ObjectiveTo identify the correlation between finger-to-floor distance(FFD) and the spinal function indices and disease activity scores of ankylosing spondylitis (AS) via a multicentre case–control study, and to calculate the optimal cutoff value of FFD using statistical methods.MethodsPatients with AS and healthy individuals were recruited, and the FFD and other spinal mobility values were measured. The correlation between the FFD and the Bath Ankylosing Spondylitis Metric Index (BASMI), Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Bath Ankylosing Spondylitis Functional Index (BASFI) was analyzed using Spearman rank correlation analysis. Receiver operating characteristic (ROC) curves of FFD stratified by gender and age were drawn and their optimal cutoff values were determined.ResultsA total of 246 patients with AS and 246 healthy subjects were recruited. The FFD was strongly correlated with BASMI (r = 0.72, p < 0.001), moderately correlated with BASFI (r = 0.50, p < 0.001) and weakly correlated with BASDAI (r = 0.36, p < 0.001). The lowest cutoff value of the FFD was 2.6 cm while the highest was 18.4 cm. Moreover, the FFD was significantly correlated with sex and age.ConclusionThere exists a strong correlation between the FFD and spinal mobility, a moderately correlation and function, which provides reliable data for the evaluation of patients with AS in clinical settings and the rapid screening of low back pain-related diseases in the general population. Furthermore, these findings have clinical potential in improving the missed diagnosis or delayed diagnosis of low back pain.

Adenovirus pneumonia is common in pediatric upper respiratory tract infection, which is comparatively easy to develop into severe cases and has a high mortality rate with many influential sequelae. As for pathogenesis, adenoviruses can directly damage target cells and activate the immune response to varying degrees. Early clinical recognition depends on patients’ symptoms and laboratory tests, including those under 2 years old, dyspnea with systemic toxic symptoms, atelectasis or emphysema in CT image, decreased leukocytes, and significantly increased C-reaction protein (CRP) and procalcitonin (PCT), indicating the possibility of severe cases. Until now, there is no specific drug for adenovirus pneumonia, so in clinical practice, current treatment comprises antiviral drugs, respiratory support and bronchoscopy, immunomodulatory therapy, and blood purification. Additionally, post-infectious bronchiolitis obliterans (PIBO), hemophagocytic syndrome, and death should be carefully noted. Independent risk factors associated with the development of PIBO are invasive mechanical ventilation, intravenous steroid use, duration of fever, and male gender. Meanwhile, hypoxemia, hypercapnia, invasive mechanical ventilation, and low serum albumin levels are related to death. Among these, viral load and serological identification are not only “gold standard” for adenovirus pneumonia, but are also related to the severity and prognosis. Here, we discuss the progress of pathogenesis, early recognition, therapy, and risk factors for poor outcomes regarding severe pediatric adenovirus pneumonia.

BackgroundThe implementation of deep learning models for medical image classification poses significant challenges, including gradual performance degradation and limited adaptability to new diseases. However, frequent retraining of models is unfeasible and raises concerns about healthcare privacy due to the retention of prior patient data. To address these issues, this study investigated privacy-preserving continual learning methods as an alternative solution.MethodsWe evaluated twelve privacy-preserving non-storage continual learning algorithms based deep learning models for classifying retinal diseases from public optical coherence tomography (OCT) images, in a class-incremental learning scenario. The OCT dataset comprises 108,309 OCT images. Its classes include normal (47.21%), drusen (7.96%), choroidal neovascularization (CNV) (34.35%), and diabetic macular edema (DME) (10.48%). Each class consisted of 250 testing images. For continuous training, the first task involved CNV and normal classes, the second task focused on DME class, and the third task included drusen class. All selected algorithms were further experimented with different training sequence combinations. The final model's average class accuracy was measured. The performance of the joint model obtained through retraining and the original finetune model without continual learning algorithms were compared. Additionally, a publicly available medical dataset for colon cancer detection based on histology slides was selected as a proof of concept, while the CIFAR10 dataset was included as the continual learning benchmark.ResultsAmong the continual learning algorithms, Brain-inspired-replay (BIR) outperformed the others in the continual learning-based classification of retinal diseases from OCT images, achieving an accuracy of 62.00% (95% confidence interval: 59.36-64.64%), with consistent top performance observed in different training sequences. For colon cancer histology classification, Efficient Feature Transformations (EFT) attained the highest accuracy of 66.82% (95% confidence interval: 64.23-69.42%). In comparison, the joint model achieved accuracies of 90.76% and 89.28%, respectively. The finetune model demonstrated catastrophic forgetting in both datasets.ConclusionAlthough the joint retraining model exhibited superior performance, continual learning holds promise in mitigating catastrophic forgetting and facilitating continual model updates while preserving privacy in healthcare deep learning models. Thus, it presents a highly promising solution for the long-term clinical deployment of such models.