Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Background/Aims: Despite the high efficacy of direct-acting antivirals (DAAs), approximately 1–3% of hepatitis C virus (HCV) patients fail to achieve a sustained virological response. We conducted a nationwide study to investigate risk factors associated with DAA treatment failure. Machine-learning algorithms have been applied to discriminate subjects who may fail to respond to DAA therapy. Methods: We analyzed the Taiwan HCV Registry Program database to explore predictors of DAA failure in HCV patients. Fifty-five host and virological features were assessed using multivariate logistic regression, decision tree, random forest, eXtreme Gradient Boosting (XGBoost), and artificial neural network. The primary outcome was undetectable HCV RNA at 12 weeks after the end of treatment. Results: The training (n=23,955) and validation (n=10,346) datasets had similar baseline demographics, with an overall DAA failure rate of 1.6% (n=538). Multivariate logistic regression analysis revealed that liver cirrhosis, hepatocellular carcinoma, poor DAA adherence, and higher hemoglobin A1c were significantly associated with virological failure. XGBoost outperformed the other algorithms and logistic regression models, with an area under the receiver operating characteristic curve of 1.000 in the training dataset and 0.803 in the validation dataset. The top five predictors of treatment failure were HCV RNA, body mass index, α-fetoprotein, platelets, and FIB-4 index. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the XGBoost model (cutoff value=0.5) were 99.5%, 69.7%, 99.9%, 97.4%, and 99.5%, respectively, for the entire dataset. Conclusions Machine learning algorithms effectively provide risk stratification for DAA failure and additional information on the factors associated with DAA failure.

Background/Aims: Chronic hepatitis C (CHC) patients who failed antiviral therapy are at increased risk for hepatocellular carcinoma (HCC). This study assessed the potential role of metformin and statins, medications for diabetes mellitus (DM) and hyperlipidemia (HLP), in reducing HCC risk among these patients. Methods: We included CHC patients from the T-COACH study who failed antiviral therapy. We tracked the onset of HCC 1.5 years post-therapy by linking to Taiwan’s cancer registry data from 2003 to 2019. We accounted for death and liver transplantation as competing risks and employed Gray’s cumulative incidence and Cox subdistribution hazards models to analyze HCC development. Results: Out of 2,779 patients, 480 (17.3%) developed HCC post-therapy. DM patients not using metformin had a 51% increased risk of HCC compared to non-DM patients, while HLP patients on statins had a 50% reduced risk compared to those without HLP. The 5-year HCC incidence was significantly higher for metformin non-users (16.5%) versus non-DM patients (11.3%; adjusted sub-distribution hazard ratio [aSHR]=1.51; P=0.007) and metformin users (3.1%; aSHR=1.59; P=0.022). Statin use in HLP patients correlated with a lower HCC risk (3.8%) compared to non-HLP patients (12.5%; aSHR=0.50; P<0.001). Notably, the increased HCC risk associated with non-use of metformin was primarily seen in non-cirrhotic patients, whereas statins decreased HCC risk in both cirrhotic and non-cirrhotic patients. Conclusions Metformin and statins may have a chemopreventive effect against HCC in CHC patients who failed antiviral therapy. These results support the need for personalized preventive strategies in managing HCC risk.

Periarticular fracture of the shoulder is a common type of fractures in the elderly. Postoperative adverse events such as internal fixation failure, humeral head ischemic necrosis and upper limb dysfunction occur frequently, which seriously endangers the exercise and health of the elderly. Compared with the fracture with normal bone mass, the osteoporotic periarticular fracture of the shoulder is complicated with slow healing and poor rehabilitation, so the clinical management becomes more difficult. At present, there is no targeted guideline or consensus for this type of fracture in China. In such context, experts from Youth Osteoporosis Group of Chinese Orthopedic Association, Orthopedic Expert Committee of Geriatrics Branch of Chinese Association of Gerontology and Geriatrics, Osteoporosis Group of Youth Committee of Chinese Association of Orthopedic Surgeons and Osteoporosis Committee of Shanghai Association of Chinese Integrative Medicine developed the Chinese expert consensus on the diagnosis and treatment of osteoporotic periarticular fracture of the shoulder in the elderly ( version 2023). Nine recommendations were put forward from the aspects of diagnosis, treatment strategies and rehabilitation of osteoporotic periarticular fracture of the shoulder, hoping to promote the standardized, systematic and personalized diagnosis and treatment concept and improve functional outcomes and quality of life in elderly patients with osteoporotic periarticular fracture of the shoulder.

