China Medical University Hospital Research Articles

Preoperative assessment of bone density for dental implantation: a comparative study of three different ROI methods

BackgroundDental cone beam computed tomography (CBCT) is commonly used to evaluate cancellous bone density before dental implant surgery. However, to our knowledge, no measurement approach has been standardized yet. This study aimed to evaluate the relationship between three different regions of interest (ROI) methods on cancellous bone density at the dental implant site using dental CBCT images.MethodsPatients’ dental CBCT images (n = 300) obtained before dental implant surgery were processed using Mimics (Materialise, Leuven, Belgium). At the potential implant sites, the rectangle, cylinder, and surrounding cylinder ROI methods were used to measure bone density. Repeated measures one-way analysis of variance was performed to compare the three ROI methods in terms of measurement results. Pearson correlation analysis was performed to identify the likely pair-wise correlations between the three ROI methods.ResultsThe density value obtained using the surrounding cylinder approach (grayscale value [GV],523.56 ± 228.03) was significantly higher than the values obtained using the rectangle (GV, 497.04 ± 236.69) and cylinder (GV,493 ± 231.19) ROI methods in terms of results. Furthermore, significant correlations were noted between the ROI methods (r > 0.965; p < 0.001).ConclusionsThe density measured using the surrounding cylinder method was the highest. The choice of method may not influence the trends of measurement results.Trial registrationThis study was approved by the Institutional Review Board of China Medical University Hospital, No. CMUH111-REC3-205. Informed consent was waived by the Institutional Review Board of China Medical University Hospital, CMUH111-REC3-205, owing to the retrospective nature of the study.

Genome‑wide association study and polygenic risk scores predict psoriasis and its shared phenotypes in Taiwan.

Psoriasis is a chronic inflammatory dermatological disease, and there is a lack of understanding of the genetic factors involved in psoriasis in Taiwan. To establish associations between genetic variations and psoriasis, a genome‑wide association study was performed in a cohort of 2,248 individuals with psoriasis and 67,440 individuals without psoriasis. Using the ingenuity pathway analysis software, biological networks were constructed. Human leukocyte antigen (HLA) diplotypes and haplotypes were analyzed using Attribute Bagging (HIBAG)‑R software and chi‑square analysis. The present study aimed to assess the potential risks associated with psoriasis using a polygenic risk score (PRS) analysis. The genetic association between single nucleotide polymorphisms (SNPs) in psoriasis and various human diseases was assessed by phenome‑wide association study. METAL software was used to analyze datasets from China Medical University Hospital (CMUH) and BioBank Japan (BBJ). The results of the present study revealed 8,585 SNPs with a significance threshold of P<5x10‑8, located within 153 genes strongly associated with the psoriasis phenotype, particularly on chromosomes 5 and 6. This specific genomic region has been identified by analyzing the biological networks associated with numerous pathways, including immune responses and inflammatory signaling. HLA genotype analysis indicated a strong association between HLA‑A*02:07 and HLA‑C*06:02 in a Taiwanese population. Based on our PRS analysis, the risk of psoriasis associated with the SNPs identified in the present study was quantified. These SNPs are associated with various dermatological, circulatory, endocrine, metabolic, musculoskeletal, hematopoietic and infectious diseases. The meta‑analysis results indicated successful replication of a study conducted on psoriasis in the BBJ. Several genetic loci are significantly associated with susceptibility to psoriasis in Taiwanese individuals. The present study contributes to our understanding of the genetic determinants that play a role in susceptibility to psoriasis. Furthermore, it provides valuable insights into the underlying etiology of psoriasis in the Taiwanese community.

Multiple polygenic risk scores can improve the prediction of systemic lupus erythematosus in Taiwan

ObjectiveTo identify new genetic variants associated with SLE in Taiwan and establish polygenic risk score (PRS) models to improve the early diagnostic accuracy of SLE.MethodsThe study enrolled 2429 patients with...

