Searchable abstracts of presentations at key conferences in endocrinology
Endocrine Abstracts (2022) 81 OC13.5 | DOI: 10.1530/endoabs.81.OC13.5

1Endocrinology - Sant’Andrea Hospital, Sapienza University of Rome; 2Department of Surgical and Medical Sciences and Translational Medicine, Sant’Andrea Hospital, sapienza University of Rome; 3Endocrinology, Department of Clinical and Molecular Medicine, Sapienza University of Rome; 4Endocrinology - Sant’Andrea Hospital, Sapienza University of Rome, Clinical and Molecular Medicine, Rome, Italy


Adrenal nodular disease is a frequently increasing in the general population with an incidence that reaches almost 10% in the seventh decade of life. More and more evidences show these lesions discovered through diagnostic imaging (CT, MRI) performed for other medical problems (incidentalomas). New radioimaging techniques, exploiting the quantitative variables of imaging, permit to identify an hypothetical pathological tissue. We have applied this potential in a retrospective series of 72 patients of both sexes with single adrenal lesion >1 cm in dimension with adrenal incidentalomas followed at our center. Patients were studied following ESE/ENSAT current criteria practice guideline in order to exclude any hormonal hypersecretion considering in this study only not secreting and cortisol secreting adrenal masses by dexamethasone-suppression test (DST). Based on cortisol value they were divided in two groups: functioning (32) and non-functioning (40) adrenal incidentalomas with cortisol values >50 nmol/l and <50 nmol/l respectively. Machine learning concept, through different algorithms offers the possibility to study several biological processes obtaining quantitative information from imaging and correlating it with outcomes. Radiomics is an emerging technique that translates radiological images into quantitative data to yield biological information and permits an in depth radiological characterization, thus improving diagnosis, decision support, and follow up monitoring. It is a multistage process in which features based on shape, pixel densities, and texture are extracted from CT or MR images. Each incidentaloma was studied in the preliminary non-contrast phase with a specific software (Mazda), surrounding a region of interest within each lesion. 314 features were extrapolated. Mean and standard deviations of features were obtained and the difference in means between the two groups was statistically analyzed. ROC curves were used to identify an optimal cut off for each variable and a prediction model was constructed via multivariate logistic regression with backward and stepwise selection. A 11-variables prediction model was constructed and a ROC curve was used to differentiate patients with high probability of functioning incidentalomas. Using a specific threshold value we obtained a sensitivity of 93.75% and a specificity of 100% in diagnosing functioning incidentaloma. Based on these results, CT texture analysis appears a promising tool in the diagnostic definition of adrenal incidentalomas.

Volume 81

European Congress of Endocrinology 2022

Milan, Italy
21 May 2022 - 24 May 2022

European Society of Endocrinology 

Browse other volumes

Article tools

My recent searches

No recent searches.