Rationale and Objectives Artificial intelligence (AI) is reshaping the future of medicine, particularly influencing specialties like radiology. While the adoption of AI continues to accelerate, its integration presents both opportunities and challenges. To date, limited research has examined the perspectives of medical students and radiology trainees in less-developed regions of China—groups essential to the future healthcare workforce. This study aimed to assess their perceptions, attitudes, usage, and concerns regarding AI in clinical practice. Materials and Methods A cross-sectional, multicenter study was conducted between February and March 2025 across three provinces in central and western China. A total of 5043 medical undergraduates and 190 radiology trainees completed a self-designed questionnaire assessing their knowledge of AI, perceived utility, and awareness of its clinical applications. Results Most medical undergraduates reported limited exposure to formal AI training or hands-on experience yet expressed general support for its clinical use. Concerns regarding data privacy, transparency, and patient trust were commonly noted. Radiology trainees demonstrated higher levels of AI education and tool utilization. Both groups agreed AI could enhance efficiency without replacing physicians and emphasized the need to address technical, legal, and ethical challenges for successful implementation. Further analysis presented both grades and gender will influence participants’ attitudes toward AI. Conclusion This study highlights a generally positive attitude toward AI among future healthcare professionals, while revealing substantial educational gaps. Structured AI training should be integrated into undergraduate and radiology curricula to better prepare trainees for AI-assisted clinical environments.

To develop and validate a deep learning model based on three-dimensional features (DL_3D) for distinguishing lung adenocarcinoma (LUAD) from tuberculoma (TBM). A total of 1160 patients were collected from three hospitals. A vision transformer network-based DL_3D model was trained, and its performance in differentiating LUAD from TBM was evaluated using validation and external test sets. The performance of the DL_3D model was compared with that of two-dimensional features (DL_2D), radiomics, and six radiologists. Diagnostic performance was assessed using the area under the receiver operating characteristic curves (AUCs) analysis. The study included 840 patients in the training set (mean age, 54.8 years [range, 19–86 years]; 514 men), 210 patients in the validation set (mean age, 54.3 years [range, 18–86 years]; 128 men), and 110 patients in the external test set (mean age, 54.7 years [range, 22–88 years]; 51 men). In both the validation and external test sets, DL_3D exhibited excellent diagnostic performance (AUCs, 0.895 and 0.913, respectively). In the test set, the DL_3D model showed better performance (AUC, 0.913; 95

To analyze the correlation between the main perfusion parameters of the left ventricle and various physiological and coronary artery disease (CAD) risk factors or comorbidities using dynamic stress computed tomography myocardial perfusion imaging (CT-MPI) in patients without obstructive coronary stenosis. This retrospective analysis included 119 patients without obstructive coronary artery stenosis in computed tomography angiography (CTA), and without perfusion defects in CT-MPI. Patients were categorized into groups based on the presence or absence of physiological and CAD risk factors or comorbidities. The global myocardial blood flow (MBF), myocardial blood volume (MBV), and perfused capillary blood volume (PCBV) of the left ventricle were compared between groups, and correlations with continuous variables were analyzed. Multivariate linear regression was used to identify independent factors. Perfusion parameters were higher (MBF, 149.41 ± 26.38 vs. 159.20 ± 21.31 ml/100 ml/min, MBV, 17.09 ± 2.37 vs.18.84 ± 1.89, and PCBV, 9.82 ± 2.21 vs. 11.47 ± 1.79 ml/100 ml [all P < 0.05]) in female patients than in male patients. Hypertension and overweight/obesity resulted in lower perfusion parameters (hypertension vs. normotension: MBF, 148.09 ± 21.15 vs. 161.47 ± 25.13 ml/100 ml/min, PCBV, 10.25 ± 2.23 vs. 11.22 ± 1.96 ml/100 ml; overweight/obesity vs. none: MBF, 148.82 ± 20.98 vs. 159.51 ± 25.44 ml/100 ml/min, PCBV, 10.20 ± 1.93 vs. 11.15 ± 2.22 ml/100 ml [all P < 0.05]). Body surface area (BSA), body mass index, stress heart rate (HR), incremental HR, coronary total plaque volume, and stress systolic blood pressure were significantly correlated with perfusion parameters (all P < 0.05). Stress HR, BSA, and hypertension were independent predictors of MBF, stress HR and sex were independent predictors of MBV, and stress HR and BSA were independent predictors of PCBV. Dynamic stress CT-MPI myocardial perfusion is affected by stress HR, sex, and BSA, and can identify early perfusion distribution in hypertension and obesity/overweight.

