Abstract 2124
Background
To explore the differences of MRI scan between pituitary adenomas and craniopharyngioma from MR image features to 3D based texture features.
Methods
A total number of 131 patients were introduced into this study (pituitary Adenomas =68; craniopharyngioma=63) with pre-surgery MRI image. Qualitative MR features and MRI texture features of lesion were evaluated using chi-square tests, Fisher exact test or Mann-Whitney U test. Multivariate logistic regression analyses were performed to access their ability as independent predictors. Accuracy measures were calculated substantially for the significant features.
Results
Five MRI features were suggested to be significantly different between pituitary adenomas and craniopharyngioma and one of them, cystic alteration could be considered as independent practical predictors. Three texture features from contrast enhanced images (Histo-Skewness, GLCM-Contrast and GLCM-Energy), two texture features from T2WI (Histo-Skewness and GLCM-Contrast) were found to be significantly related to discrimination between two types of diseases. Two texture features (Histo-Skewness and GLCM-Contrast) were found significantly to be in relation with cystic alteration.
Conclusions
Both of MRI image features and texture features, could make significant discrimination between pituitary adenoma and craniopharyngioma and represent practical diagnostic value; in addition, two type features associate with each other.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Xuelei Ma.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
4732 - Progesterone Receptor Isoform Ratio Dictates Antiprogestins/Progestins Effects on Metastatic Breast Cancer Models
Presenter: Maria Abascal
Session: Poster Display session 2
Resources:
Abstract
5737 - PAM50 and CGH-array genomic characterization of HER2-Equivocal Breast Cancers defined by the 2018 ASCO/CAP recommendations.
Presenter: Carine Ngo
Session: Poster Display session 2
Resources:
Abstract
1096 - OncotypeDX® predictive nomogram for recurrence score output: a machine learning system based on quantitative immunochemistry analysis - ADAPTED01
Presenter: Fabio Marazzi
Session: Poster Display session 2
Resources:
Abstract
5426 - Geriatric parameters predict both disease-related and patient-reported outcomes in older patients with breast cancer
Presenter: Willeke van der Plas-Krijgsman
Session: Poster Display session 2
Resources:
Abstract
5865 - Patients with a 21-gene assay in South East London differ from the TAILORx trial population
Presenter: Charalampos Gousis
Session: Poster Display session 2
Resources:
Abstract
1312 - Predictive tools in adjuvant breast cancer – what is the standard of evidence supporting their utility? A literature review examining validation of Adjuvant!, Cancermath and NHS Predict
Presenter: Alice Loft
Session: Poster Display session 2
Resources:
Abstract
2445 - Oncologic outcome of invasive lobular carcinoma: Is it different from that of invasive ductal carcinoma?
Presenter: Hee Jun Choi
Session: Poster Display session 2
Resources:
Abstract
2476 - Pathologic response and survival efficacy in patients with initial nodal involvement after neoadjuvant chemotherapy in early breast cancer
Presenter: SERAFIN MORALES Murillo
Session: Poster Display session 2
Resources:
Abstract
3761 - Chemotherapy-induced amenorrhea: prognostic impact on premenopausal Egyptian patients with breast cancer
Presenter: Khaled Abdel Karim
Session: Poster Display session 2
Resources:
Abstract
4687 - Predicting the presence of breast cancer using circulating small RNA in the serum
Presenter: Yumiko Koi
Session: Poster Display session 2
Resources:
Abstract