Affiliated Hospital Research Articles

A hierarchical multi-channel multi-modality fusion strategy for integrating digital mammography and synthesized digital breast tomosynthesis for breast mass discrimination

Digital breast tomosynthesis (DBT) combined with full-field digital mammography (FFDM), known as the “combo-mode”, can enhance breast cancer detection and discrimination. However, the DBT is not yet a standard breast cancer screening modality in most hospitals, and the “combo-mode” also doubles the dose exposure to the patient more than FFDM alone. In this study, we synthesized DBT images from FFDM to reduce the extra radiation dose and explored a methodology to effectively integrate multifaceted information from both the synthetic DBT and real FFDM. An improved conditional generative adversarial network (cGAN) network was proposed for generating synthetic DBT with image quality qualified for breast mass discrimination. A novel multiple accuracy metrics scoring (MAMS) strategy was proposed for integrating multichannel and multimodality imaging information within a hierarchical fusion framework. We retrospectively collected 441 patients with both DBT and FFDM from Nanfang Hospital (NFH) and 143 patients with only FFDM from the Second Affiliated Hospital of Guangzhou University of Chinese Medicine (GDHCM), with regions of interest (ROIs) covering the malignant, benign, and normal tissues extracted for model training and validation. The synthesized DBT exhibited satisfactory image quality and comparable discrimination ability with the real DBT on the NFH dataset. The proposed MAMS achieved an accuracy of 80%, 79%, 91%, and AUC of 87%, 85%, and 99%, respectively, for the malignant, benign, and normal tissues on the GDHCM dataset. The experimental results indicate that the findings of this study can help improve breast mass discrimination when only unimodal FFDM is available.

Development and validation of a machine learning-based nomogram for survival prediction of patients with hilar cholangiocarcinoma after curative-intent resection

Hilar cholangiocarcinoma (hCCA), a rare cancer of the biliary system, has a poor prognosis. This study aimed to investigate the risk factors affecting the survival of patients with hCCA after curative-intent resection and establish a survival predictive model. Clinical data from 340 hCCA patients who underwent curative-intent resection at the First Affiliated Hospital of Xi’an Jiaotong University between 2010 and 2021 were collected. The patients were randomly assigned to a training set and a testing set in a 7:3 ratio. Risk factors selection was performed by five machine learning (ML) algorithms, including Least Absolute Shrinkage and Selection Operator (LASSO) Regression, Forward Stepwise Cox regression, Boruta feature selection, Random Forest and eXtreme Gradient Boosting (XGBoost). A nomogram was constructed based on identified risk factors. The independent risk factors for the postoperative survival in hCCA patients included positive margin, lymph node metastasis, low total lymph node count (TLNC) and poor tumor differentiation. In the training and testing sets, the consistency index (C-index) of ML-based nomogram was 0.731 (95% CI: 0.684–0.753) and 0.714 (95% CI: 0.661–0.775), while the 3-year AUC of the nomogram was 0.784 (95% CI: 0.724–0.844) and 0.770 (95% CI: 0.763–0.867), respectively. The calibration curves for the nomogram showed good concordance. Based on the decision curve analysis, the nomogram had a good clinical application value, outperforming both the TNM staging system and the Bismuth-Corlette classification. Furthermore, patients were stratified into three groups with varying risks of overall survival (OS): the low-risk, middle-risk and high-risk group according to the nomogram, with statistically significant differences observed among these groups (p < 0.001). The ML-based nomogram provided a personalized prognostic prediction model for hCCA patients after surgical resection.

The Clinical Significance of Non-Criteria Antiphospholipid Antibodies in Atypical Antiphospholipid Syndrome.