Utility of polygenic scores across diverse diseases in a hospital cohort for predictive modeling

Polygenic scores estimate genetic susceptibility to diseases. We systematically calculated polygenic scores across 457 phenotypes using genotyping array data from China Medical University Hospital. Logistic regression models assessed polygenic scores’ ability to predict disease traits. The polygenic score model with the highest accuracy, based on maximal area under the receiver operating characteristic curve (AUC), is provided on the GeneAnaBase website of the hospital. Our findings indicate 49 phenotypes with AUC greater than 0.6, predominantly linked to endocrine and metabolic diseases. Notably, hyperplasia of the prostate exhibited the highest disease prediction ability (P value = 1.01 × 10−19, AUC = 0.874), highlighting the potential of these polygenic scores in preventive medicine and diagnosis. This study offers a comprehensive evaluation of polygenic scores performance across diverse human traits, identifying promising applications for precision medicine and personalized healthcare, thereby inspiring further research and development in this field.

Menarche-a journey into womanhood: age at menarche and health-related outcomes in East Asians.

Are there associations of age at menarche (AAM) with health-related outcomes in East Asians? AAM is associated with osteoporosis, Type 2 diabetes (T2D), glaucoma, and uterine fibroids, as demonstrated through observational studies, polygenic risk scores, genetic correlations, and Mendelian randomization (MR), with additional findings indicating a causal effect of BMI and T2D on earlier AAM. Puberty timing is linked to adult disease risk, but research predominantly focuses on European populations, with limited studies in other groups. We performed an AAM genome-wide association study (GWAS) with 57890 Han Taiwanese females and examined the association between AAM and 154 disease outcomes using the Taiwanese database. Additionally, we examined genetic correlations between AAM and 113 diseases and 67 phenotypes using Japanese GWAS summary statistics. We performed AAM GWAS and gene-based GWAS studies to obtain summary statistics and identify potential AAM-related genes. We applied phenotype, polygenic risk scores, and genetic correlation analyses of AAM to explore health-related outcomes, using multivariate regression and linkage disequilibrium score regression analyses. We also explored potential bidirectional causal relationships between AAM and related outcomes through univariable and multivariable MR analyses. Fifteen lead single-nucleotide polymorphisms and 24 distinct genes were associated with AAM in Taiwan. AAM was genetically associated with later menarche and menopause, greater height, increased osteoporosis risk, but lower BMI, and reduced risks of T2D, glaucoma, and uterine fibroids in East Asians. Bidirectional MR analyses indicated that higher BMI/T2D causally leads to earlier AAM. Our findings were specific to Han Taiwanese individuals, with genetic correlation analyses conducted in East Asians. Further research in other ethnic groups is necessary. Our study provides insights into the genetic architecture of AAM and its health-related outcomes in East Asians, highlighting causal links between BMI/T2D and earlier AAM, which may suggest potential prevention strategies for early puberty. The work was supported by China Medical University, Taiwan (CMU110-S-17, CMU110-S-24, CMU110-MF-49, CMU111-SR-158, CMU111-MF-105, CMU111-MF-21, CMU111-S-35, CMU112-SR-30, and CMU112-MF-101), the China Medical University Hospital, Taiwan (DMR-111-062, DMR-111-153, DMR-112-042, DMR-113-038, and DMR-113-103), and the Ministry of Science and Technology, Taiwan (MOST 111-2314-B-039-063-MY3, MOST 111-2314-B-039-064-MY3, MOST 111-2410-H-039-002-MY3, and NSTC 112-2813-C-039-036-B). The funders had no influence on the data collection, analyses, or conclusions of the study. No conflict of interests to declare. N/A.

Coverage for the vascular pedicle of a free fasciocutaneous flap using split thickness skin graft with auxiliary procedure to achieve a safe method with no impairment of vessels: a case series and literature review

