Search

Precise detection of recurrence in patients with treated nasopharyngeal carcinoma (NPC) facilitates timely intervention and prolongs survival. However, there is no compelling tool realizing real-time precise recurrence detection as scale hitherto. Here we present a deep learning-based sequential scan model called RAIN, harnessing 10,212 time-series follow-up head and neck magnetic resonance (MR) scans of 1,808 patients with treated NPC in a multicenter observational study (Blinded ID). The RAIN yields larger area under the receiver operating curve (AUC) values than single scan model (internal: 0.916 vs 0.855, p = 0.004; external: 0.900 vs 0.709, p

Dieulafoy's disease is characterized by abnormal submucosal arteries and results in acute luminal hemorrhage. Dieulafoy's lesions can also be found in the submucosa of the bronchus. Due to its low incidence rate and non-specific clinical symptoms, Dieulafoy's disease is easy to overlook, but can lead to massive bleeding and high rates of mortality. Therefore, improvements in the understanding of the disease are necessary. The awareness of the disease and associated diagnostic and treatment techniques have continued to improve, and thus, an increasing number of cases of Dieulafoy's disease of the bronchus have been reported. In the present review, 74 cases of Dieulafoy's disease are summarized. New technologies such as endobronchial ultrasound, narrow-band imaging, angiography and argon plasma treatment have been found to be increasingly applied to diagnose and treat Dieulafoy's disease of the bronchus. Therefore, the primary focus of this systematic review is to highlight advances in the diagnosis and treatment of bronchial Dieulafoy's disease.

Summary Background Upper gastrointestinal cancers (including oesophageal cancer and gastric cancer) are the most common cancers worldwide. Artificial intelligence platforms using deep learning algorithms have made remarkable progress in medical imaging but their application in upper gastrointestinal cancers has been limited. We aimed to develop and validate the Gastrointestinal Artificial Intelligence Diagnostic System (GRAIDS) for the diagnosis of upper gastrointestinal cancers through analysis of imaging data from clinical endoscopies. Methods This multicentre, case-control, diagnostic study was done in six hospitals of different tiers (ie, municipal, provincial, and national) in China. The images of consecutive participants, aged 18 years or older, who had not had a previous endoscopy were retrieved from all participating hospitals. All patients with upper gastrointestinal cancer lesions (including oesophageal cancer and gastric cancer) that were histologically proven malignancies were eligible for this study. Only images with standard white light were deemed eligible. The images from Sun Yat-sen University Cancer Center were randomly assigned (8:1:1) to the training and intrinsic verification datasets for developing GRAIDS, and the internal validation dataset for evaluating the performance of GRAIDS. Its diagnostic performance was evaluated using an internal and prospective validation set from Sun Yat-sen University Cancer Center (a national hospital) and additional external validation sets from five primary care hospitals. The performance of GRAIDS was also compared with endoscopists with three degrees of expertise: expert, competent, and trainee. The diagnostic accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of GRAIDS and endoscopists for the identification of cancerous lesions were evaluated by calculating the 95% CIs using the Clopper-Pearson method. Findings 1 036 496 endoscopy images from 84 424 individuals were used to develop and test GRAIDS. The diagnostic accuracy in identifying upper gastrointestinal cancers was 0·955 (95% CI 0·952–0·957) in the internal validation set, 0·927 (0·925–0·929) in the prospective set, and ranged from 0·915 (0·913–0·917) to 0·977 (0·977–0·978) in the five external validation sets. GRAIDS achieved diagnostic sensitivity similar to that of the expert endoscopist (0·942 [95% CI 0·924–0·957] vs 0·945 [0·927–0·959]; p=0·692) and superior sensitivity compared with competent (0·858 [0·832–0·880], p Interpretation GRAIDS achieved high diagnostic accuracy in detecting upper gastrointestinal cancers, with sensitivity similar to that of expert endoscopists and was superior to that of non-expert endoscopists. This system could assist community-based hospitals in improving their effectiveness in upper gastrointestinal cancer diagnoses. Funding The National Key R&D Program of China, the Natural Science Foundation of Guangdong Province, the Science and Technology Program of Guangdong, the Science and Technology Program of Guangzhou, and the Fundamental Research Funds for the Central Universities.