To analyse the correlation between non-criteria antiphospholipid antibodies (non-criteria aPLs) and clinical symptoms (recurrent miscarriage) in patients with seronegative antiphospholipid syndrome (SNAPS).Methods:Ninety-four SNAPS patients who were treated in the First Affiliated Hospital of Soochow University from May 2022 to May 2024 were included. Chemiluminescence immunoassay was used to detect anti-β2 glycoprotein I antibody (aβ2GPI), anti-cardiolipin antibody (aCL), and lupus anticoagulant. ELISA was used to detect the levels of anti-annexin A5 (ANXA5), anti-phosphatidylethanolamine antibody (aPE), anti-prothrombin IgG antibody (aPT IgG), anti-prothrombin IgM antibody (aPT IgM), anti-phosphatidylserine/prothrombin IgG antibody (aPS/PT IgG), and anti-phosphatidylserine/prothrombin IgM antibody (aPS/PT IgM). The correlation between non-criteria antiphospholipid antibodies and clinical events was analysed. Compared with the non-criteria antibody-negative group, the positive group had a higher rate of positive antinuclear antibodies (p < 0.05). The proportion of recurrent miscarriage in the positive non-criteria antibody group was significantly higher than in the negative group (p = 0.002). Compared with the negative group, patients with positive anti-annexin A5 antibody and aPT IgM were more likely to have recurrent miscarriage (p < 0.05). Compared with the aPE-negative group, those with aPE-positive were more likely to experience early miscarriage (p = 0.039). Both aPT IgM subtype (p = 0.004) and ANXA5 (p = 0.013) antibodies were the risk factors for recurrent miscarriage. Patients with positive ANXA5, aPT IgG, and aPT/PS IgM antibodies showed a significant decrease in antibody titers after treatment (p < 0.05). Detection of non-criteria aPLs contributes to improving the diagnosis of SNAPS, assessing its development trends, and intervening promptly to prevent recurrent miscarriage.

Preliminary observations of extraocular muscle changes in patients with intermittent exotropia during binocular convergence based on MRI

Objective: To investigate changes in diffusion-weighted imaging (DWI) of extraocular muscles in patients with intermittent exotropia (IXT) during binocular convergence using MRI, and to explore the imaging characteristics of extraocular muscle alterations in IXT. Methods: It was a cross-sectional study. Patients diagnosed with IXT at the First Affiliated Hospital of Anhui Medical University from January 2023 to September 2023 were enrolled as the IXT group, with age-and sex-matched healthy individuals recruited as the control group. Axial MRI scanning was used to measure DWI signal intensity values of the medial rectus, lateral rectus, and optic nerve during primary gaze and convergence gaze. The normalized DWI values between the two groups and two gaze conditions were compared. The correlation between normalized DWI value changes and the degree of strabismus in the IXT group was analyzed. Normally distributed data were analyzed using independent and paired t-tests, while non-normally distributed data were evaluated with Mann-Whitney and Wilcoxon signed-rank tests. Pearson correlation analysis was used to assess the relationship between DWI values and the degree of strabismus. Results: Eleven patients (22 eyes) were included in the IXT group (six males, five females; mean age, 24±4 years), and 10 healthy controls (20 eyes) were included (four males, six females; mean age, 24±1 years). Baseline characteristics were comparable between groups (P>0.05). In the control group, the normalized DWI value of the medial rectus was 0.86 (0.72, 0.97) during primary gaze and 0.97 (0.87, 1.20) during convergence gaze, with a significant difference (P<0.05). In the IXT group, the normalized DWI value of the medial rectus was 0.77 (0.72, 0.93) during primary gaze and 0.81 (0.69, 1.08) during convergence gaze, with no significant difference (P>0.05). A negative correlation trend was observed between changes in normalized DWI values of the medial rectus during convergence gaze and the degree of strabismus in the IXT group (r=-0.58, P=0.059). Conclusion: Patients with IXT exhibit significantly reduced changes in normalized DWI values of the medial rectus during binocular convergence compared to healthy controls.

Efficacy of extracranial vertebral artery reconstruction for symptomatic vertebral artery origin in-stent restenosis or occlusion: a ten-year retrospective cohort study.

The optimal management of symptomatic vertebral artery in-stent restenosis or occlusion (SVISRO) after stent-assisted angioplasty (SAA) at the vertebral artery origin remains unclear. This study aimed to investigate the efficacy of vertebral artery reconstruction surgery (VRS) surgery among SVISRO patients. A retrospective study was conducted to analyze the clinical data for SVISRO patients admitted to the Third Affiliated Hospital of Sun Yat-sen University between May 2011 and November 2021. The stroke recurrence and stroke-free rates during the follow-up (FU) period and the patients' neurological statuses at the last FU were compared between a VRS group and an intensive medical therapy (IMT) group. Sixty-two SVISRO patients with an average age of 60.1 ± 8.3 years and an average FU duration of 67.6 ± 32.5 months were studied. The VRS group had significantly fewer stroke recurrences than the IMT group did (5.7% vs. 25.9%, p = 0.034). A 154-month stroke-free rate of 73.8% (95% CI, 24.5% to 93.7%) was observed in the VRS group, whereas the IMT group had a stroke-free rate of 33.9% (95% CI, 17.3% to 75.0%). The hazard ratio (log-rank) between the two groups was 0.234 (95% CI, 0.063-0.871; p = 0.048). The modified Rankin scale score for the VRS group was significantly better than that for the IMT group at the final FU (p = 0.032). In patients with SVISRO secondary to SAA, VRS targeting the extracranial vertebral artery (V1-V2) appears to reduce stroke recurrence, increase the stroke-free rate and improve neurological status.