To compare the diagnostic efficacy of different relative myocardial blood flow (MBF) ratios in computed tomography perfusion (CTP) for myocardial ischemia in patients with obstructive coronary artery disease (CAD). Between October 2020 and March 2024, patients with suspected or known obstructive CAD who underwent CTP + coronary computed tomography angiography and invasive coronary angiography/fractional flow reserve were retrospectively selected. Patients and vessels were categorized into ischemia and non-ischemia groups. The diagnostic efficacies of the three relative MBF ratios were compared in patients with obstructive CAD. This study included 48 patients (144 vessels). Notably, 34 of the 48 patients (70.83

Nuclear medicine is an interdisciplinary field that integrates basic science with clinical medicine. The traditional classroom teaching model lacks interactive and efficient teaching methods and does not adequately address the learning needs and educational goals associated with standardized training for residents. The teaching model that combines Small Private Online Courses (SPOCs) with a flipped classroom approach is more aligned with the demands of real-life scenarios and workplace requirements, thereby assisting students in developing comprehensive literacy and practical problem-solving skills. However, this innovative teaching model has yet to be implemented in Nuclear medicine courses. This study aimed to explore whether the post-training competence for residents can be improved based on this new teaching model. A total of 103 first-year residents from Sichuan Provincial People’s Hospital were randomly assigned to either an experimental group (n = 52) or a control group (n = 51) between July 2019 and June 2023. The experimental group utilized a SPOC and flipped classroom-blended teaching model, while the control group received traditional lecture-based learning (LBL). We assessed the theoretical evaluation scores and questionnaire responses from both groups to determine the effectiveness of the new pedagogical approach. Residents in the experimental group demonstrated a superior understanding of nuclear medicine content compared to those in the control group, achieving higher scores on pre-class assessments, after-class tests, and final exams (P < 0.01). A majority of the residents in the experimental group expressed that the innovative teaching model, which integrated SPOC and a flipped classroom approach, significantly enhanced their motivation and contributed to the development of their ‘professional skills,’ ‘patient care,’ ‘interaction and teamwork,’ ‘teaching proficiency,’ and ‘learning capabilities’. The teaching satisfaction survey indicated that the experimental group reported significantly higher levels of ‘overall satisfaction,’ as well as greater satisfaction with ' teaching methodologies ' and ' fulfillment of targeted clinical skills,’ compared to the control group (P < 0.01). The SPOC and flipped classroom teaching model is better than traditional LBL in enriching residents’ professional knowledge and cultivating their post-training competence. It can effectively promote educational quality, improve residents’ learning, and enhance their satisfaction.

IntroductionStress dynamic computed tomography myocardial perfusion imaging (CT-MPI) is an accurate quantitative method for diagnosing myocardial ischemia in coronary artery disease (CAD). However, its clinical application has been limited, partly due to the varied cutoff values for absolute myocardial blood flow (MBFa) and the uncertain value of the relative myocardial blood flow ratio (MBF-ratio). This study aimed to compare the diagnostic efficacy of and investigate the optimal cutoff values for MBFa and the MBF-ratio in CT-MPI for diagnosing myocardial ischemia in patients with hemodynamically significant CAD.MethodsPatients with suspected or known hemodynamically significant CAD who underwent CT-MPI + CT angiography and invasive coronary angiography (ICA)/fractional flow reserve (FFR) between October 2020 and December 2023 were retrospectively evaluated. ICA ≥80% or FFR ≤0.8 were set as the diagnostic standards for functional ischemia. The patients and vessels were categorized into ischemic and non-ischemic groups, and differences in MBFa and the MBF-ratio were compared between the groups. The area under the curve (AUC) and optimal cutoff values were calculated. Diagnostic efficacy parameters, such as sensitivity, specificity, and accuracy, were also compared. In addition, a consistency test was performed.ResultsA total of 46 patients (mean age: 65.37 ± 8.25 years; 120 vessels) were evaluated. Hemodynamically significant stenosis was detected in 30/46 patients (48%) and 81/120 vessels (67.5%). The MBFa and MBF-ratio values were significantly lower in the ischemic than in the non-ischemic group; in the per-vessel analysis, the MBFa values were 73 vs. 128 (P < 0.001) and the MBF-ratio values were 0.781 vs. 0.856 (P < 0.001), respectively. The optimal cutoff values for MBFa and the MBF-ratio were 117.71 and 0.67, respectively. MBFa demonstrated a sensitivity, specificity, accuracy, AUC, positive predictive value, negative predictive value, and kappa value of 97.44%, 74.07%, 81.66%, 0.936 [95% confidence interval (CI): 0.876–0.973, P < 0.001], 63.33%, 98.36%, and 0.631 (95% CI: 0.500–0.762), respectively. The corresponding values for the MBF-ratio were 92.31%, 85.19%, 87.5%, 0.962 (95% CI: 0.911–0.989, P < 0.001), 75%, 95.83%, and 0.731 (95% CI: 0.606–0.857, P < 0.001), with no significant difference (P = 0.1225).ConclusionBoth MBFa and the MBF-ratio exhibit excellent diagnostic performance for myocardial ischemia in patients with hemodynamically significant CAD. The MBF-ratio is more robust than MBFa for interpreting CT-MPI findings in clinical practice, which is useful for radiologists and clinicians implementing CT-MPI.

