Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 163
Filtrar
1.
BMC Med Imaging ; 21(1): 142, 2021 10 02.
Artigo em Inglês | MEDLINE | ID: mdl-34600486

RESUMO

BACKGROUND: Automated language analysis of radiology reports using natural language processing (NLP) can provide valuable information on patients' health and disease. With its rapid development, NLP studies should have transparent methodology to allow comparison of approaches and reproducibility. This systematic review aims to summarise the characteristics and reporting quality of studies applying NLP to radiology reports. METHODS: We searched Google Scholar for studies published in English that applied NLP to radiology reports of any imaging modality between January 2015 and October 2019. At least two reviewers independently performed screening and completed data extraction. We specified 15 criteria relating to data source, datasets, ground truth, outcomes, and reproducibility for quality assessment. The primary NLP performance measures were precision, recall and F1 score. RESULTS: Of the 4,836 records retrieved, we included 164 studies that used NLP on radiology reports. The commonest clinical applications of NLP were disease information or classification (28%) and diagnostic surveillance (27.4%). Most studies used English radiology reports (86%). Reports from mixed imaging modalities were used in 28% of the studies. Oncology (24%) was the most frequent disease area. Most studies had dataset size > 200 (85.4%) but the proportion of studies that described their annotated, training, validation, and test set were 67.1%, 63.4%, 45.7%, and 67.7% respectively. About half of the studies reported precision (48.8%) and recall (53.7%). Few studies reported external validation performed (10.8%), data availability (8.5%) and code availability (9.1%). There was no pattern of performance associated with the overall reporting quality. CONCLUSIONS: There is a range of potential clinical applications for NLP of radiology reports in health services and research. However, we found suboptimal reporting quality that precludes comparison, reproducibility, and replication. Our results support the need for development of reporting standards specific to clinical NLP studies.

2.
BMC Public Health ; 21(1): 1885, 2021 Oct 19.
Artigo em Inglês | MEDLINE | ID: mdl-34663285

RESUMO

BACKGROUND: China has introduced a series of stricter policies to criminalize drunk driving and increase penalties since May 2011. However, there is no previous study examining the time-varying impacts of drunk driving regulations on road traffic fatalities based on daily data. METHODS: We collected 6536 individual data of road traffic deaths (RTDs) in Guangzhou from 2008 to 2018. The quasi-Poisson regression models with an inclusion of the intervention variable and the interaction of intervention variable and a function of time were used to quantify the time-varying effects of these regulations. RESULTS: During the 11-year study period, the number of population and motor vehicles showed a steady upward trend. However, the population- and motor vehicles- standardized RTDs rose steadily before May 2011, the criminalizing drunk driving intervention was implemented and gradually declined after that. The new drunk driving intervention were associated with an average risk reduction of RTDs (ER = -9.01, 95% eCI: - 10.05% to - 7.62%) during the 7.7 years after May 2011. On average, 75.82 (95% eCI, 54.06 to 92.04) RTDs per 1 million population annually were prevented due to the drunk driving intervention. CONCLUSION: These findings would provide important implications for the development of integrated intervention measures in China and other countries attempting to reduce traffic fatalities by stricter regulations on drunk driving.

3.
Commun Biol ; 4(1): 1019, 2021 08 31.
Artigo em Inglês | MEDLINE | ID: mdl-34465850

RESUMO

Despite the uniform mortality in pancreatic adenocarcinoma (PDAC), clinical disease heterogeneity exists with limited genomic differences. A highly aggressive tumor subtype termed 'basal-like' was identified to show worse outcomes and higher inflammatory responses. Here, we focus on the microbial effect in PDAC progression and present a comprehensive analysis of the tumor microbiome in different PDAC subtypes with resectable tumors using metagenomic sequencing. We found distinctive microbial communities in basal-like tumors and identified an increasing abundance of Acinetobacter, Pseudomonas and Sphingopyxis to be highly associated with carcinogenesis. Functional characterization of microbial genes suggested the potential to induce pathogen-related inflammation. Host-microbiota interplay analysis provided new insights into the tumorigenic role of specific microbiome compositions and demonstrated the influence of host genetics in shaping the tumor microbiome. Taken together, these findings indicated that the tumor microbiome is closely related to PDAC oncogenesis and the induction of inflammation. Additionally, our data revealed the microbial basis of PDAC heterogeneity and proved the predictive value of the microbiome, which will contribute to the intervention and treatment of disease.

