ABSTRACT
Problem: Chest radiography is a crucial tool for diagnosing thoracic disorders, but interpretation errors and a lack of qualified practitioners can cause delays in treatment. Aim: This study aimed to develop a reliable multi-classification artificial intelligence (AI) tool to improve the accuracy and efficiency of chest radiograph diagnosis. Methods: We developed a convolutional neural network (CNN) capable of distinguishing among 26 thoracic diagnoses. The model was trained and externally validated using 795,055 chest radiographs from 13 datasets across 4 countries. Results: The CNN model achieved an average area under the curve (AUC) of 0.961 across all 26 diagnoses in the testing set. COVID-19 detection achieved perfect accuracy (AUC 1.000, [95% confidence interval {CI}, 1.000 to 1.000]), while effusion or pleural effusion detection showed the lowest accuracy (AUC 0.8453, [95% CI, 0.8417 to 0.8489]). In external validation, the model demonstrated strong reproducibility and generalizability within the local dataset, achieving an AUC of 0.9634 for lung opacity detection (95% CI, 0.9423 to 0.9702). The CNN outperformed both radiologists and nonradiological physicians, particularly in trans-device image recognition. Even for diseases not specifically trained on, such as aortic dissection, the AI model showed considerable scalability and enhanced diagnostic accuracy for physicians of varying experience levels (all P < 0.05). Additionally, our model exhibited no gender bias (P > 0.05). Conclusion: The developed AI algorithm, now available as professional web-based software, substantively improves chest radiograph interpretation. This research advances medical imaging and offers substantial diagnostic support in clinical settings.
ABSTRACT
Background: To determine the effectiveness and safety of different patch materials in the treatment of pediatric patients with congenital supravalvular aortic stenosis (SVAS). Methods: 218 consecutive SVAS patients (age < 14 years) who underwent surgery from Beijing Fuwai and Yunnan Fuwai hospital between 2002 and 2020 were included. Patients were divided into the pericardium patch group (133 (61.0%)), modified patch group (43 (19.7%)) and artificial patch group (42 (19.3%)). The primary safety endpoint was patch-related adverse complications (post-operation patch hemorrhage or aortic sinus aneurysm at 2-year follow-up). The primary effectiveness outcome was the re-operation or restenosis at 2-year follow-up. Multivariable cox regression was used to obtain the hazard ratio (HR). Results: The median age at operation was 43.5 months (IQR 24.0-73.0). Only three patients had patch-related adverse complications, and no difference existed among the three groups (p = 0.763). After a median follow-up of 24.0 months (IQR 6.0-48.0), patients with a pericardium patch had a lower re-operation or restenosis rate compared with the other two groups (pericardium patch vs modified patch, HR = 0.30, 95% CI 0.12-0.77; pericardium patch vs artificial patch, HR = 0.33, 95% CI 0.13-0.82), even in the main subgroup and sensitivity analysis. Conclusions: In pediatric patients, the safety of autologous pericardium patch is acceptable, along with lower rates of middle-term re-operation or restenosis. Clinical Trial Registration: http://www.chictr.org.cn, number: ChiCTR2300067851.
ABSTRACT
OBJECTIVES: To investigate the short-term/long-term impact of pulmonary artery intramural hematoma (PA-IMH) in patients with acute Stanford type A aortic dissection (ATAAD) following surgical repair. MATERIALS AND METHODS: Consecutive patients with ATAAD who received surgical repair at Beijing and Yunnan Fuwai Hospital in 2010-2021 were retrospectively reviewed. Patients with hemorrhage extending along the PA were identified as the PA-IMH group. Multivariable logistics regression was used to obtain the odds ratio (OR), and the Kaplan-Meier method was used to estimate the survival rate. RESULTS: Of the 2046 ATAAD patients, 324 (15.8%) patients were identified with PA-IMH, and 1722 (84.2%) were without PA-IMH. PA-IMH had a higher prevalence in patients with older age, female gender, aortic IMH, and type II aortic dissection. PA-IMH patients incurred excess early mortality compared with non-PA-IMH patients (9.3% vs. 5.6%, OR = 1.86, 95%CI 1.19-2.91, p = 0.006). The results were stable in the subgroup analysis, with an increased risk in older (> 70 years) or DeBakey type II ATAAD patients. Notably, an increase in the degree and extent of PA-IMH exacerbated the risk of early mortality. However, after landmark analysis at 30-day postsurgery, no significant difference was noted in the long-term outcomes between PA-IMH and non-PA-IMH groups (p = 0.440). The 5-year survival rates were 87.1% (95%CI: 83.3%, 91.1%) and 90.1% (95%CI: 88.5%, 91.7%), respectively. CONCLUSIONS: The presence of PA-IMH in ATAAD patients is common and is independently associated with increased early mortality after surgical repair, especially in those with older age (> 70) or type II dissection. However, such detrimental effects do not persist in the long-term follow-up among patients who survived hospital discharge. CLINICAL RELEVANCE STATEMENT: We confirmed that PA-IMH significantly increases early postoperative mortality in patients with acute type A aortic dissection, especially in older patients or DeBakey type II dissection. This should prompt further investigation of the incremental role of PA-IMH in this pathology. KEY POINTS: Acute type A aortic dissection mortality gets worse when pulmonary artery intramural hematoma is present. Pulmonary artery-intramural hematoma increased the risk of early mortality but not affect long-term prognosis. Further research should investigate the effects of pulmonary artery intramural thrombus on aortic dissection.
ABSTRACT
BACKGROUND: Quadricuspid aortic valve (QAV) is a rare congenital heart disease with a limited body of literature. This retrospective cohort study investigates QAV morphology, function, and clinical outcomes. METHODS: Echocardiography was used to assess valvular function. Morphological characteristics such as phenotypes, raphe, regurgitant orifice area (ROA), and aortic dilation (diameter >40 âmm) were assessed by cardiac CT. Patients were followed up for the combined event of all-cause death and aortic valve replacement (AVR). RESULTS: Ninety QAV patients (screened from 322385 CT scans) were included (mean age 55.2 â± â13.6 years, 61.1 â% male). Isolated significant aortic regurgitation (AR) was present in 75.6 â% of patients. The cohort was dominated by type I (four equal leaflets, 37.8 â%) and type II (3 larger and 1 smaller leaflets, 42.2 â%) QAV. Fused raphe was present in 26.7 â% of patients. ROACT was correlated with AR severity and aortic dilation (41.1 â%, n â= â37). Among patients without AVR at baseline (n â= â60), one died and 17 underwent AVR during a median follow-up of 35.0 months (IQR:17.3-62.8). ROACT was associated with an increasing risk of combined event (as a categorical variable with a cut-off of 21.4 âmm2, HR â= â4.25, 95%CI 1.49-12.17, p â= â0.007; as a continuous variable (per mm2 increment), HR â= â1.04, 95%CI 1.01-1.07, p â= â0.003). Additionally, ROACT had incremental prognostic value when added to the AR severity model (area under the receiver-operating characteristic curve increased from 86.8 to 88.4, p â= â0.004). CONCLUSION: QAV is characterized by variable anatomy, progressive AR, concomitant cusp fusion and aortic enlargement. ROACT may be a potential ancillary prognostic marker in patients with QAV.