BackgroundThe vascular pedicle of a free flap is the most critical structure that determines its viability. Most of the times it is covered with local skin flaps raised from the recipient site or with part of the free flap itself. However, there are conditions in which the vascular pedicle can be covered with skin graft. The purpose of the present study is to describe our experience in the use of split-thickness skin grafts (STSG) as an auxiliary procedure for pedicle coverage.MethodsAll patients who underwent microvascular fasciocutaneous free flap reconstruction at the Department of Plastic Surgery of China Medical University Hospital in Taichung from 1986 to 2021 were retrospectively evaluated. Patients who met all of the following criteria were eligible for the study: microvascular free flap reconstruction of any region of the body with a fasciocutaneous flap and cases where tension was detected during skin closure over the pedicle of the flap and STSG was applied as a cover.ResultsThere were 14 cases in this series treated from 1986 to 2021. Among them, 11 cases had no additional skin at the proximal end of the free flap, nor local flap at disposal in the recipient site to cover the vessels. In 3 other cases a vascular bridge flap was used for cross-leg flap transfer without possibility of tension free tubulization to protect the vessels. All reconstruction were successful.ConclusionsDuring microvascular transfer of free flaps, if no skin flap is available to cover the vascular pedicle, skin graft can be used to protect the vessels without compromising the circulation of the flap. Our results, in accordance with the literature, supports the safety of this technique when direct closure of the skin above and near the pedicle is not possible.Level of evidenceLevel IV, Therapeutic.

A274 COVID-19 INFECTION RISK BETWEEN VACCINATED PATIENTS WITH ULCERATIVE COLITIS AND CROHN’S DISEASE : A RETROSPECTIVE COHORT STUDY IN TAIWAN

Abstract Background The effectiveness of coronavirus disease 2019 (COVID-19) vaccine against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in patients with inflammatory bowel disease (IBD) is well established, but it is not clear whether patients with ulcerative colitis (UC) and Crohn’s disease (CD) have different risks of COVID-19 infection after vaccination. Aims The aim of this study was to compare the risk of COVID-19 infection after vaccination in patients with UC and CD. Moreover, we also investigated risk factors related to COVID-19 infection and collected data on any adverse events and disease symptoms. Methods We conducted a retrospective cohort study in adult IBD patients who had received at least two doses of COVID-19 vaccination and compared the prevalence of infection between UC patients and CD patients using the medical records of China Medical University Hospital between 1 January 2020 and 31 March 2023. All data management and odds ratios calculations were performed using version 9.4 of SAS software (SAS Institute, Cary, NC, USA). Results Total 169 IBD patients (96 with UC, 73 with CD) were included in this study. UC patients were older than CD patients (44.92±13.72 vs. 37.27±15.27 years, p=0.0008)(Table 1). A high proportion were male (IBD 70.41%; UC 65.63%, and CD 76.71%). Most (57.4%) received three doses of COVID-19 vaccines. Azathioprine, steroid, and mesalazine medication histories were different between the groups (pampersand:003C0.05). COVID-19 infection prevalence was 49.11% (UC 56.25%, CD 39.73%; p=0.0333) (Figure 1). Logistic regression analysis suggested UC patients had an odds ratio of 1.95 (95% CI=1.05-3.62) and adjusted odds ratio of 2.78 (95% CI=1.20-6.44) for COVID-19 infection compared to CD patients, indicating a 1.95-fold to 2.78-fold higher risk (Figure 2). There was a trend towards a decreasing risk of COVID-19 infection with increasing number of vaccine doses. Medication was not identified as a risk factor. (Figure 3) Conclusions Our study identified a greater risk of COVID-19 infection in vaccinated patients with UC than those with CD. However, most patients who became infected with COVID-19 did not have severe symptoms and only one case was hospitalized. The mortality rate of COVID-19 in IBD patients is not specifically higher than the general population. Supporting the view that SARS-CoV-2 vaccination is effective in patients with IBD. Odds ratios (95% confidence interval) for logistic regression model predicting COVID-19 infection among vaccinated IBD patients Figure 1. The medications used before COVID-19 infection between vaccinated patients with ulcerative colitis (UC) and Crohn’s disease (CD). Figure2. The prevalence of COVID-19 infection between vaccinated patients with ulcerative colitis (UC) and Crohn's disease (CD). Figure 3. The symptoms after COVID-19 infection between vaccinated patients with ulcerative colitis (UC) and Crohn's disease (CD). Funding Agencies None

P264 The utility of ANCA in Ulcerative Colitis and Crohn’s Disease: A Retrospective Cohort Study in Taiwan