4.
BMJ Open ; 11(9): e049581, 2021 09 06.
Artigo em Inglês | MEDLINE | ID: mdl-34489283

RESUMO

OBJECTIVES: To evaluate the cost-effectiveness of four different primary screening strategies: high-risk factor questionnaire (HRFQ) alone, single immunochemical faecal occult blood test (iFOBT), double iFOBT and HRFQ+double iFOBT for colorectal cancer (CRC) screening compared with no screening using the Markov model. METHODS: Treeage Pro V.2011 software was used to simulate the Markov model. The incremental cost-effectiveness ratio, which was compared with the willingness-to-pay (WTP) threshold, was used to reflect the cost-effectiveness of the CRC screening method. One-way sensitivity analysis and probabilistic sensitivity analysis were used for parameter uncertainty. RESULTS: All strategies had greater effectiveness because they had more quality-adjusted life years (QALYs) than no screening. When the WTP was ¥435 762/QALY, all screening strategies were cost-effective compared with no screening. The double iFOBT strategy was the best-buy option compared with all other strategies because it had the most QALYs and the least cost. One-way sensitivity analysis showed that the sensitivity of low-risk adenoma, compliance with colonoscopy and primary screening cost were the main influencing factors comparing single iFOBT, double iFOBT and HRFQ+double iFOBT with no screening. However, within the scope of this study, there was no fundamental impact on cost-effectiveness. Probabilistic sensitivity analysis showed that when the WTP was ¥435 762/QALY, the probabilities of the cost-effectiveness acceptability curve with HRFQ alone, single iFOBT, double iFOBT and HRFQ+double iFOBT were 0.0%, 5.3%, 69.3% and 25.4%, respectively. CONCLUSIONS: All screening strategies for CRC were cost-effective compared with no screening strategy. Double iFOBT was the best-buy option compared with all other strategies. The significant influencing factors were the sensitivity of low-risk polyps, compliance with colonoscopy and cost of primary screening.


Assuntos
Neoplasias Colorretais , Detecção Precoce de Câncer , China , Colonoscopia , Neoplasias Colorretais/diagnóstico , Análise Custo-Benefício , Humanos , Cadeias de Markov , Programas de Rastreamento , Sangue Oculto , Anos de Vida Ajustados por Qualidade de Vida
5.
Emerg Microbes Infect ; 10(1): 1751-1759, 2021 Dec.
Artigo em Inglês | MEDLINE | ID: mdl-34396940

RESUMO

The effectiveness of inactivated SARS-CoV-2 vaccines against the Delta variant, which has been associated with greater transmissibility and virulence, remains unclear. We conducted a test-negative case-control study to explore the vaccine effectiveness (VE) in real-world settings. We recruited participants aged 18-59 years who consisted of SARS-CoV-2 test-positive cases (n = 74) and test-negative controls (n = 292) during the outbreak of the Delta variant in May 2021 in Guangzhou city, China. Vaccination status was compared to estimate The VE of SARS-CoV-2 inactivated vaccines. A single dose of inactivated SARS-CoV-2 vaccine yielded the VE of only 13.8%. After adjusting for age and sex, the overall VE for two-dose vaccination was 59.0% (95% confidence interval: 16.0% to 81.6%) against coronavirus disease 2019 (COVID-19) and 70.2% (95% confidence interval: 29.6-89.3%) against moderate COVID-19 and 100% against severe COVID-19 which might be overestimated due to the small sample size. The VE of two-dose vaccination against COVID-19 reached 72.5% among participants aged 40-59 years, and was higher in females than in males against COVID-19 and moderate diseases. While single dose vaccination was not sufficiently protective, the two-dose dosing scheme of the inactivated vaccines was effective against the Delta variant infection in real-world settings, with the estimated efficacy exceeding the World Health Organization minimal threshold of 50%.


