Piotr Durniat, Tomasz Hałas, Jakub Kukowski, and Wiktor Sadowy from our Faculty created a calculator for medicine doctors. The solution proposed by our students aims to facilitate the work of endocrinologists. It provides support to doctors during diagnosis.

The tool is available online for free. Doctors enter data into the calculator based on the ultrasound image.

– Our calculator allows endocrinologists to quickly analyse a patient's thyroid in order to assess its condition. The doctor provides the tool with the description of what they see in the image. Based on this information, the tool returns the result and indicates the tumour category – says Piotr Durniat.

Thyroid nodules are divided into several categories, depending on their characteristics (e.g., shape, edges, presence of microcalcifications, vascularisation). Each category corresponds to a different risk of malignancy and requires different recommendations for further diagnostics.

– The special calculator will aid endocrinologists in the diagnosis process – emphasises Med. Dr. Tomasz Tomkalski, endocrinologist and head of the Clinic of Endocrinology, Diabetology and Internal Diseases Wmed at the Marciniak Hospital in Wrocław.

Benefits for the patient

The calculator is an additional task. It was developed as part of a scientific and implementation project aimed at creating a system that would classify the malignancy of tumours based on ultrasound images.

– We are not medical experts. Doctors tell us what the demand is, and based on that, we try to find a solution to the problem –  explains Wiktor Sadowy. The aim of the project is to reduce the number of removed thyroids.

– The biopsy results are difficult to interpret due to the thyroid's vascularity. Because of this ambiguity, thyroids are excessively removed – explains Jakub Kukowski.

– The issue particularly arises when the nodule is classified as category 3 or 4 on the EU-TIRADS scale (a scale used for assessing focal thyroid changes). Here, even a biopsy cannot determine whether the thyroid should be removed, so the procedure is performed usually because of the risk. Thanks to our project, artificial intelligence is intended to assist in assessments and prevent unnecessary excisions – notes Tomasz Hałas.

The students are supervised by Prof. Halina Kwaśnicka from the Department of Artificial Intelligence.

Scale of the problem

The occurrence of thyroid nodules, including thyroid cancers, is one of the main endocrine-related issues in the population of Lower Silesia. It is estimated that palpable thyroid nodules are found in as many as 5.3% of women and 0.8% of men. These changes usually require further diagnosis, such as a thyroid biopsy.

Thyroid cancer is the most common malignant tumour of the endocrine glands. It occurs in most cases between the ages of 40 and 50.