The incidence of hypoxic pulmonary hypertension (HPH) is increasing. Accumulating evidence suggests that long noncoding RNAs (lncRNAs) play an important role in HPH, but the functions and mechanism have yet to be fully elucidated. In the present study, we established a HPH rat model with 8 h of hypoxia exposure (10% O2) per day for 21 days. High-throughput sequencing identified 60 differentially expressed (DE) lncRNAs, 20 DE miRNAs and 695 DE mRNAs in rat lung tissue. qRT-PCR verified the accuracy of the results. The DE mRNAs were significantly enriched in immune response, inflammatory response, leukocyte migration, cell cycle, cellular response to interleukin-1, IL-17 signalling pathway, cytokine–cytokine receptor interaction and Toll-like receptor signalling pathway. According to the theory of competing endogenous RNA (ceRNA) networks, lncRNA–miRNA–mRNA network was constructed by Cytoscape software, 16 miRNAs and 144 mRNAs. The results suggested that seven DE lncRNAs (Ly6l, AABR07038849.2, AABR07069008.2, AABR07064873.1, AABR07001382.1, AABR07068161.1 and AABR07060341.2) may serve as molecular sponges of the corresponding miRNAs and play a major role in HPH.

Background: No consensus in sequence selection for artificial intelligence model development has been achieved, we aimed to explore whether contrast-enhanced magnetic resonance imaging (ceMRI) could be substituted in the identification and segmentation of nasopharyngeal carcinoma (NPC) with the aid of deep learning models in a large-scale cohort. Methods: A total of 4,478 eligible individuals were randomly split into training, validation and test sets. The diagnostic performance between NPC and benign hyperplasia and segmentation efficacy in NPC were compared among self-constrained 3D DenseNet models developed using axial T1-weighted imaging (T1WI), T2WI or enhanced T1WI (T1WIC) images separately. Findings: All developed models exhibited similar satisfactory diagnostic performance in discriminating NPC from benign hyperplasia, attaining over accuracy over 99.00% in all T stages of NPC. However, the T1WIC model yielded a significantly higher dice similarity coefficient (DSC) and lower average surface distance (ASD) than either the T1WI model or T2WI model in the test set of NPC cohort (DSC, 0.768±0.070 vs 0.759±0.065 and 0.755±0.071, p < 0.001; ASD, 1.573±0.954 mm vs 1.661±0.898 mm and 1.722±1.133 mm, p < 0.01).Interpretation: Unenhanced MRI could substitute ceMRI in discriminating NPC from benign hyperplasia but not in the segmentation for NPC with the aid of deep learning models and ceMRI remained the optimal sequence for developing segmentation models for NPC , indicating the feasibility of reducing the use of MR contrast in screening program. Funding: This work was supported by the National Natural Science Foundation of China [Grant No. 31900461, 81702873, 81872375, 81572665, 81802712]. Declaration of Interest: Nothing to disclose. Ethical Approval: The ethics committee of the authors’ institution approved the study protocol and patient consent was waived due to the retrospective nature of the study.

To explore potential critical genes and identify circular RNAs (circRNAs) that act as the competitive endogenous RNA (ceRNA) in a hypoxic pulmonary hypertension (HPH) rat model. Constructed rat model, and a bioinformatics method was used to analyse differentially expressed (DE) genes and construct a circRNA-miRNA-mRNA ceRNA regulatory network. Then, qRT-PCR was used to verify. The significant DEcircRNAs/DEmiRNAs/DEmRNAs was showed, and a ceRNA network with 8 DEcircRNAs, 9 DEmiRNAs and 46 DEmRNAs were constructed. The functional enrichment suggested the inflammatory response, NF-κB signalling, MAPK cascade and Toll-like receptor were associated with HPH. Further assessment confirmed that circ_002723, circ_008021, circ_016925 and circ_020581 could have a potential ceRNA mechanism by sponging miR-23a or miR-21 to control downstream target gene and be involved in the pathophysiology of HPH. The qRT-PCR validation results were consistent with the RNA-Seq results. This study revealed potentially important genes, pathways and ceRNA regulatory networks in HPH.