Retrospective review: single- and multidonor washed microbiota transplantation have equivalent efficacy in the treatment of autism

BackgroundAutism spectrum disorder (ASD) is a serious neurodevelopmental disorder with no effective treatment. This study explored the short-term clinical effects of washed microbiota transplantation (WMT) with different numbers of donors on autism.MethodsConsecutive ASD patients treated with two continuous WMT courses from March 2020 to March 2022 at the First Affiliated Hospital of Guangdong Pharmaceutical University were retrospectively assessed. Basic information, aberrant behavior checklist (ABC) scores, childhood autism rating scale (CARS) scores, sleep disturbance scale for children (SDSC) scores, adverse reactions, and feces were collected.ResultsForty-four patients were included (single-donor group: 17 patients; multidonor group: 27 patients). The CARS, ABC and SDSC scores didn’t differ between the two groups before treatment. After two courses, the scores for the 44 patients were lower than those at baseline (P&amp;lt;0.05), with no severe adverse reactions observed. After the first course, the mean ABC (P=0.049) and SDSC (P=0.019) scores were significantly different between the single-donor and multidonor groups, but the difference disappeared after two courses. The alpha-diversity of the faecal flora in the effective-group was greater than that in the ineffective-group (Shannon index P=0.0018). Lactobacillus was the predominant genus in the effective group, whereas Faecalibacterium, Campylobacter, and Sphingomonas were predominant genera in the ineffective group.ConclusionAfter two WMT courses, the symptoms of ASD improved, with good short-term treatment efficacy. The ASD symptom improvement did not differ between the single-donor and multidonor groups. Changes in the alpha-diversity and abundance of the faecal microbiota after WMT may be related to treatment efficacy.

Knowledge, attitudes, and willingness of patients with thyroid diseases toward thyroid thermal ablation techniques

To investigate the knowledge, attitudes, and willingness (KAW) of patients with thyroid disease regarding thyroid thermal ablation techniques and explore the factors associated with KAW. This cross-sectional survey was conducted at Yantai Hospital of Shandong Wendeng Orthopaedics & Traumatology and Yantai Affiliated Hospital of Binzhou Medical College between October 2022 and March 2023. This study included 632 patients; 66.14% were female. The mean knowledge, attitude, and willingness scores were 6.03 ± 2.42 (possible range: 0–10), 17.52 ± 2.91 (possible range: 5–25), and 33.02 ± 6.34 (possible range: 8–40), indicating poor knowledge, positive attitudes, and proactive willingness. Multivariable analysis showed that ≥ 51 years old, urban areas, consuming alcohol, medical treatment, and surgical treatment were independently associated with adequate knowledge. The knowledge scores, ≥ 51 years old, females, urban areas, medical treatment, and surgical treatment were independently associated with a positive attitude. Only the attitude scores were independently associated with proactive willingness. Patients with thyroid diseases have poor knowledge, positive attitudes, and proactive practice toward thermal ablation. Sustained efforts are required to increase knowledge about thyroid thermal ablation techniques.

Innovative multimodal educational strategies: assessing the impact of integrative teaching methods on standardized neurology resident training

ObjectiveThis study explores the application effects of a novel teaching model that combines micro-learning, bedside teaching, and case-based learning (CBL) in the standardized training of neurology residents.MethodsFrom January 2022 to December 2024, a total of 103 neurology residents undergoing standardized training at the First Affiliated Hospital of Zhengzhou University were selected as the study subjects. A quantitative research design was employed, and participants were randomly divided into an experimental group and a control group. The experimental group (51 participants) utilized a novel teaching model that integrated micro-learning, bedside teaching, and CBL, while the control group (52 participants) employed Lecture-Based Learning (LBL) methods. The study focused on several indicators, including theoretical assessment scores from standardized training, student satisfaction and interest in the teaching content, clinical competency evaluations, and self-assessments by instructors.ResultsThe novel teaching model combining micro-learning, bedside teaching, and CBL significantly outperformed traditional teaching methods in improving residents’ theoretical assessment scores (83.02 ± 7.33 vs. 76.04 ± 7.19) and clinical competency evaluations, particularly in comprehensive ability assessments (85.55 ± 5.24 vs. 75.10 ± 7.86). Students also reported a marked increase in satisfaction with the teaching content and learning interest (P < 0.01). Furthermore, instructors expressed a high level of recognition for the new course model in their post-class self-evaluations (P < 0.01).ConclusionThe findings of this study indicate that the integration of micro-learning, bedside teaching, and CBL can provide a more effective teaching model for the training of neurology residents, offering strong support for future reforms and developments in medical education.