Assuntos
Vacinas contra COVID-19/normas , COVID-19/prevenção & controle , SARS-CoV-2/genética , Adolescente , Adulto , Distribuição por Idade , COVID-19/classificação , Vacinas contra COVID-19/administração & dosagem , Estudos de Casos e Controles , China , Surtos de Doenças , Feminino , Variação Genética , Humanos , Masculino , Pessoa de Meia-Idade , Vacinas de Produtos Inativados/administração & dosagem , Vacinas de Produtos Inativados/normas , Adulto Jovem
8.
Ecotoxicol Environ Saf ; 222: 112498, 2021 Oct 01.
Artigo em Inglês | MEDLINE | ID: mdl-34265527

RESUMO

As the major constituents of PM2.5, carbonaceous constituents and inorganic ions have attracted emerging attentions on their health risks, particularly on cardiorespiratory diseases. However, evidences on the risks of PM2.5 constituents on other diseases (eg. nervous disease, genitourinary disease, neoplasms and endocrine disease) remain scarce. In our study, we firstly calculated residuals of PM2.5 constituents regressed on PM2.5 to remove the confounding effect of PM2.5. Then, generalized additive model (GAM) was used to assess impacts of residuals of PM2.5 constituents on mortality from 36 diseases (10 broad categories and 26 subcategories) during 2011-2015 in Guangzhou, China. Results of constituent-residual models showed that only EC, OC and NO3- were significantly associated with all-cause mortality, with per IQR change in corresponding constituent residuals related to percentage changes of 1.69% (95% CI: 0.42, 2.97), 1.94% (95% CI: 0.37, 3.54) and 2.59% (95% CI: 1.02, 4.18) at lag 03 days. All these pollutants were significantly associated with elevated mortality risk of cardiovascular disease, but only EC was significantly associated with respiratory mortality, and NO3- with endocrine disease and neoplasm. For more specific causes, the highest effect estimates of EC and NO3-were both observed on mortality from other form of heart disease, and OC on intentional self-harm, with estimates of 11.45% (95% CI: 2.74, 20.91), 12.59% (95% CI: 1.41, 25.02) and 18.01% (95% CI: 2.14, 36.36), respectively. Our findings highlighted that stricter emission control measures are still warranted to reduce air pollution level and protect the public health.


Assuntos
Poluentes Atmosféricos , Poluição do Ar , Doenças Cardiovasculares , Poluentes Atmosféricos/análise , Poluentes Atmosféricos/toxicidade , Poluição do Ar/efeitos adversos , Poluição do Ar/análise , Doenças Cardiovasculares/induzido quimicamente , China/epidemiologia , Humanos , Material Particulado/análise , Material Particulado/toxicidade
9.
Pharmaceutics ; 13(5)2021 May 19.
Artigo em Inglês | MEDLINE | ID: mdl-34069645

RESUMO

Inspired by molecular self-assembly, which is ubiquitous in natural environments and biological systems, self-assembled peptides have become a research hotspot in the biomedical field due to their inherent biocompatibility and biodegradability, properties that are afforded by the amide linkages forming the peptide backbone. This review summarizes the biological advantages, principles, and design strategies of self-assembled polypeptide systems. We then focus on the latest advances in in situ self-assembly of polypeptides in medical applications, such as oncotherapy, materials science, regenerative medicine, and drug delivery, and then briefly discuss their potential challenges in clinical treatment.

10.
BMC Med Inform Decis Mak ; 21(1): 179, 2021 06 03.
Artigo em Inglês | MEDLINE | ID: mdl-34082729

RESUMO

BACKGROUND: Natural language processing (NLP) has a significant role in advancing healthcare and has been found to be key in extracting structured information from radiology reports. Understanding recent developments in NLP application to radiology is of significance but recent reviews on this are limited. This study systematically assesses and quantifies recent literature in NLP applied to radiology reports. METHODS: We conduct an automated literature search yielding 4836 results using automated filtering, metadata enriching steps and citation search combined with manual review. Our analysis is based on 21 variables including radiology characteristics, NLP methodology, performance, study, and clinical application characteristics. RESULTS: We present a comprehensive analysis of the 164 publications retrieved with publications in 2019 almost triple those in 2015. Each publication is categorised into one of 6 clinical application categories. Deep learning use increases in the period but conventional machine learning approaches are still prevalent. Deep learning remains challenged when data is scarce and there is little evidence of adoption into clinical practice. Despite 17% of studies reporting greater than 0.85 F1 scores, it is hard to comparatively evaluate these approaches given that most of them use different datasets. Only 14 studies made their data and 15 their code available with 10 externally validating results. CONCLUSIONS: Automated understanding of clinical narratives of the radiology reports has the potential to enhance the healthcare process and we show that research in this field continues to grow. Reproducibility and explainability of models are important if the domain is to move applications into clinical use. More could be done to share code enabling validation of methods on different institutional data and to reduce heterogeneity in reporting of study properties allowing inter-study comparisons. Our results have significance for researchers in the field providing a systematic synthesis of existing work to build on, identify gaps, opportunities for collaboration and avoid duplication.


