The thyroid gland is a hormone-producing gland located in the front of the neck, below the larynx. Diseases of the thyroid gland can be divided into two main groups: those that involve a change in hormone production and secretion (hyperthyroidism and hypothyroidism), and those that affect the gland's shape and size (goiter, thyroid cancer) [1].
During a medical ultrasound examination (US), thyroid cancer appears as irregular tissue, namely nodules. However, there are different types of nodules that look similar but have different origins than malignant cancer. It is therefore crucial to correctly classify these nodules in order to choose the appropriate treatment.
When the appearance of the nodule in US images is not obvious or well-defined, it is usually necessary to perform an ultrasound-guided fine-needle aspiration biopsy with cytological evaluation (FNC) to diagnose the nodule. A study conducted by UNN radiologists [2] concludes that an expert can distinguish between benign and malignant nodules using only US images without FNC, and even differentiate between multiple histopathologies in thyroid nodules. Because there is sufficient relevant visual information in US images, there is potential for a reduction in the number of invasive and time-consuming FNC biopsies and diagnostic procedures.
Artificial intelligence provides tools that can help us correctly classify nodules and estimate the probability of malignant cancer. Therefore, SPKI is starting this project in collaboration with UNN, Helse Nord IKT, and Visual Intelligence (UiT) to develop an ultrasound decision support system for patients in Helse Nord.
The project also includes the appointment of M
aria Bolomiti, a clinician in a 50% Ph.D. position over 6 years funded through research grants from Helse Nord, which you can read more about here. Maria will contribute her expertise in clinical medicine and collaborate closely with the machine learning community at UNN to ensure a comprehensive approach to the research.
Through this project, the aim is to develop an effective tool that can contribute to a better and more accurate diagnosis of thyroid cancer, as well as reduce the burden on patients and healthcare professionals by minimizing the need for unnecessary invasive procedures.