Abstract Background Inflammatory bowel disease (IBD) includes two chronic idiopathic inflammatory diseases: ulcerative colitis (UC) and Crohn disease (CD). The diagnosis of IBD depends on clinical, endoscopic, histological, radiological and biochemical criteria, which may be invasive, time consuming and usually not accepted by patients with IBD. Antineutrophil cytoplasmic antibodies (ANCAs) is believed to be related to IBD. The aim of this study was to investigate these serological markers and the evidence for their use in the diagnosis and management of IBD in Taiwan. Methods We conducted a retrospective cohort study in adult IBD patients who had received the serum examinations using the medical records of China Medical University Hospital between 1 January 1980 and 31 October. 2023. Results A total of 301 IBD patients (194 with UC, 107 with CD) were included in this current study. A high proportion were male (IBD 70.41%; UC 65.63%, and CD 76.71%). The mean diagnostic age of these enrolled patients was 42.9 years. In our patients, the prevalence of positive anti-HCV Ab is 0.0% in UC patients and 0.96% in CD patients. However, the prevalence of positive HBsAg was 14.5% in UC patients and 9.4% in CD patients. The prevalence of pANCA was 21.2% in UC patients and 1.4% in CD patients, respectively. The prevalence of cANCA was 3.1% in UC patients and 0.0% in CD patients. This pattern of low sensitivity and high specificity for pANCA is also seen in UC patients in various cohort studies. The serum positivity of p-ANCA was significantly higher in UC patients (Figure 1). UC patients with positive p-ANCA were older than CD patients with positive p-ANCA (44.9 vs. 37.3 years). Furthermore, we found that the incidence of positive p-ANCA in UC patients with E1, E2, and E3 at diagnosis was 3.5%, 39.2%, and 57.1%, respectively (Table 1). The serum levels of ANCA-IgG were indeed higher in patients in the severe group than those in the moderate and mild groups. Conclusion Serological biomarkers have been demonstrated to be a series of rapid, non-invasive approaches for assessments of early diagnosis, disease activity and prognosis for IBD. ANCA may be helpful in the early diagnosis of UC and in differentiating it from CD. UC patients with positive ANCA have a higher rate of intestinal mucosal vasculitis than UC patients with negative ANCA. Furthermore, ANCA may also contributes to the pathogenesis of HBV-related systemic vasculitis. Therefore, the management of HBV-related vasculitis includes control the immune complex formation and reaction as well as antiviral agents to reduce the antigenic load resulting in reduction of inflammation. Moreover, effective biomarkers with high sensitivity and specificity need to be investigated in the future.

A Computer-Assisted Diagnostic Method for Accurate Detection of Early Nondisplaced Fractures of the Femoral Neck.

Nondisplaced femoral neck fractures are sometimes misdiagnosed by radiographs, which may deteriorate into displaced fractures. However, few efficient artificial intelligent methods have been reported. We developed an automatic detection method using deep learning networks to pinpoint femoral neck fractures on radiographs to assist physicians in making an accurate diagnosis in the first place. Our proposed accurate automatic detection method, called the direction-aware fracture-detection network (DAFDNet), consists of two steps, namely region-of-interest (ROI) segmentation and fracture detection. The first step removes the noise region and pinpoints the femoral neck region. The fracture-detection step uses a direction-aware deep learning algorithm to mark the exact femoral neck fracture location in the region detected in the first step. A total of 3840 femoral neck parts in anterior-posterior (AP) pelvis radiographs collected from the China Medical University Hospital database were used to test our method. The simulation results showed that DAFDNet outperformed the U-Net and DenseNet methods in terms of the IOU value, Dice value, and Jaccard value. Our proposed DAFDNet demonstrated over 94.8% accuracy in differentiating non-displaced Garden type I and type II femoral neck fracture cases. Our DAFDNet method outperformed the diagnostic accuracy of general practitioners and orthopedic surgeons in accurately locating Garden type I and type II fracture locations. This study can determine the feasibility of applying artificial intelligence in a clinical setting and how the use of deep learning networks assists physicians in improving correct diagnoses compared to the current traditional orthopedic manual assessments.