MSH6 as a prognostic biomarker in bladder cancer and its correlation with immunity

This study aims to investigate the role of MSH6 in the diagnosis and prognosis of bladder cancer and its association with immunity. Various analyses were conducted on The Cancer Genome Atlas (TCGA) data, and the results were validated using the Gene Expression Omnibus (GEO) database. The potential mechanism of MSH6 in bladder cancer was revealed using R tools, Gene Set Enrichment Analysis (GSEA), and Gene Ontology (GO) analysis. Bladder cancer patients admitted to the Affiliated Hospital of Inner Mongolia Medical University from June 2023 to December 2023 were enrolled. Pathological specimens of bladder cancer and adjacent tissues were obtained. Immunohistochemical results showed the expression of MSH6 in bladder cancer tissues was higher than in adjacent tissues. High expression of MSH6 was identified as an independent prognostic factor for bladder cancer. The occurrence of bladder cancer was also influenced by age, pathological T stage, and pathological stage. According to the Kaplan–Meier survival curve, the overall survival rate of the MSH6 high-expression group was lower (P = 0.004). After MSH6 was knocked down in cell experiments, the proliferation, colony formation, and migration abilities of bladder cancer cells were significantly inhibited. MSH6 is significantly associated with the diagnosis and prognosis of bladder cancer, suggesting its potential utility as a biomarker for these purposes. MSH6 is closely related to the tumor immune microenvironment, implying a significant role in immunotherapy.

Applying Large Language Models for Surgical Case Length Prediction

Accurate prediction of surgical case duration is critical for operating room (OR) management, as inefficient scheduling can lead to reduced patient and surgeon satisfaction while incurring considerable financial costs. To evaluate the feasibility and accuracy of large language models (LLMs) in predicting surgical case length using unstructured clinical data compared to existing estimation methods. This was a retrospective study analyzing elective surgical cases performed between January 2017 and December 2023 at a single academic medical center and affiliated community hospital ORs. Analysis included 125 493 eligible surgical cases, with 1950 used for LLM fine-tuning and 2500 for evaluation. An additional 500 cases from a community site were used for external validation. Cases were randomly sampled using strata to ensure representation across surgical specialties. Eleven LLMs, including base models (GPT-4, GPT-3.5, Mistral, Llama-3, Phi-3) and 2 fine-tuned variants (GPT-4 fine-tuned, GPT-3.5 fine-tuned), were used to predict surgical case length based on clinical notes. The primary outcome was average error between predicted and actual surgical case length (wheels-in to wheels-out time). The secondary outcome was prediction accuracy, defined as predicted length within 20% of actual duration. Fine-tuned GPT-4 achieved the best performance with a mean absolute error (MAE) of 47.64 minutes (95% CI, 45.71-49.56) and R2 of 0.61, matching the performance of current OR scheduling (MAE, 49.34 minutes; 95% CI, 47.60-51.09; R2, 0.63; P = .10). Both GPT-4 fine-tuned and GPT-3.5 fine-tuned significantly outperformed current scheduling methods in accuracy (46.12% and 46.08% vs 40.92%, respectively; P < .001). GPT-4 fine-tuned outperformed all other models during external validation with similar performance metrics (MAE, 48.66 minutes; 95% CI, 45.31-52.00; accuracy, 46.0%). Base models demonstrated variable performance, with GPT-4 showing the highest performance among non-fine-tuned models (MAE, 59.20 minutes; 95% CI, 56.88 - 61.52). The findings in this study suggest that fine-tuned LLMs can predict surgical case length with accuracy comparable to or exceeding current institutional scheduling methods. This indicates potential for LLMs to enhance operating room efficiency through improved case length prediction using existing clinical documentation.

Is sterility essential for hand-drying products in surgical hand antisepsis? A controlled before-and-after study.

Is sterility essential for hand-drying products in surgical hand antisepsis? A controlled before-and-after study.

