The researchers from the Medical University of Bialystok (Marta Wielogorska-Partyka, Katarzyna Siewko, Anna Poplawska-Kita, Angelika Buczynska, Piotr Mysliwiec, Adam Jacek Kretowski, Agnieszka Adamska) along with co-authors from other institutions published the article „Patient classification and attribute assessment based on machine learning techniques in the qualification process for surgical treatment of adrenal tumours.” in Scientific reports.
Adrenal incidentaloma (AI) is an asymptomatic tumour of the adrenal gland that is detected using CT scans and presents a clinical challenge. The aim of the groundbreaking study by UMB scientists was to compare different machine learning techniques in qualifying patients for adrenalectomy and to select the most accurate algorithm. The analysis included data from 33 patients with incidental adrenal tumours, using algorithms such as SVM, KNN and random forest.
The results showed that the SVM classifier achieved the highest accuracy (91%) in identifying patients requiring adrenalectomy, outperforming the accuracy of medical specialists (64%). Key features for classification included tumour homogeneity, maximum tumour diameter and patient obesity. Despite the limitations of the small sample size, the study highlights the huge potential of artificial intelligence in supporting patient surgical decisions. For years now, artificial intelligence has been one of the areas in which the Medical University of Bialystok remains one of the regional leaders.
This work was partially supported by W/WI-IIT/3/2020 and WZ/WI-IIT/4/2023 grants from Bialystok University of Technology and funded with resources for research by the Ministry of Science and Higher Education in Poland.
Link to the article: Patient classification and attribute assessment based on machine learning techniques in the qualification process for surgical treatment of adrenal tumours.