Effectiveness of Chinese herbal medicine compared with angiotensin II receptor blockers in patients with diabetic kidney disease: A hospital-based matched cohort study

Angiotensin II receptor blockers (ARBs) are one of the standard treatments for diabetic kidney disease (DKD). Some patients may opt for Chinese herbal medicine (CHM) of their own free will. However, there is no real-world evidence regarding the effectiveness and safety of CHM. We aimed to explore the effectiveness of CHM for DKD in comparison to ARBs. We enrolled 732 DKD patients (72 used only CHM and 661 used ARBs) from 2007 to 2016, and all patients were followed until December 2016 at China Medical University Hospital in Taiwan. A total of 355 ARB users and 71 CHM users were analyzed after propensity score matching. The estimated glomerular filtration rate (eGFR) after treatment was 84.9 ± 28.1 ml/min/1.73 m2 in CHM users, which was higher than that (67.8 ± 35.4 ml/min/1.73 m2) in ARB users (p < 0.001). The change in the eGFR was −6.0 ± 21.4 ml/min/1.73 m2 in CHM users and −12.9 ± 24.8 ml/min/1.73 m2 in ARB users (p = 0.029). The blood urea nitrogen (BUN) and creatinine levels of patients taking CHM were 22 ± 16 mg/dl and 0.9 ± 0.4 mg/dl, respectively, and were lower than those (30 ± 28 mg/dl and 1.7 ± 2.0 mg/dl) of patients taking ARBs (p = 0.025 and p = 0.003). Using linear regression with adjustments for age, sex, BMI, baseline eGFR, and HbA1c levels, we found that the declines in the eGFR/baseline eGFR and changes in the urine albumin–creatinine ratio (ACR) were comparable between the two groups (p = 0.86 and 0.73). This study suggests that CHM may have comparable effectiveness to ARBs, which provides insights for further investigations.

The largest genome-wide association study for breast cancer in Taiwanese Han population.

Breast cancer is a molecularly heterogeneous disease, and multiple genetic variants contribute to its development and prognosis. Most of previous genome-wide association studies (GWASs) and polygenic risk scores (PRSs) analyses focused on studying breast cancers of Caucasian populations, which may not be applicable to other population. Therefore, we conducted the largest breast cancer cohort of Taiwanese population to fill in the knowledge gap. A total of 152,534 Participants recruited by China Medical University Hospital between 2003 and 2019 were filtered by several patient selection criteria and GWAS quality control steps, resulting in the inclusion of 2496 cases and 9984 controls for this study. We then conducted GWAS for all breast cancers and PRS analyses for all breast cancers and the four breast cancer subtypes, including luminal A, luminal B, basal-like, and HER2-enriched. The GWAS analyses identified 113 SNPs, 50 of which were novel. The PRS models for all breast cancers and the luminal A subtype showed positively correlated trends between the PRS and the risk of developing breast cancer. The odds ratios (95% confidence intervals) for the groups with the highest PRS in all breast cancers and the luminal A subtype were 5.33 (3.79-7.66) and 3.55 (2.13-6.14), respectively. In summary, we explored the association of genetic variants with breast cancer in the largest Taiwanese cohort and developed two PRS models that can predict the risk of developing any breast cancer and the luminal A subtype in Taiwanese women.

Unlocking Colchicine's Untapped Potential: A Paradigm Shift in Hepatocellular Carcinoma Prevention.

Background: Liver cancer and notably hepatocellular carcinoma (HCC), results in significantly high mortality rates worldwide. Chronic hepatitis and fatty liver, recognized precursors, underscore the imperative need for effective preventive strategies. This study explores colchicine, traditionally acknowledged for its anti-inflammatory properties and investigates its potential in liver cancer prevention. Methods: Utilizing the iHi Data Platform of China Medical University Hospital, Taiwan, this study analyzed two decades of medical data, incorporating 10,353 patients each in the Colchicine and Non-Colchicine cohorts, to investigate the association between colchicine use and liver cancer risk. Results: The study identified that colchicine users exhibited a 19% reduction in liver cancer risk, with a multivariable-adjusted odds ratio of 0.81 after accounting for confounding variables. Additionally, the influence of gender and comorbidities like diabetes mellitus on liver cancer risk was identified, corroborating the existing literature. A notable finding was that the prolonged use of colchicine was associated with improved outcomes, indicating a potential dose-response relationship. Conclusions: This study proposes a potential new role for colchicine in liver cancer prevention, extending beyond its established anti-inflammatory applications. While the findings are promising, further research is essential to validate these results. This research may serve as a foundation for future studies, aiming to further explore colchicine's role via clinical trials and in-depth investigations, potentially impacting preventive strategies for liver cancer.