Purpose: The aim of this study was to investigate the efficacy of various epidermal growth factor receptor (EGFR)–tyrosine kinase inhibitors (TKIs) plus bevacizumab in advanced EGFR-mutant lung adenocarcinoma patients. Materials and Methods: From August 2016 to October 2020, we enrolled advanced lung adenocarcinoma patients harboring exon 19 deletion or L858R receiving gefitinib, erlotinib and afatinib plus bevacizumab as the first-line treatment for the purposes of analysis. Results: A total of 36 patients were included in the final analysis. Three patients received gefitinib, 17 received erlotinib, and 16 received afatinib combined with bevacizumab as the first-line treatment. The objective response rate was 77.8%, and disease control rate was 94.4%. The overall median progression-free survival (PFS) was 16.4 months, while the median PFS was 17.1 months in patients with exon 19 deletion, and 16.2 months in patients with L858R mutation (p=0.311). Regarding the use of different EGFR-TKIs, the median PFS was 17.1 months in the erlotinib group and 21.6 months in the afatinib group (p=0.617). In patients with brain metastasis at baseline, the median PFS was 18.9 months in the erlotinib group and 16.4 months in the afatinib group (p=0.747). Amongst patients harboring exon 19 deletion, the median PFS was 16.2 months in the erlotinib group and not-reached in the afatinib group (p=0.141). In patients with L858R mutation, the median PFS was 18.9 months in the erlotinib group and 16.2 months in the afatinib group (p=0.481). Conclusion Our research demonstrates that not only erlotinib combined with bevacizumab, but also afatinib plus bevacizumab as first-line treatment, provides solid clinical efficacy in advanced EGFR-mutant lung adenocarcinoma patients.

Background: and Purpose To explore the causal relationships of elements of the exposome with ischemic stroke and its subtypes at the omics level and to provide evidence for stroke prevention. Methods We conducted a Mendelian randomization study between exposure and any ischemic stroke (AIS) and its subtypes (large-artery atherosclerotic disease [LAD], cardioembolic stroke [CE], and small vessel disease [SVD]). The exposure dataset was the UK Biobank involving 361,194 subjects, and the outcome dataset was the MEGASTROKE consortium including 52,000 participants. Results: We found that higher blood pressure (BP) (systolic BP: odds ratio [OR], 1.02; 95% confidence interval [CI], 1.01 to 1.04; diastolic BP: OR, 1.03; 95% CI, 1.01 to 1.05; pulse pressure: OR, 1.03; 95% CI, 1.00 to 1.06), atrial fibrillation (OR, 1.18; 95% CI, 1.13 to 1.25), and diabetes (OR, 1.13; 95% CI, 1.07 to 1.18) were significantly associated with ischemic stroke. Importantly, higher education (OR, 0.69; 95% CI, 0.60 to 0.79) decreased the risk of ischemic stroke. Higher systolic BP (OR, 1.06; 95% CI, 1.02 to 1.10), pulse pressure (OR, 1.08; 95% CI, 1.02 to 1.14), diabetes (OR, 1.28; 95% CI, 1.13 to 1.45), and coronary artery disease (OR, 1.58; 95% CI, 1.25 to 2.00) could cause LAD. Atrial fibrillation could cause CE (OR, 1.90; 95% CI, 1.71 to 2.11). For SVD, higher systolic BP (OR, 1.04; 95% CI, 1.00 to 1.07), diastolic BP (OR, 1.06; 95% CI, 1.01 to 1.12), and diabetes (OR, 1.22; 95% CI, 1.10 to 1.36) were causal factors. Conclusions The study revealed elements of the exposome causally linked to ischemic stroke and its subtypes, including conventional causal risk factors and novel protective factors such as higher education.