ESPE2024 Rapid Free Communications Pituitary, Neuroendocrinology and Puberty 2 (6 abstracts)
MRI-based radiomics of the pituitary gland is highly predictive of CentralPrecocious Puberty in girls: pilot study
Maddalena Petraroli 1 , Viviana Dora Patianna 1 , Michele Maddalo 2 , Marco Masetti 3 , Francesca Ormitti 4 , Alice Fulgoni 1 , Margherita Gnocchi 3 , Benedetta Piccolo 1 , Emanuela Claudia Turco 1 , Nicola Sverzellati 1,5 , Caterina Ghetti 2 , Susanna Maria Roberta Esposito 1,3 & Maria Elisabeth Street 1,3
1Unit of Paediatrics, University Hospital of Parma, Parma, Italy. 2Medical Physics Unit, University Hospital of Parma, Parma, Italy. 3Department of Medicine and Surgery, University of Parma, Parma, Italy. 4Neuroradiology Unit, University Hospital of Parma, Parma, Italy. 5Unit of Radiological Sciences, University Hospital of Parma, University of Parma, Parma, Italy
Introduction: The diagnostic gold standard for Central Precocious puberty (CPP) is the gonadotropin-releasing hormone (GnRH) stimulation test. MR imaging of the brain (MRI) and the hypothalamus-pituitary region is required to exclude organic causes.
Objective: The aim of the study was to explore a radiomic model that could assist physicians in the diagnostic workup of CPP.
Methods: 45 girls with a confirmed diagnosis of CPP and 47 age-matched pre-pubertal female subjects (control group) were retrospectively enrolled. Two readers (R1, R2) with different levels of expertise on pediatric neuroradiology blindly segmented the pituitary gland on MRI studies for radiomic features (RFs) calculation and performed a manual estimation of pituitary volume (ellipsoid approximation - EA). Cross-validated linear discriminant analysis was used to develop for each reader both a radiomic model and a clinical reference model based on the manually-measured pituitary volume. Radiomics was compared against the EA in terms of: 1) predictive performances (metrics: ROC-AUC, accuracy, sensitivity and specificity); 2) reliability of predictors between readers (metric: intraclass correlation coefficient, ICC); 3) consistency of performance metrics between readers (absolute difference between R1 and R2 mean performances).
Results: The radiomic model significantly improved the diagnostic sensitivity with respect to the EA of the pituitary gland (0.78 versus 0.69 in validation set, P <0.001) and achieved a well-balanced trade-off between sensitivity and specificity. Radiomic predictors demonstrated higher inter-reader reliability (ICC>0.57) with respect to the EA predictor (ICC=0.46). Moreover, the radiomic model showed a greater consistency between R1 and R2 findings, with a mean difference among all performance metrics of 0.06±0.04, which was lower than the difference observed for EA (0.10±0.02).
Conclusion: Radiomics of the pituitary gland alone demonstrated a greater potential in diagnosing CPP and an increased reliability with respect to EA. This opens a way to diagnosing CPP based on MRI of the pituitary gland in addition to clinical and hormonal data. However, further studies are warranted to validate these preliminary data.
Article tools
My recent searches
My recently viewed abstracts
Authors
ESPE Abstracts
© Bioscientifica 2024 |
Privacy policy |
Cookie settings
BiosciAbstracts
Bioscientifica Abstracts is the gateway to a series of products that provide a permanent, citable record of abstracts for biomedical and life science conferences.
Find out more