Assuntos
Sistemas de Informação em Radiologia , Radiologia , Humanos , Aprendizado de Máquina , Processamento de Linguagem Natural , Reprodutibilidade dos Testes
11.
Front Pharmacol ; 12: 603734, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34149403

RESUMO

Drynariae Rhizoma (DR) has been demonstrated to be effective in promoting fracture healing in clinical use. In the study, we tried to predicate potential signaling pathways and active ingredients of DR via network pharmacology, uncover its regulation mechanism to improve large bone defects by in vivo and in vitro experiment. We total discovered 18 potential active ingredients such as flavonoids and 81 corresponding targets, in which mitogen-activated protein kinase (MAPK) signaling pathway has the highest correlation with bone defects in pathway and functional enrichment analysis. Therefore, we hypothesized that flavonoids in DR improve large bone defects by activating MAPK signaling pathway. Animal experiments were carried out and all rats randomly divided into TFDR low, medium, and high dosage group, model group and control group. 12 weeks after treatment, according to X-ray and Micro-CT, TFDR medium dosage group significantly promote new bone mineralization compared with other groups. The results of HE and Masson staining and in vitro ALP level of BMSC also demonstrated the formation of bone matrix and mineralization in the TFDR groups. Also, angiographic imaging suggested that flavonoids in DR promoting angiogenesis in the defect area. Consistently, TFDR significantly enhanced the expression of BMP-2, RUNX-2, VEGF, HIF-1 in large bone defect rats based on ELISA and Real-Time PCR. Overall, we not only discover the active ingredients of DR in this study, but also explained how flavonoids in DR regulating MAPK signaling pathway to improve large bone defects.

12.
Pathobiology ; 88(4): 277-288, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-33882484

RESUMO

BACKGROUND: The dysregulation of proliferation and migration of vascular smooth muscle cells (VSMCs) is one of the major causes of atherosclerosis (AS). Accumulating studies confirm that Kruppel-like factor 4 (KLF4) can regulate the proliferation and differentiation of VSMCs through multiple signaling pathways. However, the mechanism of KLF4 dysregulation remains unknown. METHODS: Apolipoprotein E-knockout (ApoE-/-) mice and human VSMCs were used to establish AS animal model and cell model, respectively. qRT-PCR was employed to determine the expressions of miR-506-3p and KLF4. Cell Counting Kit -8, Transwell, TUNEL assays, and flow cytometry were performed to measure the proliferation, migration, and apoptosis of VSMCs. The upstream miRNAs of KLF4 were predicted by microT, miRanda, miRmap, and TargetScan databases. The interaction between KLF4 and miR-506-3p was confirmed using qRT-PCR, Western blot, and luciferase reporter gene assay. RESULTS: KLF4 expression was significantly decreased in the VSMCs of ApoE-/- mice fed with high-fat diet and in human VSMCs treated with oxidized low-density lipoprotein in time-dependent and dose-dependent manners. The transfection of miR-506-3p mimics or KLF4 shRNA promoted the proliferation and migration of VSMCs but inhibited the apoptosis while miR-506-3p inhibitors and pcDNA3.1-KLF4 exerted opposite effects. Additionally, KLF4 was confirmed as a target gene of miR-506-3p and could be negatively regulated by miR-506-3p. CONCLUSION: MiR-506-3p can promote the proliferation and migration of VSMCs via targeting KLF4, which can probably contribute to the pathogenesis of AS.