Effects of different embryo culture media on birthweight following assisted reproductive technology

Abstract STUDY QUESTION Does the type of embryo culture medium affect the birthweight of newborns conceived by ART? SUMMARY ANSWER After fresh embryo transfers, singleton newborns in the G5 and HTF groups exhibited higher birthweight z-scores and increased risks of being large for gestational age (LGA) compared to those in the Cook group. WHAT IS KNOWN ALREADY Current studies have not yet determined whether embryo culture medium affects birthweight, and if such an effect does exist, the significance of embryo culture medium among all the influencing factors is not yet clear. STUDY DESIGN, SIZE, DURATION A retrospective cohort study including 23,403 fresh ET cycles between 1st January 2010 and 31st December 2022 at the reproductive medical center of a university—affiliated hospital was conducted. PARTICIPANTS/MATERIALS, SETTING, METHODS We performed a retrospective cohort study including 23,403 fresh ET cycles. Four embryo culture media were analyzed: Cook, G5-PLUS, G5 and HTF. Multiple linear regression analysis was used to evaluate potential associations between embryo culture medium and birthweight z-score. Logistic regression analysis was used to evaluate potential associations between embryo culture medium and the risk of LGA and macrosomia. Random forest models were constructed to conduct significance analysis of all factors that may affect birthweight z-score. MAIN RESULTS AND THE ROLE OF CHANCE The cohort comprised 4453, 8460, 7463 and 3027 singletons in the Cook, G5-PLUS, G5 and HTF groups, respectively. Compared to the Cook group, newborns in the G5 and HTF groups had a higher birthweight z-scores (increased by 0.069 units, P &amp;lt; 0.001, and 0.073 units, P = 0.002, respectively) and a higher risks of LGA (OR: 1.25, 95% CI: 1.12-1.39, P &amp;lt; 0.001; OR: 1.20, 95% CI: 1.05-1.37, P = 0.009, respectively), while newborns in the G5 group also had a higher risk of macrosomia (OR: 1.21, 95% CI: 1.06-1.39, P = 0.006). Of the main factors influencing birthweight, the embryo culture medium had a moderate significance. LIMITATIONS, REASONS FOR CAUTION Due to commercial and regulatory reasons, various culture media were used for different periods, thus resulting in the different number of cases for the four culture media groups. WIDER IMPLICATIONS OF THE FINDINGS Our analysis revealed that the type of embryo culture medium directly affected birthweight z-scores and the risk of LGA and macrosomia in newborns conceived by ART. Consequently, the selection of embryo culture medium shoud be made cautiously. In addition, new embryo culture media require comprehensive safety verification before being marketed. STUDY FUNDING/COMPETING INTEREST(S) This study was supported by National Key Research and Development Program of China [2023YFC2705604] and the National Natural Science Foundation of China [82071721, 82371706 and 82288102]. All authors have no conflicts of interest to report. TRIAL REGISTRATION NUMBER N/A.

Association of visceral obesity indicators with prostate cancer: a cross-sectional study from Xinjiang

BackgroundProstate cancer (PCa) is a common malignancy among men worldwide, and its risk is strongly associated with obesity, especially visceral obesity. Visceral obesity has been assessed by the visceral adiposity index (VAI), cardiometabolic index (CMI), and lipid accumulation product (LAP), but their associations with PCa remain underexplored. This study investigated the relationship between these visceral obesity indicators and PCa risk.MethodsData for this cross-sectional study were obtained from the First Affiliated Hospital of Xinjiang Medical University from 2022-2023, and 730 participants were screened for the study. A total of 102 PCa patients were included as the PCa group and 102 healthy individuals as the control group using propensity score matching (PSM). We collected anthropometric data (height, weight, waist circumference) and blood biochemical parameters from participants to calculate the VAI, CMI and LAP. These indicators’ association and predictive efficacy with PCa were assessed by logistic regression, restricted cubic spline (RCS), and receiver operating characteristic (ROC) curve analysis. The robustness of these results was further examined through sensitivity analyses.ResultsVAI, CMI, and LAP were higher in the PCa group than in the control group (P&amp;lt;0.05). Logistic regression models showed that VAI, CMI, and LAP were positively associated with PCa. This association of VAI and CMI shows robustness in sensitivity analysis. Compared with the first quartile (Q1), the fourth quartile’s (Q4) VAI, CMI and LAP were linked to an increased risk of PCa (OR: 9.07, 95% CI: 3.21-25.65; OR: 11.10, 95% CI: 3.87-31.83; OR: 3.01, 95% CI: 1.17-7.76, respectively). RCS analysis showed that VAI and CMI were nonlinearly associated with PCa risk, and LAP was linearly associated with PCa risk. The area under the ROC curve (AUC) of VAI, CMI, and LAP was 0.721 (95% CI: 0.651-0.791), 0.711 (95% CI: 0.639-0.782), and 0.593 (95% CI: 0.515-0.671), respectively.ConclusionsVisceral obesity indicators are closely associated with PCa, of which VAI and CMI show good predictive value and robustness, and can be used as potential biomarkers for assessing PCa risk.

