Objective To construct machine learning models based on preoperative inflammatory and radiological features for the prediction of glioma grading and isocitrate dehydrogenase (IDH) mutation status, and to analyze application values of these models and identify the optimal predictive models. Methods A retrospective analysis was conducted on the data of pathologically confirmed glioma patients admitted to Tongji Hospital Affiliated to Tongji University from March 2019 to March 2023. LASSO regression was used to screen feature variables, and predictive models were constructed based on logistic regression (LR), random forest (RF), support vector machine (SVM), gradient boosting decision tree (XGBoost) and K-nearest neighbor (KNN) algorithms. The model performance was comprehensively evaluated using metrics including discrimination ability, area under the precision-recall curve (AUC), accuracy, F1 score and Brier score. The DeLong test was adopted to compare the AUC values among different models; Friedman rank-sum test was used to determine the overall performance differences of the models, with the Nemenyi test applied for multiple comparison correction. Results In the task of glioma grading prediction, the LR model achieved the highest comprehensive score (0.726), and no significant difference was observed between the LR model and the other four models; age was positively correlated with glioma grading (P=0.003). In the task of IDH mutation status prediction, the XGBoost model obtained the highest comprehensive score (0.832), which was superior to the LR (0.762, P=0.035) and KNN models (0.754, P=0.025), while no statistical differences were found between the XGBoost model and the RF or SVM models. Conclusions The LR model for glioma grading prediction and XGBoost model for IDH mutation prediction constructed based on a task-oriented strategy achieve a favorable interpretability while ensuring optimized performance, thereby providing reliable decision support for the individualized diagnosis and treatment of glioma.

Objective To compare the epidemiological characteristics of COVID-19 in Guangzhou and Wenzhou, and evaluate the effectiveness of their prevention and control measures. Methods Data of COVID-19 cases reported in Guangzhou and Wenzhou as of 29 February, 2020 were collected. The incidence curves of COVID-19 in two cities were constructed. The real time reproduction number ( R t ) of COVID-19 in two cities was calculated respectively. Results A total of 346 and 465 confirmed COVID-19 cases were analysed in Guangzhou and Wenzhou, respectively. In two cities, most cases were aged 30-59 years (Guangzhou: 54.9%; Wenzhou: 70.3%). The incidence curve peaked on 27 January, 2020 in Guangzhou and on 26 January, 2020 in Wenzhou, then began to decline in both cities. The peaks of imported COVID-19 cases from Hubei occurred earlier than the peak of COVID-19 incidences in two cities, and the peak of imported cases from Hubei occurred earlier in Wenzhou than in Guangzhou. In early epidemic phase, imported cases were predominant in both cities, then the number of local cases increased and gradually took the dominance in Wenzhou. In Guangzhou, the imported cases was still predominant. Despite the different epidemic pattern, the R t and the number of COVID-19 cases declined after strict prevention and control measures were taken in Guangzhou and in Wenzhou. Conclusion The time and scale specific differences of imported COVID-19 resulted in different epidemic patterns in two cities, but the spread of the disease were effectively controlled after taking strict prevention and control measures.

Objective To explore the imaging biomarker for early diagnosis and disease course monitoring of Parkinson disease (PD) in arterial spin labeling (ASL) technique. Methods Between July, 2014 and May, 2017, 23 patients with PD underwent magnetic resonance imaging (MRI) and ASL examinations in our hospital, including 13 in the early stage and 10 in advanced stages. Voxel-based analysis (VBA) was used to observe the regional cerebral blood flow (rCBF) characteristics in PD patients in different stages and three-dimensional continuous arterial spin labeling (3D CASL) was used to analyze the mean cerebral blood flow (mCBF). Results No significant difference was found in mCBF among PD patients in the early stage, patients in advanced stages and normal control subjects (P=0.30). Compared with the normal control group, the patients with early-stage PD had decreased rCBF in resting state mainly in the right superior occipital gyrus and the right superior frontal gyrus as revealed by VBA (P<0.001);the patients with advanced PD showed decreased rCBF mainly in the left precentral gyrus and the postcentral gyrus (P<0.001). The patients with advanced PD exhibited lowered rCBF in the right substantia nigra and the bilateral corpus callosum as compared with the early-stage patients (P<0.001). Conclusions VBA of ASL reveals rCBF alterations in association with the disease progression in PD patients, suggesting that this technique might provide assistance in identification of potential markers for early PD diagnosis and for monitoring the disease course.

Objective To explore the imaging biomarker for early diagnosis and disease course monitoring of Parkinson disease (PD) in arterial spin labeling (ASL) technique. Methods Between July, 2014 and May, 2017, 23 patients with PD underwent magnetic resonance imaging (MRI) and ASL examinations in our hospital, including 13 in the early stage and 10 in advanced stages. Voxel-based analysis (VBA) was used to observe the regional cerebral blood flow (rCBF) characteristics in PD patients in different stages and three-dimensional continuous arterial spin labeling (3D CASL) was used to analyze the mean cerebral blood flow (mCBF). Results No significant difference was found in mCBF among PD patients in the early stage, patients in advanced stages and normal control subjects (P=0.30). Compared with the normal control group, the patients with early-stage PD had decreased rCBF in resting state mainly in the right superior occipital gyrus and the right superior frontal gyrus as revealed by VBA (P<0.001);the patients with advanced PD showed decreased rCBF mainly in the left precentral gyrus and the postcentral gyrus (P<0.001). The patients with advanced PD exhibited lowered rCBF in the right substantia nigra and the bilateral corpus callosum as compared with the early-stage patients (P<0.001). Conclusions VBA of ASL reveals rCBF alterations in association with the disease progression in PD patients, suggesting that this technique might provide assistance in identification of potential markers for early PD diagnosis and for monitoring the disease course.

Objective To compare and analyze the diagnostic value of prenatal ultrasound and MRI in fetal thoracic abnormalities, as well as the advantages and disadvantages, respectively, and to explore the clinical value of the combined use in diagnosing fetal thoracic abnormalities.Methods The prenatal ultrasound and MRI images of total 94 cases with thoracic abnormalities were analyzed retrospectively.All the patients received MRI exams within 2 days after the preliminary ultrasound diagnosis of abnormalities.All cases were confirmed by autopsy or postnatal follow-ups.Results 94 cases of thoracic abnormalities included 48 cases of cystic adenoma abnormalities, 33 cases of bronchopulmonary sequestration, 10 cases of diaphragmatic hernia and 3 cases of primary pulmonary hypoplasia.The diagnosis coincidence rate of ultrasound was 82.98% (78/94), while the combined use was 93.62% (88/94).There were statistical differences between the combined use and single ultrasound examination in detecting fetal thoracic abnormalities(P<0.05).Conclusion Both ultrasound and MRI could diagnose fetal thoracic abnormalities well and had its own advantages and disadvantages.The combined use of ultrasound and MRI could improve the sensitivity and specificity of prenatal diagnostic accuracy and have a better advantage in diagnosing fetal thoracic abnormalities.