13.
Sichuan Da Xue Xue Bao Yi Xue Ban ; 52(2): 207-215, 2021 Mar.
Artigo em Chinês | MEDLINE | ID: mdl-33829693

RESUMO

Objective: To investigate whether long-term exposure to inhaled sevoflurane, a volatile anesthetic, causes abnormal activities and memory impairment related to attention-deficit/hyperactivity disorder (ADHD) in neonatal rats. Methods: On postnatal day 5 (P5), Sprague-Dawley rats were randomly assigned to two sevoflurane subgroups and two control subgroups and underwent experimental intervention. The two sevoflurane (SEVO) subgroups were exposed to 3% sevoflurane for 2 h and 4 h respectively, while the two control subgroups were given pure oxygen for the same amount and duration. Behavioral tests, including open-field test (OFT), five-choice serial reaction time task (5-CSRTT), fear-conditioning (FC) and Morris water maze (MWM), were applied to evaluate changes in cognition, memory, anxiety and ADHD-related behavioral changes in the rats in adolescence (-P25) and in adulthood (-P65). Results: In OFT, the SEVO 2 h and SEVO 4 h subgroups displayed activity level and exploratory behaviors similar to those of the control subgroups on P21 and P61, with no statistically significant difference identified in the data. 5-CSRTT results on P25 and P65 indicated no statistically significant difference between the SEVO subgroups and the control subgroups in regard to ADHD-related abnormal behaviors, including number of immature reaction, rate of correct response and omission rate. In the FC experiment, SEVO 4 h group had a shorter freezing period and longer period of freezing latency ( P=0.029) in comparison to the control groups. The results of the MWM test showed that the escape latency period of rats in the SEVO 4 h group was significantly prolonged on the second day and the third day, compared to the control groups ( P<0.05). The average swimming speed of SEVO groups did no exhibit any statistically significant difference on P69 or P76. The time the SEVO 4 h group spent in the target quadrant was significantly shorter than that of the control group ( P=0.039) and percentage of distance traveled in the target quadrant was significantly reduced compared to that the control group ( P=0.048). Conclusion: The findings suggest that four hours of inhaled sevoflurane exposure in neonate rats may cause memory impairment, but does no increase risks for ADHD-related abnormal activities.


Assuntos
Anestésicos Inalatórios , Transtorno do Deficit de Atenção com Hiperatividade , Anestésicos Inalatórios/toxicidade , Animais , Animais Recém-Nascidos , Transtorno do Deficit de Atenção com Hiperatividade/induzido quimicamente , Aprendizagem em Labirinto , Ratos , Ratos Sprague-Dawley , Sevoflurano
14.
Front Chem ; 9: 522708, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-33796503

RESUMO

Zirconia nanoparticles (ZrO2 NPs) are commonly used in the field of biomedical materials, but their antitumor activity and mechanism is unclear. Herein, we evaluated the anti-tumor activity of ZrO2 NPs and explored the anti-tumor mechanism. The results of in vitro and in vivo experiments showed that the level of intracellular reactive oxygen species (ROS) in HeLa cells was elevated after ZrO2 NPs treatment. Transmission electron microscopy (TEM) showed that after treatment with ZrO2 NPs, the mitochondria of HeLa cells were swollen, accompanied with the induction of autophagic vacuoles. In addition, flow cytometry analysis showed that the apoptotic rate of HeLa cells increased significantly by Annexin staining after treatment with ZrO2 NPs, and the mitochondrial membrane potential (MMP) was reduced significantly. The proliferation of HeLa cells decreased as indicated by reduced Ki-67 labeling. In contrast, TUNEL-positive cells in tumor tissues increased after treatment with ZrO2 NPs, which is accompanied by increased expression of mitochondrial apoptotic proteins including Bax, Caspase-3, Caspase-9, and Cytochrome C (Cyt C) and increased expression of autophagy-related proteins including Atg5, Atg12, Beclin-1, and LC3-II. Treating HeLa cells with N-acetyl-L-cysteine (NAC) significantly reduced ROS, rate of apoptosis, MMP, and in vivo anti-tumor activity. In addition, apoptosis- and autophagy-related protein expressions were also suppressed. Based on these observations, we conclude that ZrO2 NPs induce HeLa cell death through ROS mediated mitochondrial apoptosis and autophagy.