Randomised controlled trial comparing low doses of aspirin in the prevention of pre-eclampsia (ASAPP): a study protocol.

Pre-eclampsia (PEC) is a morbid and potentially lethal complication of pregnancy and is more common in women with risk factors such as hypertension, diabetes, autoimmune disease, kidney disease or multifetal pregnancies. Low dose aspirin (ASA) is currently the only prophylactic therapy known to decrease PEC in this patient population. However, currently, there is no prospective literature comparing various low-dose ASA formulations in the risk reduction of PEC. In the USA, the currently available low-dose ASA is over-the-counter and found in 81 mg tablets. Therefore, when clinicians initiate low-dose ASA therapy, they may prescribe one or two tablets of 81 mg per day without comparative evidence to guide their decision. Our objective is to prospectively compare pregnant patients on 81 mg vs 162 mg of ASA and determine a possible dose response in the prevention of PEC. We designed a pragmatic phase 3 prospective randomised open label blinded-end point clinical trial with parallel assignment between two groups of pregnant people at high risk for PEC, as defined by the US Preventive Services Task Force and American College of Obstetricians and Gynecologists (ACOG). The primary outcome is the incidence of preterm (<37 weeks' gestation) PEC and PEC with severe features. Secondary outcomes include adherence to therapy, maternal and fetal complications of PEC, and time-to-event development of PEC across the two treatment groups. Participants are recruited from a tertiary care academic institution and its affiliated community hospital and clinics if they present for prenatal care at <16 weeks' gestation. Women are randomised and accordingly instructed to either take one or two tablets of 81 mg ASA daily. Participants are followed from recruitment until 6 weeks postpartum, with adherence questionnaires collected at 18-22 weeks, 24-28 weeks and 34-38 weeks gestation via a Simple Medication Adherence Questionnaire. After the first 100 pregnancies are completed, a prespecified interim analysis is performed to assess feasibility, tailor power calculations, and ensure safety. The second stage comprises the continuation of recruitment and randomisation to the accrual ceiling as determined by the first stage. Analysis will be completed in an intention-to-treat manner. Our research protocol and safety monitoring protocol have been approved by the Weill Cornell Medicine institutional review board (IRB) and the Office of Human Research Protection. Evaluation for adverse events will be monitored throughout the study period. Adverse events will be assessed at each study visit. NCT04070573.

Explainable machine learning model for predicting the transarterial chemoembolization response and subtypes of hepatocellular carcinoma patients

BackgroundHepatocellular carcinoma (HCC), the third leading cause of cancer-related deaths globally, faces heterogeneous responses to transarterial chemoembolization (TACE) in intermediate-stage disease. We developed a Machine Learning (ML)-based model integrating routine clinical variables to preoperatively predict TACE efficacy, enabling tailored TACE candidate selection and optimized therapeutic decision-making.MethodsThis retrospective multicentre study enrolled treatment-naive HCC patients undergoing initial TACE from two independent cohorts: the First Affiliated Hospital of Wenzhou Medical University (training cohort) and Wenzhou Central Hospital (external validation cohort). Through recursive feature elimination (RFE), we systematically developed prediction models employing ten distinct ML algorithms. The SHAP algorithm was implemented to enhance model interpretability, while patient stratification was subsequently performed using PCA and K-means clustering to facilitate comprehensive prognostic analysis.ResultsWe retrospectively collected 382 unresectable HCC patients from the First Affiliated Hospital of Wenzhou Medical University and 52 from Wenzhou Central Hospital. RFE method identified 10 predictors for constructing ML models. XGBoost and CatBoost outperformed other algorithms, achieving AUCs of 0.796–0.799 (internal test) and 0.785–0.791 (external validation) with balanced accuracy (76-76.8%). SHAP interpretability revealed tumor burden and hepatic function markers as key determinants of TACE resistance. K-means clustering stratified patients into two prognostically distinct subgroups: Cluster B showed significantly longer survival than Cluster A (HR = 0.36, 95%CI:0.26–0.49, P < 0.001), confirming the clinical relevance of ML-selected features.ConclusionWe developed and validated an interpretable ML-based system integrating predictive modelling and patient clustering to individualize TACE efficacy prediction and clinical risk stratification for HCC patients.