Background:Noninvasively detecting epidermal growth factor receptor (EGFR) mutation status in lung adenocarcinoma patients before targeted therapy remains a challenge. This study aimed to develop a 3-dimensional (3D) convolutional neural network (CNN)-based deep learning model to predict EGFR mutation status using computed tomography (CT) images. Methods:We retrospectively collected 660 patients from 2 large medical centers. The patients were divided into training (n=528) and external test (n=132) sets according to hospital source. The CNN model was trained in a supervised end-to-end manner, and its performance was evaluated using an external test set. To compare the performance of the CNN model, we constructed 1 clinical and 3 radiomics models. Furthermore, we constructed a comprehensive model combining the highest-performing radiomics and CNN models. The receiver operating characteristic (ROC) curves were used as primary measures of performance for each model. Delong test was used to compare performance differences between different models. Results:Compared with the clinical [training set, area under the curve (AUC) =69.6%, 95% confidence interval (CI), 0.661-0.732; test set, AUC =68.4%, 95% CI, 0.609-0.752] and the highest-performing radiomics models (training set, AUC =84.3%, 95% CI, 0.812-0.873; test set, AUC =72.4%, 95% CI, 0.653-0.794) models, the CNN model (training set, AUC =94.3%, 95% CI, 0.920-0.961; test set, AUC =94.7%, 95% CI, 0.894-0.978) had significantly better predictive performance for predicting EGFR mutation status. In addition, compared with the comprehensive model (training set, AUC =95.7%, 95% CI, 0.942-0.971; test set, AUC =87.4%, 95% CI, 0.820-0.924), the CNN model had better stability. Conclusions:The CNN model has excellent performance in non-invasively predicting EGFR mutation status in patients with lung adenocarcinoma and is expected to become an auxiliary tool for clinicians.

RATIONALE AND OBJECTIVES:This study aimed to non-invasively predict epidermal growth factor receptor (EGFR) mutation status in patients with lung adenocarcinoma using multi-phase computed tomography (CT) radiomics features. MATERIALS AND METHODS:A total of 424 patients with lung adenocarcinoma were recruited from two hospitals who underwent preoperative non-enhanced CT (NE-CT) and enhanced CT (including arterial phase CT [AP-CT], and venous phase CT [VP-CT]). Patients were divided into training (n = 297) and external validation (n = 127) cohorts according to hospital. Radiomics features were extracted from the NE-CT, AP-CT, and VP-CT images, respectively. The Wilcoxon test, correlation analysis, and simulated annealing were used for feature screening. A clinical model and eight radiomics models were established. Furthermore, a clinical-radiomics model was constructed by incorporating multi-phase CT features and clinical risk factors. Receiver operating characteristic curves were used to evaluate the predictive performance of the models. RESULTS:The predictive performance of multi-phase CT radiomics model (AUC of 0.925 [95% CI, 0.879-0.971] in the validation cohort) was higher than that of NE-CT, AP-CT, VP-CT, and clinical models (AUCs of 0.860 [95% CI,0.794-0.927], 0.792 [95% CI, 0.713-0.871], 0.753 [95% CI, 0.669-0.838], and 0.706 [95% CI, 0.620-0.791] in the validation cohort, respectively) (all P < 0.05). The predictive performance of the clinical-radiomics model (AUC of 0.927 [95% CI, 0.882-0.971] in the validation cohort) was comparable to that of multi-phase CT radiomics model (P > 0.05). CONCLUSION:Our multi-phase CT radiomics model showed good performance in identifying the EGFR mutation status in patients with lung adenocarcinoma, which may assist personalized treatment decisions.

