Publications by authors named "Runrong Chen"
End-to-end deep learning for the diagnosis of pelvic and sacral tumors using non-enhanced MRI: a multi-center study.
NPJ Precis Oncol
August 2025
This study developed an end-to-end deep learning (DL) model using non-enhanced MRI to diagnose benign and malignant pelvic and sacral tumors (PSTs). Retrospective data from 835 patients across four hospitals were employed to train, validate, and test the models. Six diagnostic models with varied input sources were compared.
View Article and Find Full Text PDFHepatic fat infiltration and clinical indicators in polycystic ovary syndrome: insights from MRI HISTO sequences.
Reprod Biomed Online
September 2025
Research Questions: Is the 3.0 Tesla magnetic resonance imaging (MRI) high-resolution tissue spectroscopy (HISTO) sequence efficacious as a non-invasive method for assessing metabolic-associated fatty liver disease (MAFLD) in patients with polycystic ovary syndrome (PCOS)? Are there associations between the HISTO-proton density fat fraction (PDFF) value and liver fat infiltration, liver fat content, clinical indicators and body composition?
Design: MRI HISTO sequences of the liver were acquired in 28 patients with PCOS and 37 age-matched healthy controls. Fasting blood chemistry including liver function tests, lipid profile, glucose and insulin, morphometric characteristics, and body composition analysis were recorded.
MRI radiomics model for predicting TERT promoter mutation status in glioblastoma.
Brain Behav
December 2023
Article Synopsis
- The study investigates the predictive capability of MRI-based models for TERT promoter mutations in glioblastoma patients, aiming to enhance prognosis and treatment strategies.
- Using a dataset of 143 patients, the researchers evaluated 2553 features through various classification algorithms and found that a model utilizing recursive feature elimination and linear discriminant analysis performed best.
- Ultimately, the findings suggest that the radiomics model centered on ADC entropy is effective in predicting TERT mutations, which are linked to poorer patient outcomes.
Article Synopsis
- The study aimed to explore the effectiveness of a multi-parameter MRI-based radiomics nomogram in predicting TERT promoter mutations and prognosis in glioblastoma patients.
- A cohort of 152 GBM patients was analyzed, from which over 2,800 radiomics features were extracted to create a predictive nomogram that combined radiomics and clinical data.
- The results showed that the random forest algorithm had the highest diagnostic accuracy for TERT mutation identification, and the nomogram demonstrated strong predictive power and was validated as a useful tool for assessing patient risk and prognosis.