15.
Biomed Pharmacother ; 138: 111480, 2021 Jun.
Artigo em Inglês | MEDLINE | ID: mdl-33774316

RESUMO

Osteogenesis and angiogenesis acts as an essential role in repairing large tibial defects (LTDs). Total flavonoids of rhizoma drynariae (TFRD), a traditional Chinese medicinal herb, is reported to show anabolic effects on fracture healing. However, whether TFRD could improve the bone formation and angiogenesis in LTDs remains unknown. The purpose of this study was to evaluate the effect of TFRD on bone formation and angiogenesis in LTDs in distraction osteogenesis (DO). Using a previously established fracture model, LTD rats was established with circular external fixator (CEF). All rats then randomly divided into TFRD low dosage group (with DO), TFRD medium dosage group (with DO), TFRD high dosage group (with DO), model group (with DO) and blank group (without DO). Twelve weeks after treatment, according to X-ray and Micro-CT, TFRD groups (especially in medium dosage group) can significantly promote the formation of a large number of epiphyses and improve new bone mineralization compared with model group, and the results of HE and Masson staining and in vitro ALP level of BMSC also demonstrated the formation of bone matrix and mineralization in the TFRD groups. Also, angiographic imaging suggested that total flavonoids of TFRD was able to promote angiogenesis in the defect area. Consistently, TFRD significantly increased the levels of BMP-2, SMAD1, SMAD4, RUNX-2, OSX and VEGF in LTD rats based on ELISA and Real-Time PCR. In addition, we found that ALP activity of TFRD medium dosage group reached a peak after 10 days of induction through BMSC cell culture in vitro experiment. TFRD promoted bone formation in LTD through activation of BMP-Smad signaling pathway, which provides a promising new strategy for repairing bone defects in DO surgeries.


Assuntos
Densidade Óssea/fisiologia , Proteína Morfogenética Óssea 2/metabolismo , Flavonoides/farmacologia , Polypodiaceae , Proteínas Smad/metabolismo , Tíbia/metabolismo , Animais , Densidade Óssea/efeitos dos fármacos , Flavonoides/isolamento & purificação , Masculino , Ratos , Ratos Sprague-Dawley , Rizoma , Transdução de Sinais/efeitos dos fármacos , Transdução de Sinais/fisiologia , Tíbia/diagnóstico por imagem , Tíbia/efeitos dos fármacos , Microtomografia por Raio-X/métodos
16.
Cancer Med ; 10(8): 2865-2876, 2021 04.
Artigo em Inglês | MEDLINE | ID: mdl-33724715

RESUMO

BACKGROUND: Cancer is an important focus of public health worldwide. This study aims to provide a comprehensive overview of temporal trends in incidence and mortality of leading cancer in Guangzhou, China from 2004 to 2015. METHODS: Data were collected from the population-based registry in Guangzhou. Age-standardized incidence rate (ASIR) and age-standardized mortality rate (ASMR) were calculated and Joinpoint regression was used for evaluating the average annual percent changes (AAPC) among the entire study period and the estimated annual percent changes (EAPC) in time segments. The effects of age, period, and birth cohort were assessed by the age-period-cohort model. RESULTS: The age-standardized incidence and mortality by the world standard population decreased significantly among males with AAPC of -1.7% (95% CI: -3.0%, 0.2%) and -2.7% (95% CI: -4.3%, -1.1%) for all malignancies during 2004-2015, while among females, the age-standardized incidence had a non-significant reduction with AAPC of -1.3% (95% CI: -2.8%, 0.2%) and the age-standardized mortality demonstrated a remarkable decline (AAPC -2.0%, 95% CI: -3.6%, -0.3%). For males, the most commonly diagnosed cancers were trachea, bronchus, and lung (TBL), liver, colorectal, nasopharyngeal, stomach, and prostate cancer. For females, breast, TBL, colorectal, liver stomach, and thyroid cancer ranked the top. Unfavorable trends were observed in ASIR of colorectal, thyroid, and prostate cancer. APC models yielded different ages, periods, and birth cohort effect patterns by cancer sites. CONCLUSIONS: Cancer burden remained a public health challenge in Guangzhou as the aging population and lifestyles changes, despite declines in incidence and mortality rates in some cancers. Surveillance of cancer trends contributed to valuable insights into cancer prevention and control.


Assuntos
Neoplasias/epidemiologia , Fatores Etários , China/epidemiologia , Estudos de Coortes , Feminino , Humanos , Incidência , Masculino , Mortalidade/tendências , Neoplasias/mortalidade , Análise de Regressão
17.
Sci Prog ; 104(1): 36850421997286, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-33661721