The aim of this study was to explore the related factors linked to the development and infectivity of tuberculosis. This was achieved by comparing the clinical characteristics of patients with pulmonary tuberculosis (TB) who tested positive in smear Mycobacterium tuberculosis tests with this who tested negative in smear mycobacterium tests but positive in sputum Gene Xpert tests. We gathered clinical data of 1612 recently hospitalized patients diagnosed with pulmonary tuberculosis who tested positive either in sputum Gene-Xpert test or sputum smear Mycobacterium tuberculosis tests. The data was collected from January 1, 2018 to August 5, 2023, at Sichuan Provincial People's Hospital. We conducted separately analyzes and comparisons of the clinical characteristics between the two groups of patients, aiming to discussed the related factors influencing the development and infectivity of tuberculosis. In comparison to the GeneXpert positive group, the sputum smear positive group exhibited a higher proportion of elderly patients (aged 75-89) and individuals classified as underweight (BMI < 18.5 kg/m(2)). Furthermore, this group was more prone to experiencing symptoms such as weight loss, coughing and sputum production, hemoptysis, shortness of breath, and difficulty breathing. Moreover, they are also more likely to develop extrapulmonary tuberculosis, such as tuberculous meningitis, tuberculous pleurisy, and tuberculous peritonitis. These clinical features, when present, not only increase the likelihood of a positive result in sputum smear tests but also suggest a high infectivity of pulmonary tuberculosis. Elderly individuals (aged 75 to 89) who are underweight (BMI < 18.5 kg/m(2)), display symptom of cough, expectoration, hemoptysis and dyspnea-particularly cough and expectoration-and those with extra pulmonary tuberculosis serve as indicators of highly infectious pulmonary tuberculosis patients. These patients may present with more severe condition, carrying a higher bacteria, and being more prone to bacterial elimination. Identification of these patients is crucial, and prompt actions such as timely and rapid isolation measures, cutting off transmission routes, and early empirical treatment of tuberculosis are essential to control the development of the disease.

Background This study aimed to investigate the relationship between active pulmonary tuberculosis (TB) and type 2 diabetes mellitus (T2DM) by analysing the clinical features and computed tomography (CT) findings of patients with active pulmonary TB and comorbid T2DM (TB-DM) in the LiangShan Yi regions. Methods We collected data from 154 hospitalised patients with TB-DM initially confirmed at an infectious disease hospital in the Liangshan Yi Autonomous Prefecture between 1 and 2019, and 31 December 2021. These were matched by sex and age ± 3 years to 145 hospitalised patients with initially confirmed pulmonary TB without comorbid T2DM (TB-NDM) over the same period. The clinical characteristics of the two groups were analysed separately. Three group-blinded radiologists independently analysed the CT findings and classified them into mild-to-moderate and severe groups. Severe chest CT lesion refers to a lesion that is less diffused or moderately dense and either exceeds the total volume of one lung, a high-density fused lesion greater than one-third of the volume of one lung, or a cavitary lesion with a maximum diameter ≥ 4 cm. Results No significant differences were observed in the presentation of clinical features. Regarding the severity of chest CT manifestation, patients with TB-DM had significantly more severe TB than those with TB-NDM (89.61% vs. 68.97%, P < 0.0001). Regarding CT findings, patients with TB-DM had higher proportions of consolidation (79.22% vs. 52.41%, P < 0.0001), cavitary lesions (85.06% vs. 59.31%, P < 0.0001), bronchiectasis (71.43% vs. 31.03%, P < 0.0001), exudative lesions (88.96% vs. 68.28%, P < 0.0001), and fibrous lesions (93.51% vs. 68.97%, P < 0.0001) than patients with TB-NDM. In conclusion, patients with TB-DM have more severe pulmonary TB CT findings than those without. There were no significant differences in the distribution of lesions in the lung lobes between TB-DM and TB-NDM patients. Conclusions Among patients hospitalised with pulmonary TB, those with T2DM had more severe findings on chest CT than those without T2DM. However, the clinical presentation was not significantly different.