Long-Term Outcomes of Percutaneous Balloon Compression for Trigeminal Neuralgia: A Retrospective Single-Center Study

Objective: This study aimed to evaluate the long-term outcomes of percutaneous balloon compression (PBC) in individuals diagnosed with trigeminal neuralgia (TN) and to assess whether procedural modifications influence the durability of pain relief. Methods: This retrospective study included patients diagnosed with TN who underwent PBC at the Department of Neurosurgery, Affiliated Hospital of Beihua University, between January 2017 and December 2025. Two procedural protocols were compared. In group A (treated from January 2017 to December 2020), balloon compression was maintained at baseline pressure for 180 seconds after achieving the optimal “pear shape.” In group B (treated from January 2021 to December 2025), compression was maintained for 120 seconds at a pressure increased by 10% above baseline after pear shape formation. Postoperative pain relief, recurrence rates, and complication rates were compared between the 2 groups. Results: Immediate postoperative pain relief rates within 24 hours did not differ significantly between the 2 groups (90.7% versus 97.6%; P=0.109). However, the 5-year recurrence rate was significantly lower in group B (9.6%) compared with group A (22.4%; P=0.043). The incidence of complications, excluding numbness, showed no significant difference between the groups (22.2% versus 27.1%; P=0.522). Conclusions: A modified PBC technique involving an additional 10% increase in pressure for 120 seconds after optimal balloon positioning was associated with improved long-term pain control in patients with TN, without a corresponding increase in procedural complications. These findings suggest that procedural optimization may enhance the durability of therapeutic outcomes in PBC for this patient population.

Neoadjuvant immunochemotherapy versus chemotherapy for elderly patients with IB–IIIB non-small-cell lung cancer in real-world practice

BackgroundImmunochemotherapy is currently the standard treatment plan for non-small-cell lung cancer (NSCLC) patients with no driver gene mutations. However, it is uncertain whether immunochemotherapy is effective or safe for elderly patients (≥ 65 years). Therefore, this study compared preoperative immunochemotherapy and chemotherapy with stage IB-IIIB NSCLC in elderly patients (≥ 65 years) for safety and effectiveness.MethodsBetween 2016 and 2022, all elderly patients (≥ 65 years) with IB–IIIB NSCLC who underwent 2–4 cycles of neoadjuvant immunochemotherapy or chemotherapy at the Department of Thoracic Surgery, the First Affiliated Hospital, Zhejiang University School of Medicine were consecutively included in this retrospective study. In this study, we followed up patients for at least 1 year after surgery or until they abandoned treatment. The primary endpoints of this study were disease-free survival (DFS) and overall survival (OS). The secondary endpoints were composed of pathological response [major pathological response (MPR) and pathological complete remission (pCR)], adverse events (AEs) and objective response rate (ORR).ResultsWe included 140 patients in our study and classified them into two groups according to their neoadjuvant treatment regimen: chemotherapy (n = 47) and immunochemotherapy (n = 93). In the chemotherapy group, the median DFS was 19.4 months [95% confidence interval [CI], 2.4 to 36.4], but not achieved by the immunochemotherapy group (hazard ratio [HR] for disease progression, disease recurrence, or death, 0.247; 95% CI, 0.122 to 0.501; P < 0.001). The chemotherapy group had a median OS of 41.8 months (95% CI, 9.1 to 74.5), while the median OS of the immunochemotherapy group was not achieved (HR for death,0.265; 95% CI, 0.115 to 0.611; P < 0.001). The immunochemotherapy group had a significantly higher ORR than the chemotherapy group (73.1% vs. 29.8%, P < 0.001). The immunochemotherapy group had 19.4% of grade 3–4 AEs, while the chemotherapy group had 8.5% (P = 0.156). The immunochemotherapy group showed a significantly higher rate of MPR than the chemotherapy group (64.4% vs. 25.8%, P < 0.001). The immunochemotherapy group had a pCR rate of 32.2%, while the chemotherapy group had a pCR rate of 16.1% (P = 0.101).ConclusionsNeoadjuvant immunochemotherapy was associated with improved survival rates, ORR, and MPR rates compared to chemotherapy alone in elderly patients (≥ 65 years) with IB-IIIB NSCLC, without a significant increase in AEs.

Surgeon Preferences, Surgical Location, and Timing of Repair Drive Achilles Tendon Rupture Repair Cost.