RESUMO

Immune cells have emerged as key regulators in the occurrence and development of multiple tumor types. However, it is unclear whether immune-related genes (IRGs) and the tumor immune microenvironment can predict prognosis for patients with gastric cancer (GC). The mRNA expression data in GC tissues (n = 368) were obtained from The Cancer Genome Atlas (TCGA) database. Differentially expressed IRGs in patients with GC were determined using a computational difference algorithm. A prognostic signature was constructed using COX regression and random survival forest (RSF) analyses. In addition, datasets related to "gemcitabine resistance" and "trastuzumab resistance" (GSE58118 and GSE77346) were obtained for GEO database, and DEGs associated with drug-resistance were screened. Then, we analyzed correlations between gene expression and cancer immune infiltrates via Tumor Immune Estimation Resource (TIMER) site. The cBioportal database was used to analyze drug-resistant gene mutation status and survival. One hundred and fifty-five differentially expressed IRGs were screened between GC and normal tissues, and a prognostic signature consisting of four IRGs (NRP1, PPP3R1, IL17RA, and FGF16) was closely related to the overall survival (OS). According to cutoff value of risk score, patients were divided into high-risk and low-risk group. Patients in the high-risk group had shorter OS compared to the low-risk group in both the training (p < 0.0001) and testing sets (p = 0.0021). In addition, we developed a 5-IRGs (LGR6, DKK1, TNFRSF1B, NRP1, and CXCR4) signature which may participate in drug resistance processes in GC. Survival analysis showed that patients with drug-resistant gene mutations had shorter OS (p = 0.0459) and DFS (p < 0.001). We constructed four survival-related IRGs and five IRGs related to drug resistance which may contribute to predict the prognosis of GC.

18.
Front Med (Lausanne) ; 8: 632210, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-33693019

RESUMO

Aim: This study aimed to use machine learning algorithms to identify critical preoperative variables and predict the red blood cell (RBC) transfusion during or after liver transplantation surgery. Study Design and Methods: A total of 1,193 patients undergoing liver transplantation in three large tertiary hospitals in China were examined. Twenty-four preoperative variables were collected, including essential population characteristics, diagnosis, symptoms, and laboratory parameters. The cohort was randomly split into a train set (70%) and a validation set (30%). The Recursive Feature Elimination and eXtreme Gradient Boosting algorithms (XGBOOST) were used to select variables and build machine learning prediction models, respectively. Besides, seven other machine learning models and logistic regression were developed. The area under the receiver operating characteristic (AUROC) was used to compare the prediction performance of different models. The SHapley Additive exPlanations package was applied to interpret the XGBOOST model. Data from 31 patients at one of the hospitals were prospectively collected for model validation. Results: In this study, 72.1% of patients in the training set and 73.2% in the validation set underwent RBC transfusion during or after the surgery. Nine vital preoperative variables were finally selected, including the presence of portal hypertension, age, hemoglobin, diagnosis, direct bilirubin, activated partial thromboplastin time, globulin, aspartate aminotransferase, and alanine aminotransferase. The XGBOOST model presented significantly better predictive performance (AUROC: 0.813) than other models and also performed well in the prospective dataset (accuracy: 76.9%). Discussion: A model for predicting RBC transfusion during or after liver transplantation was successfully developed using a machine learning algorithm based on nine preoperative variables, which could guide high-risk patients to take appropriate preventive measures.