Shielding analysis of the proton therapy facility at China Medical University Hospital: comparison of a simplified approach with Monte Carlo simulations.

This study comprehensively compared two approaches for analyzing the shielding design of the proton therapy facility at China Medical University Hospital. The first approach essentially involved two approximate models: one for estimating the transmitted radiation through thick shields, and one for estimating radiation streaming at locations near a maze entrance. The second approach relied on Monte Carlo simulations for predicting the radiation field in a complex environment. A total of 22 beam loss scenarios were considered, and dose rates at 32 locations around the facility were estimated using the two approaches. The comparison results demonstrated that the simplified approach proposed in this study can yield fairly accurate or conservative estimates for quickly performing shielding design or dose assessment in a real-world proton therapy facility.

Increased intraoperative vein ligation in microsurgical varicocelectomy is associated with pain improvement.

Varicocele is a major cause of male infertility. However, few studies have discussed the potential associations between the pain caused by varicocele and preoperative and intraoperative factors. The aim of this study was to evaluate factors potentially associated with changes in pain score after microsurgical varicocelectomy. This retrospective study was conducted between August 2020 and August 2022 at China Medical University Hospital in Taichung, Taiwan. Patient characteristics including age, body mass index, semen analysis, testicular volume, and the number of veins ligated were collected. Preoperative and intraoperative factors were analyzed to determine if they were correlated with changes in numeric rating scale (NRS) after microsurgical varicocelectomy. A total of 44 patients with clinical varicocele underwent subinguinal microsurgical varicocelectomy and were analyzed. The overall pain resolution rate was 91%, and the average satisfaction score after surgery was 9.2 according to their subjective feelings. Multivariate analysis revealed that severe varicocele grade (odds ratio [OR] 16.5, 95% confidence interval [CI] 3.01-90.47; P = .018) and the number of veins ligated (OR 6, 95% CI 1.6-22.48; P = .013), were significantly associated with changes in NRS after surgery. In addition, the area under the receiver operating characteristic curve for changes in NRS and the total number of veins ligated was 0.869. Microsurgical varicocelectomy had a high success rate for scrotal pain and satisfaction. Severe varicocele grade and the number of veins ligated in microsurgical varicocelectomy were associated with postoperative pain improvement.

Correlation between impulse oscillometry parameters and test for respiratory and asthma control in kids (TRACK) in asthma control of preschoolers with asthma

Little research has been conducted to evaluate the correlation between impulse oscillometry (IOS), Childhood Asthma Control Test (C-ACT), and Test for Respiratory and Asthma Control in Kids (TRACK). This study was conducted at China Medical University Hospital between September 1, 2019, and March 31, 2021. Children aged 2-6 years who had been diagnosed with asthma with acute exacerbation were enrolled and followed-up until the end of the study. Correlations between the parameters of IOS, C-ACT and TRACK were assessed. The validity and reliability of TRACK were verified. A total of 114 children with asthma and acute exacerbations were recruited. Their mean age was 4.1±1.1 years, and 60.5% were males. After a year of treatment, the change of R5-R20 from baseline 0.64±0.38kPa/L/s to 12th month 0.48±0.2kPa/L/s (p=0.022). TRACK and C-ACT scores were significantly correlated during the observation period. R5-R20 in IOS at baseline and at the 12th month of follow-up as well as the change in IOS parameters were significantly associated with C-ACT (p=0.003, 0.015, and 0.001, respectively). R5% and R5-R20 changes in IOS were associated with TRACK (p=0.04 and 0.025, respectively). Sensitivity and specificity of TRACK were 80.8% (67.5-90.4) and 100% (94.1-100), respectively, with cut-off points >95 and AUC 93.8%. TRACK score appears to have a stronger association with the IOS parameter than C-ACT score. Our findings indicate that TRACK is a valid tool for assessing asthma control in preschool children.

