies are needed in order to evaluate this. Artificial intelligence and three-dimensional ultrasound is a continuously growing branch of gynecologic ultrasound, and many improvements are expected to increase diagnostic confidence and optimal clinical workflows.