Midsubstance Achilles tendon repair is a common procedure, yet the factors influencing its cost have been underexamined and often imprecisely understood. Using time-driven activity-based costing (TDABC), we examined variation in total cost, compared patient-, surgeon-, and surgery-specific characteristics between high- and non-high-cost repairs, and assessed factors associated with total cost to identify specific cost drivers that might reduce expenses without compromising clinical outcomes. Patients undergoing midsubstance Achilles tendon rupture repair between January 3, 2022 and December 28, 2023 at 2 academic medical centers and their affiliated community hospitals and ambulatory surgical centers were identified. Exclusion criteria included revision procedures and those with concurrent procedures besides fasciotomy. TDABC methodology was used to determine total cost across 3 clinical phases: preoperative, intraoperative, and postoperative. Cost was normalized per institution requirements. Bivariate analysis was performed across all characteristics between high-cost (top decile) procedures and all others. Spearman correlation between operative time and total cost was assessed. Multivariable linear regression was used to identify key cost drivers. Among 341 primary midsubstance Achilles rupture repairs performed by 8 surgeons, the most expensive procedure was 6 times costlier than the least expensive one. Eighty-five percent of the cost, on average, was incurred intraoperatively. A "good" correlation was found between operative time and total cost (P < .05). When accounting for covariates, increasing time between rupture and repair was associated with higher cost (P < .05). Surgical location, surgical approach, and repair suture were also associated with total cost (P < .05). Substantial variability in the cost of primary midsubstance Achilles rupture repair is driven by factors including timing of surgical repair, intraoperative surgeon-specific characteristics (ie, surgical approach and suture choice), and surgical location. Further standardization of treatment approach and decreasing time to the operating room could lower cost variability and improve the value of care for patients with midsubstance Achilles tendon rupture undergoing surgery.

Short-Interval Pregnancy Following Delivery in Catholic-Affiliated Versus Non-Catholic-Affiliated Hospitals Among Patients Insured Through the Medicaid Program.

As the number of U.S. Catholic-affiliated hospitals increases, more patients who receive care at Catholic hospitals may face restrictions to reproductive healthcare services, including postpartum contraception and sterilization. Little is known about how Catholic hospital affiliation affects the likelihood of a short-interval pregnancy following a delivery. A short interpregnancy interval is associated with poor obstetric outcomes in subsequent pregnancies. In this retrospective study of Medicaid claims data, we examined if an index delivery at a Catholic-affiliated hospital was associated with a short-interval pregnancy among patients in nine states (2010-2014). We used Cox proportional hazards regression to compare the time to subsequent pregnancy for those who delivered at Catholic-affiliated hospitals versus non-Catholic-affiliated hospitals. We also estimated the likelihood of a subsequent pregnancy by 6, 12, and 18 months following an index delivery. We matched 1,197,898 index births to a provider and the religious affiliation of the delivery hospital. Across all nine states, 20.9% (249,764) of deliveries were at Catholic hospitals. After adjusting for maternal age, maternal race and ethnicity, and rural residence, subjects who delivered in a Catholic hospital had a 5% increased adjusted odds of conception by 18 months (95% confidence interval: 1.04-1.07; adjusted odds ratio: 1.07, 95% confidence interval: 1.05-1.10 by 6 months and adjusted odds ratio: 1.06, 95% confidence interval: 1.05-1.08 by 12 months for Catholic vs. non-Catholic hospitals). Medicaid enrollees who deliver at Catholic hospitals are at increased risk for short-interval pregnancies.

Independent Association of Sleep Apnea-Specific Hypoxic Burden and Sleep Breathing Impairment Index with Thyroid Function in Obstructive Sleep Apnea: A Retrospective Study

PurposeThis study aimed to explore thyroid function changes in patients with obstructive sleep apnea (OSA) and analyze the relationships among the sleep apnea-specific hypoxic burden (SASHB), the sleep breathing impairment index (SBII), and the function during different sleep stages.MethodsThis retrospective study included 452 patients with OSA who visited the Second Affiliated Hospital of Xi’an Jiaotong University between August 2017 and March 2024. The severity of OSA was evaluated, grouping patients by their apnea–hypopnea index (AHI), SASHB, SBII, and both SASHB and SBII during non-rapid eye movement (NREM) and rapid eye movement (REM) sleep. Multiple linear regression analyzed the independent relationships between the AHI, SASHB, SBII, and thyroid function indicators while adjusting for potential confounders.ResultsSignificant intergroup differences were observed in thyroid function indicators under various grouping methods, with different trends. After adjusting for confounding factors, SASHB, SASHB during NREM sleep (NREM-SASHB), SASHB during REM sleep (REM-SASHB), SBII, SBII during NREM sleep (NREM-SBII), and SBII during REM sleep (REM-SBII) were independently associated with elevated serum free triiodothyronine (FT3) levels (p < 0.05). Similar results were noted in the male patients, whereas no significant associations were observed in the female patients.ConclusionThere is an association between OSA and thyroid function, with SASHB and SBII independently linked to elevated FT3 levels across different sleep stages and sex subgroups. Future research should further explore these mechanisms to optimize clinical management and treatment strategies for patients with OSA.