Noninvasive radiomics-based method for evaluating idiopathic central precocious puberty in girls |
| |
Authors: | Hongyang Jiang Zhenyu Shu Xiaoming Luo Meizhen Wu Mei Wang Qi Feng Junfa Chen Chunmiao Lin Zhongxiang Ding |
| |
Affiliation: | 1.Department of Radiology, Zhejiang Provincial People’s Hospital, People’s Hospital of Hangzhou Medical College, Hangzhou, China; 2.Department of Pediatrics, Zhejiang Provincial People’s Hospital, People’s Hospital of Hangzhou Medical College, Hangzhou, China; 3.Department of Radiology, Affiliated Hangzhou First People’s Hospital, Zhejiang University School of Medicine, Hangzhou, China |
| |
Abstract: | ObjectiveTraditional approaches that involve measuring the height and volume of the pituitary by magnetic resonance imaging (MRI) are unreliable. We investigated the use of a more accurate method using texture analysis to evaluate idiopathic central precocious puberty (ICPP) by MRI.MethodsIn total, 352 texture features of the pituitary were extracted from 12 healthy girls and 18 girls with ICPP. A LASSO regression model and linear regression model were used to create the prediction model. Pearson’s correlation analysis and receiver operating characteristic curves were used to evaluate the predictive performance.ResultsThe radiomics score had a significant linear relationship with the luteinizing hormone concentration and the luteinizing hormone/follicle-stimulating hormone ratio. The radiomics score showed better predictive performance than traditional pituitary measurements. The area under the curve of the radiomics score, pituitary height, and variable combinations was 0.759 (95% confidence interval [CI], 0.583–0.936), 0.681 (95% CI, 0.483–0.878), and 0.829 (95% CI, 0.681–0.976), respectively.ConclusionCombination of the radiomics score with pituitary height measurements allows for better evaluation of the pituitary during diagnostic imaging, indicating satisfactory potential for efficacy assessments. |
| |
Keywords: | Radiomics texture features precocious puberty magnetic resonance imaging pituitary regression model |
|
|