19.
J Biomed Inform ; 116: 103728, 2021 04.
Artigo em Inglês | MEDLINE | ID: mdl-33711543

RESUMO

BACKGROUND: Diagnostic or procedural coding of clinical notes aims to derive a coded summary of disease-related information about patients. Such coding is usually done manually in hospitals but could potentially be automated to improve the efficiency and accuracy of medical coding. Recent studies on deep learning for automated medical coding achieved promising performances. However, the explainability of these models is usually poor, preventing them to be used confidently in supporting clinical practice. Another limitation is that these models mostly assume independence among labels, ignoring the complex correlations among medical codes which can potentially be exploited to improve the performance. METHODS: To address the issues of model explainability and label correlations, we propose a Hierarchical Label-wise Attention Network (HLAN), which aimed to interpret the model by quantifying importance (as attention weights) of words and sentences related to each of the labels. Secondly, we propose to enhance the major deep learning models with a label embedding (LE) initialisation approach, which learns a dense, continuous vector representation and then injects the representation into the final layers and the label-wise attention layers in the models. We evaluated the methods using three settings on the MIMIC-III discharge summaries: full codes, top-50 codes, and the UK NHS (National Health Service) COVID-19 (Coronavirus disease 2019) shielding codes. Experiments were conducted to compare the HLAN model and label embedding initialisation to the state-of-the-art neural network based methods, including variants of Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). RESULTS: HLAN achieved the best Micro-level AUC and F1 on the top-50 code prediction, 91.9% and 64.1%, respectively; and comparable results on the NHS COVID-19 shielding code prediction to other models: around 97% Micro-level AUC. More importantly, in the analysis of model explanations, by highlighting the most salient words and sentences for each label, HLAN showed more meaningful and comprehensive model interpretation compared to the CNN-based models and its downgraded baselines, HAN and HA-GRU. Label embedding (LE) initialisation significantly boosted the previous state-of-the-art model, CNN with attention mechanisms, on the full code prediction to 52.5% Micro-level F1. The analysis of the layers initialised with label embeddings further explains the effect of this initialisation approach. The source code of the implementation and the results are openly available at https://github.com/acadTags/Explainable-Automated-Medical-Coding. CONCLUSION: We draw the conclusion from the evaluation results and analyses. First, with hierarchical label-wise attention mechanisms, HLAN can provide better or comparable results for automated coding to the state-of-the-art, CNN-based models. Second, HLAN can provide more comprehensive explanations for each label by highlighting key words and sentences in the discharge summaries, compared to the n-grams in the CNN-based models and the downgraded baselines, HAN and HA-GRU. Third, the performance of deep learning based multi-label classification for automated coding can be consistently boosted by initialising label embeddings that captures the correlations among labels. We further discuss the advantages and drawbacks of the overall method regarding its potential to be deployed to a hospital and suggest areas for future studies.


Assuntos
COVID-19 , Codificação Clínica/métodos , Redes Neurais de Computação , SARS-CoV-2 , COVID-19/epidemiologia , Codificação Clínica/estatística & dados numéricos , Aprendizado Profundo , Registros Eletrônicos de Saúde/estatística & dados numéricos , Humanos , Informática Médica , Pandemias/estatística & dados numéricos , Reino Unido/epidemiologia
20.
Medicine (Baltimore) ; 100(8): e24842, 2021 Feb 26.
Artigo em Inglês | MEDLINE | ID: mdl-33663105

RESUMO

ABSTRACT: An association between animals and volatile anaesthetic requirements has been shown; however, evidence related to the postoperative outcome of human patients is lacking. Our aim was to investigate whether there is a difference in the requirement for sevoflurane among people undergoing gastrointestinal surgery.We observed 390 adult patients who underwent gastrointestinal surgery with an American Society of Anesthesiologists physical status of I or II with an expected surgery duration of > 2 hours. We used the bispectral index (BIS) to guide the regulation of end-tidal sevoflurane concentration (ETsevo). The mean ETsevo from 20 minutes after endotracheal intubation to 2 hours after the start of surgery was calculated for all patients. Differential sevoflurane requirements were identified according to ETsevo. The BIS, ETsevo, heart rate, mean arterial pressure, dose of sufentanil and cisatracurium, tracheal extubation time, incidence of intraoperative awareness, and incidence of postoperative nausea and vomiting were compared between patients with a low requirement for sevoflurane (group L) and patients with a high requirement for sevoflurane (group H).The mean ETsevo of the 390 patients was 1.55% ±â€Š0.26%. Based on our definition, patients with an ETsevo of < 1.29% were allocated to the low requirement group (group L; n = 69), while patients with an ETsevo of > 1.81% were allocated to the high requirement group (group H; n = 78). The ETsevo of group L was significantly lower than the ETsevo of group H (1.29% ±â€Š0.014% vs 1.82% ±â€Š0.017%, P < .001). There was no significant difference in the ETsevo, BIS, heart rate, mean arterial pressure, dose of sufentanil and cisatracurium, tracheal extubation time, incidence of intraoperative awareness, and incidence of postoperative nausea and vomiting. The tracheal extubation time in the L group was significantly shorter than that in the H group. No intraoperative awareness occurred.There was a significant difference in the requirement for sevoflurane in adult patients. The tracheal extubation time in group L was significantly shorter than that in group H.


Assuntos
Período de Recuperação da Anestesia , Anestesia Geral/métodos , Anestésicos Inalatórios/farmacocinética , Sevoflurano/farmacocinética , Extubação/efeitos adversos , Anestésicos Inalatórios/administração & dosagem , Estudos de Casos e Controles , Monitores de Consciência , Procedimentos Cirúrgicos do Sistema Digestório/métodos , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Estudos Prospectivos , Sevoflurano/administração & dosagem
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA
...