A Combination of FPU-Net and Feature Clustering Methods for Accurate Segmentation of Femoral Neck in Radiographic Diagnosis.

In this study, we develop an innovative method that assists computer-aided diagnosis in the determination process of the exact location of the femoral neck junction in plain radiographs. Our algorithm consists of two phases, i.e., coarse prediction and fine matching, which are implemented by supervised deep learning method and unsupervised clustering, respectively. In coarse prediction, standard masks are first produced by a specialist and trained in our proposed feature propagation network (FPU-Net) with supervised learning on the femoral neck dataset. In fine matching, the standard masks are first classified into different categories using our proposed three parameters with unsupervised learning. The predicted mask from FPU-Net is matched with each category of standard masks by calculating the values of intersection of union (IOU), and finally the predicted mask is substituted by the standard mask with the largest IOU value. A total of 4320 femoral neck parts in anterior-posterior (AP) pelvis radiographs collected from China Medical University Hospital database were used to test our method. Simulation results show that, on the one hand, compared with other segmentation methods, the method proposed in this paper has a larger IOU value and better suppression of noise outside the region of interest; on the other hand, the introduction of unsupervised learning for fine matching can help in the accurate localization segmentation of femoral neck images. Accurate femoral neck segmentation can assist surgeons to diagnose and reduce the misdiagnosis rate and burden.

Machine learning of cell population data, complete blood count, and differential count parameters for early prediction of bacteremia among adult patients with suspected bacterial infections and blood culture sampling in emergency departments.

Bacteremia is a life-threatening complication of infectious diseases. Bacteremia can be predicted using machine learning (ML) models, but these models have not utilized cell population data (CPD). The derivation cohort from emergency department (ED) of China Medical University Hospital (CMUH) was used to develop the model and was prospectively validated in the same hospital. External validation was performed using cohorts from ED of Wei-Gong Memorial Hospital (WMH) and Tainan Municipal An-Nan Hospital (ANH). Adult patients who underwent complete blood count (CBC), differential count (DC), and blood culture tests were enrolled in the present study. The ML model was developed using CBC, DC, and CPD to predict bacteremia from positive blood cultures obtained within 4h before or after the acquisition of CBC/DC blood samples. This study included 20,636 patients from CMUH, 664 from WMH, and 1622 patients from ANH. Another 3143 patients were included in the prospective validation cohort of CMUH. The CatBoost model achieved an area under the receiver operating characteristic curve of 0.844 in the derivation cross-validation, 0.812 in the prospective validation, 0.844 in the WMH external validation, and 0.847 in the ANH external validation. The most valuable predictors of bacteremia in the CatBoost model were the mean conductivity of lymphocytes, nucleated red blood cell count, mean conductivity of monocytes, and neutrophil-to-lymphocyte ratio. ML model that incorporated CBC, DC, and CPD showed excellent performance in predicting bacteremia among adult patients with suspected bacterial infections and blood culture sampling in emergency departments.

Effects of Human Leukocyte Antigen DRB1 Genetic Polymorphism on Anti-Cyclic Citrullinated Peptide (ANTI-CCP) and Rheumatoid Factor (RF) Expression in Rheumatoid Arthritis (RA) Patients.