Objectives We aimed to develop a dual-task model to detect and segment nasopharyngeal carcinoma (NPC) automatically in magnetic resource images (MRI) based on deep learning method, since the differential diagnosis of NPC and atypical benign hyperplasia was difficult and the radiotherapy target contouring of NPC was labor-intensive. Materials and methods A self-constrained 3D DenseNet (SC-DenseNet) architecture was improved using separated training and validation sets. A total of 4100 individuals were finally enrolled and split into the training, validation and test sets at a proximate ratio of 8:1:1 using simple randomization. The diagnostic metrics of the established model against experienced radiologists was compared in the test set. The dice similarity coefficient (DSC) of manual and model-defined tumor region was used to evaluate the efficacy of segmentation. Results Totally, 3142 nasopharyngeal carcinoma (NPC) and 958 benign hyperplasia were included. The SC-DenseNet model showed encouraging performance in detecting NPC, attained a higher overall accuracy, sensitivity and specificity than those of the experienced radiologists (97.77% vs 95.87%, 99.68% vs 99.24% and 91.67% vs 85.21%, respectively). Moreover, the model also exhibited promising performance in automatic segmentation of tumor region in NPC, with an average DSC at 0.77 ± 0.07 in the test set. Conclusions The SC-DenseNet model showed competence in automatic detection and segmentation of NPC in MRI, indicating the promising application value as an assistant tool in clinical practice, especially in screening project.

We conduct a study to explore the influence of impaired renal function on prognosis in Acute pulmonary embolism (APE) patients. A meta-analysis was performed using the EMBASE and PubMed databases for relevant publications reporting the impact of impaired renal function on the clinical outcomes in patients with APE. Eventually, 17 articles were included in our analysis. The results suggested that renal insufficiency (RI) is a predictor of poor prognosis in APE patients(short-term mortality: pooled OR = 2.83, 95%CI: 2.20���3.63; long-term mortality: pooled OR = 2.30, 95%CI: 1.72���3.08; adverse outcomes: pooled OR = 3.02, 95%CI: 2.60���3.51). The short-term and long-term mortality rates of APE patients with RI were both higher than those in patients without RI. In addition, acute kidney injury(AKI) could serve as a predictive factor of poor prognosis (pooled OR = 2.75, 95%CI: 2.45���3.08), and it doubles the overall mortality rate in APE patients. However, chronic kidney disease (CKD) did not predict poor prognosis in APE patients (pooled OR = 1.94, 95%CI: 0.99���3.81), although it could slightly increase the overall mortality rate in APE patients. RI and AKI could be included in the prognosis evaluation for APE, but the impact of CKD in APE patients has yet to be determined.

Objectives Patients with pulmonary embolism (PE) have a higher red cell distribution width (RDW), which may complicate the interpretation of the prognosis of PE. We aim to evaluate the diagnostic utility of the RDW in patients with PE. Methods Eligible studies were identified by searching PubMed and EMBASE from inception to December 23, 2018. The data were pooled using Stata 12 to evaluate the prognostic value. The sources of heterogeneity were detected. Results For the analysis of prognosis, 7 studies with 2146 patients were included, with median RDW values ranging from 13.4% to 15.7% and a median cut-off point of 18.27%. A total of 247 patients (11.5% mortality) among the 2146 participants died. Overall, RDW was positively associated with the prognosis of PE in the overall analysis (HR = 1.19, 95% CI = 1.10-1.29, I2 = 73.4%; P = 0.000). Subgroup analyses by study design, sex and follow-up time indicated that they were not sources of heterogeneity. Conclusions An elevated RDW has the potential to be a biomarker for prognosis in PE patients.

Phosphatase and tensin homolog (PTEN) plays an important role in the pathogenesis of hypoxic pulmonary hypertension (HPH). A decrease in PTEN expression is associated with the hypermethylation of its promoter. However, whether the demethylation of the PTEN gene could attenuate HPH remains unknown. 5-Aza-2'-deoxycytidine (5-Aza-dC) is a DNA methyltransferase (DNMT) inhibitor. The present study was designed to investigate the effects and mechanisms of 5-Aza-dC on HPH. The proliferation, migration and apoptosis of rat pulmonary artery smooth muscle cells (PASMCs) induced by hypoxia and treated with 5-Aza-dC were detected. The expression of PTEN and DNMTs and the PTEN methylation status of PASMCs were detected. SD rats were randomly divided into normal group, hypoxia group and hypoxia + 5-Aza-dC group. The expression of PTEN was decreased, the expression of DNMTs was increased, and the methylation status of PTEN was increased in hypoxia-induced PASMCs. However, 5-Aza-dC can rescue the decreased PTEN, inhibit DNMT levels in a dose-dependent manner and suppress PTEN methylation. Furthermore, the demethylation of PTEN, which was induced by 5-Aza-dC, inhibited the proliferation, migration and promoted apoptosis in PASMCs. In vivo studies further demonstrated that the expression of PTEN, mean pulmonary artery pressure and right ventricular hypertrophy index in HPH rats was attenuated by 5-Aza-dC. 5-Aza-dC also suppressed the expression of DNMTs and PTEN methylation in the lungs of HPH rats. These results indicated that PTEN promoter methylation status is involved in HPH. 5-Aza-dC, as a DNMT inhibitor, has the potential to attenuate HPH via demethylation of the PTEN promoter.