Background Tuberculosis (TB) has a high morbidity and mortality rate, and its prevention and treatment focus is on impoverished areas. The Liangshan Yi Autonomous Prefecture is a typical impoverished area in western China with insufficient medical resources and high HIV positivity. However, there have been few reports of TB and drug resistance in this area. Methods We collected the demographic and clinical data of inpatients with sputum smear positive TB between 2015 and 2021 in an infectious disease hospital in the Liangshan Yi Autonomous Prefecture. Descriptive analyses were used for the epidemiological data. The chi-square test was used to compare categorical variables between the drug-resistant and drug-susceptible groups, and binary logistic regression was used to analyse meaningful variables. Results We included 2263 patients, 79.9% of whom were Yi patients. The proportions of HIV (14.4%) and smoking (37.3%) were higher than previously reported. The incidence of extrapulmonary TB (28.5%) was high, and the infection site was different from that reported previously. When drug resistance gene detection was introduced, the proportion of drug-resistant patients became 10.9%. Patients aged 15–44 years (OR 1.817; 95% CI 1.162–2.840; P < 0.01) and 45–59 years (OR 2.175; 95% CI 1.335–3.543; P < 0.01) had significantly higher incidences of drug resistance than children and the elderly. Patients with a cough of ≥ 2 weeks had a significantly higher chance of drug resistance than those with < 2 weeks or no cough symptoms (OR 2.069; 95% CI 1.234–3.469; P < 0.01). Alcoholism (OR 1.741; 95% CI 1.107–2.736; P < 0.05) and high bacterial counts on sputum acid-fast smears (OR 1.846; 95% CI 1.115–3.058; P < 0.05) were significant in the univariate analysis. Conclusions Sputum smear-positive TB predominated in Yi men (15–44 years) with high smoking, alcoholism, and HIV rates. Extrapulmonary TB, especially abdominal TB, prevailed. Recent drug resistance testing revealed higher rates in 15–59 age group and ≥ 2 weeks cough duration. Alcohol abuse and high sputum AFB counts correlated with drug resistance. Strengthen screening and supervision to curb TB transmission and drug-resistant cases in the region.

Purpose:This study aimed to investigate clinical features and computed tomography (CT) manifestations of rifampicin primary drug-resistant pulmonary tuberculosis in Liangshan Yi Autonomous Prefecture.Patients and Methods:A total of 100 inpatients with confirmed primary rifampicin-resistant pulmonary tuberculosis were recruited from January 2020 to December 2022 at an infectious disease hospital located in the Liangshan Yi Autonomous Prefecture. Additionally, 100 inpatients with confirmed drug-susceptible pulmonary tuberculosis during the same period were matched to the rifampicin-resistant group based on gender, age, and ethnicity. The clinical characteristics of the two groups were recorded separately. Furthermore, the CT manifestations in these patients were independently analyzed by three radiologists.Results:The results showed that comorbid diabetes mellitus was more prevalent in the drug-resistant tuberculosis (DR-TB) group than in the drug-susceptible tuberculosis (DS-TB) group (9% vs 0%, p=0.0032). In terms of imaging presentation, DR-TB patients exhibited a higher frequency of calcifications (55% vs 35.00%, p=0.0068), greater median number of cavities (5 vs 2, p=0.0027), and larger maximum cavity diameter (52.08±25.55 mm vs 42.72±17.48 mm, p=0.0097). Additionally, bilateral involvement was more common in DR-TB patients at the site of the lesion (89% vs 76%, p=0.0246), with a higher prevalence in the right middle (82% vs 68%, p=0.0332), right lower (82% vs 68%, p=0.0332), left upper (91% vs 77%, p=0.0113), and left lower lobes (92% vs 66%, p<0.0001). Conversely, the involvement of only one lobe was less frequent in patients with DR-TB than in those with DS-TB (4% vs 13%, p=0.0398), whereas the involvement of all five lobes was more common (68% vs 51%, p=0.0209).Conclusion:Patients with DR-TB exhibit a higher prevalence of severe imaging manifestations, highlighting the importance of CT in the early detection and diagnosis of DR-TB.