Rheumatoid arthritis (RA) is a systemic disease characterized by non-infectious inflammation of the joints and surrounding tissues, which can cause severe health problems, affect the patient's daily life, and even cause death. RA can be clinically diagnosed by the occurrence of blood serological markers, rheumatoid factor (RF) and anti-cyclic citrullinated peptide antibody (anti-CCP). However, about 20% of RA patients exhibit negative results for both markers, which makes RA diagnosis difficult and, therefore, may delay the effective treatment. Previous studies found some evidence that human leukocyte antigen (HLA)-related genes might be the susceptibility genes for RA and their polymorphisms might contribute to varieties of susceptibility and disease severity. This study aimed for the genetic polymorphisms of the RA patient genome and their effects on the RA patient's serological makers, RF and anti-CCP. A total of 4580 patients' electronic medical records from 1992 to 2020 were retrieved from the China Medical University Hospital database. The most representative single-nucleotide polymorphisms (SNPs) were identified through a genome-wide association study (GWAS) followed by enzyme-linked immunosorbent assay (ELISA) validation using the blood from 30 additional RA patients. The results showed significant changes at the position of chromosome 6 with rs9270481 being the most significant locus, which indicated the location of the HLA-DRB1 gene. Further, patients with the CC genotype at this locus were more likely to exhibit negative results for RF and anti-CCP than those with the TT genotype. The C allele was also more likely to be associated with negative results for RF and anti-CCP. The results demonstrated that a genetic polymorphism at rs9270481 affected the expression of RF and anti-CCP in RA patients, which might indicate the necessity to develop a personalized treatment plan for each individual patient based on the genetic profile.

Aspirin Use and the Risk of Hepatocellular Carcinoma: A Meta-analysis.

*College of Medicine, Tzu Chi University, Hualien, Taiwan †Division of Hepatogastroenterology, Department of Internal Medicine, Taichung Tzu Chi Hospital, Taichung, Taiwan ‡Department of Public Health, College of Public Health §Department of Medicine, College of Medicine, China Medical University, Taichung, Taiwan ∥Department of Family Medicine, China Medical University Hospital, Taichung, Taiwan The authors declare that they have nothing to disclose. Address correspondence to: Shih-Wei Lai, MD, Department of Family Medicine, China Medical University Hospital, No 2, Yu-De Road, Taichung City 404, Taiwan (e-mail: [email protected]).

Efficacy of Warfarin Therapy Guided by Pharmacogenetics: A Real-world Investigation Among Han Taiwanese

The anticoagulation activity of warfarin in populations with CYP2C9, VKORC1, and CYP4F2 variants differs between individuals and is correlated with poor international normalized ratio (INR) control. Pharmacogenetics-guided warfarin dosing has been successfully developed for patients with genetic variations in recent years. However, few real-world data have been used to investigate the INR and warfarin dosage and the time to target INR. This study examined the largest collection of genetic and clinical real-world data related to warfarin to provide further evidence supporting the benefits of pharmacogenetics in clinical outcomes. We retrieved a total of 69,610 INR-warfarin records after the index date from 2,613 patients in the China Medical University Hospital database between January 2003 and December 2019. Each INR reading was obtained from the latest laboratory data after the hospital visit date. Patients with a history of malignant neoplasms or pregnancy before the index date were excluded, as were patients without data on INR measurements after the fifth day of prescription, genetic information, or gender variables. The primary outcomes were the INR and warfarin dosage during days 7, 14, 28, 56, and 84 after prescription. The secondary outcome was the time required to reach the INR ranges of 1.5 to 3.0 and >4.0. A total of 59,643 INR-warfarin records from 2188 patients were retrieved. The average INR was higher for homozygous carriers of the minor allele at CYP2C9 and VKORC1 during the first 7 days (1.83 [1.03] [CYP2C9*1] and 2.46 [1.44] [CYP2C9*3], P < 0.001; 1.39 [0.36] [rs9923231 G/G], 1.55 [0.79] [rs9923231 G/A], and 1.96 [1.13] [rs9923231 A/A], P < 0.001) than for the wild-type allele. These patients with variants required lower warfarin doses than those with the wild-type allele during the first 28 days. CYP4F2 variant patients seemed to require higher doses of warfarin than those in the wild-type group; however, no significant difference in the average INR was observed (1.95 [1.14] [homozygous V433 carriers], 1.78 [0.98] [heterozygous V433M carriers], and 1.66 [0.91] [homozygous M433 carriers], P=0.016). Our study indicates that genetic variants in the Han population may enhance warfarin responsiveness, which holds clinical relevance. An increased warfarin dosage was not linked to a shorter time to therapeutic INR between CYP4F2 variant patients and those with a wild-type allele. Assessing CYP2C9 and VKORC1 genetic polymorphisms before initiating warfarin treatment in real-world practice is essential for potentially vulnerable patients and is likely to optimize therapeutic dosing.