Background This study aimed to investigate the diagnostic value of machine-learning (ML) models with multiple classifiers based on non-enhanced CT Radiomics features for differentiating anterior mediastinal cysts (AMCs) from thymomas, and high-risk from low risk thymomas. Methods In total, 201 patients with AMCs and thymomas from three centers were included and divided into two groups: AMCs vs. thymomas, and high-risk vs low-risk thymomas. A radiomics model (RM) was built with 73 radiomics features that were extracted from the three-dimensional images of each patient. A combined model (CM) was built with clinical features and subjective CT finding features combined with radiomics features. For the RM and CM in each group, five selection methods were adopted to select suitable features for the classifier, and seven ML classifiers were employed to build discriminative models. Receiver operating characteristic (ROC) curves were used to evaluate the diagnostic performance of each combination. Results Several classifiers combined with suitable selection methods demonstrated good diagnostic performance with areas under the curves (AUCs) of 0.876 and 0.922 for the RM and CM in group 1 and 0.747 and 0.783 for the RM and CM in group 2, respectively. The combination of support vector machine (SVM) as the feature-selection method and Gradient Boosting Decision Tree (GBDT) as the classification algorithm represented the best comprehensive discriminative ability in both group. Comparatively, assessments by radiologists achieved a middle AUCs of 0.656 and 0.626 in the two groups, which were lower than the AUCs of the RM and CM. Most CMs exhibited higher AUC value compared to RMs in both groups, among them only a few CMs demonstrated better performance with significant difference in group 1. Conclusion Our ML models demonstrated good performance for differentiation of AMCs from thymomas and low-risk from high-risk thymomas. ML based on non-enhanced CT radiomics may serve as a novel preoperative tool.

Objective Fluorodeoxyglucose Positron emission tomography/computerized tomography (FDG PET/CT) has become popular for diagnosing periprosthetic joint infections (PJI). However, the diagnostic accuracy for this technique has varied from report to report. This meta-analysis was performed to assess the accuracy of FDG PET/CT for PJI diagnosis. Material and Methods We conducted a systematic search of online academic databases for all studies reporting the diagnostic accuracy of FDG PET/CT for PJI. Meta-analysis was performed using STATA software. Results 23 studies, containing data on 1,437 patients, met inclusion criteria. Pooled sensitivity and specificity of FDG PET/CT for diagnosing PJI were 85% (95% CI, 76%, 91%) and 86% (95% CI, 78%, 91%), respectively with an AUC of 0.92. LRP was 6.1 (95% CI, 3.8, 9.7) and LRN was 0.17 (0.11, 0.28), indicating that FDG PET/CT cannot be used for confirmation or exclusion of PJI. There was significant inter-study heterogeneity, but no significant publication bias was noted. Conclusions Our study found that FDG PET/CT has an important role as a diagnostic tool for PJI with high sensitivity and specificity. Further studies exploring its accuracy in different PJI locations remain necessary.

Purpose: The aim of this study was to investigate the diagnostic value of machine-learning models with radiomic features and clinical features in preoperative differentiation of common lesions located in the anterior skull base. Methods: A total of 235 patients diagnosed with pituitary adenoma, meningioma, craniopharyngioma, or Rathke cleft cyst were enrolled in the current study. The discrimination was divided into three groups: pituitary adenoma vs. craniopharyngioma, meningioma vs. craniopharyngioma, and pituitary adenoma vs. Rathke cleft cyst. In each group, five selection methods were adopted to select suitable features for the classifier, and nine machine-learning classifiers were employed to build discriminative models. The diagnostic performance of each combination was evaluated with area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity calculated for both the training group and the testing group. Results: In each group, several classifiers combined with suitable selection methods represented feasible diagnostic performance with AUC of more than 0.80. Moreover, the combination of least absolute shrinkage and selection operator as the feature-selection method and linear discriminant analysis as the classification algorithm represented the best comprehensive discriminative ability. Conclusion: Radiomics-based machine learning could potentially serve as a novel method to assist in discriminating common lesions in the anterior skull base prior to operation.