Background: Upper gastrointestinal (UGI) cancers (including esophageal cancer, gastric cancer, etc.) are the most common cancers worldwide. Detection at early stage is the key for increasing the overall survival. Artificial intelligence (AI) using deep learning algorithms have made remarkable progress in medical imaging. Here we constructed a Gastrointestinal Artificial Intelligence Diagnostic System (GRAIDS) to improve the diagnostic accuracy of upper gastrointestinal (UGI) cancer by analyzing imaging data from clinical endoscopies. Methods: A multi-institutional, diagnostic study from six hospitals of different tiers across China was initiated. We included a total of 1,036,496 endoscopy images from 84,424 individuals to develop and validate the GRAIDS. We evaluated the performance of the GRAIDS using an internal validation set and a prospective validation set from a national hospital and five external validation sets from primary care hospitals. We also compared the diagnostic performance of the GRAIDS with that of endoscopists with three degrees of expertise (expert, competent, and trainee). Findings: The diagnostic accuracy in identifying UGI cancer was 0.955 in the internal validation set and 0.927 in the prospective set, and ranged from 0.915 to 0.977 in the five external validation sets. Compared with the expert endoscopists, the GRAIDS achieved a comparable sensitivity to experts (0.942 versus 0.945; P=0.692) and a superior sensitivity to competent (0.858, P

Elevated red blood cell distribution width (RDW) may correlate with a worse prognosis in pulmonary hypertension (PH), though results to date are inconsistent. The goal of this study is to detect the impact of RDW on the prognosis of PH.PubMed and EMBASE databases were searched from their inception to July 22, 2019 for relevant publications reporting the relationship between RDW and the prognosis of PH. A meta-analysis was performed, and the heterogeneity across the included studies was evaluated using I and Q statistics. We conducted sensitivity and subgroup analyses to detect sources of heterogeneity. In addition, potential publication bias was evaluated by Begg's and Egger's tests.In total, 1236 publications were retrieved, and 7 eligible publications with 666 PH patients were included in our meta-analysis. The results suggested that increased RDW can predict worse prognosis in PH (hazard ratio (HR) = 1.27, 95% confidence interval (CI) 1.11-1.45). According to subgroup analysis, study design, region, various endpoints, time of follow-up, and patient age were not sources of heterogeneity. In addition, RDW showed prognostic value in retrospective studies (HR = 1.32, 95%CI 1.15-1.51) but not in prospective studies (HR = 1.14, 95%CI 0.78-1.67). Additionally, RDW may serve as a predictive biomarker of PH in Europe (HR = 1.33, 95%CI 1.18-1.49) but not in Asia (HR = 1.20, 95%CI 0.90-1.58). Further analysis indicated that the prognostic value of RDW was influenced by patient age (>44 years: HR = 1.34, 95%CI 1.17-1.55; ≤44 years: HR = 1.20, 95%CI 0.90-1.58) and follow-up (

Linkage studies suggested that a quantitative trait locus (QTL) for blood pressure (BP) was present in a region on chromosome 17 (Chr 17) of Dahl salt-sensitive (DSS) rats. A subsequent congenic strain targeting this QTL, however, could not confirm it. These conflicting results called into question the validity of localization of a QTL by linkage followed by the use of a congenic strain made with an incomplete chromosome coverage. To resolve this issue, we constructed five new congenic strains, designated C17S.L1 to C17S.L5, that completely spanned the ±2 LOD confidence interval supposedly containing the QTL. Each congenic strain was made by replacing a segment of the DSS rat by that of the normotensive Lewis (LEW) rat. The only section to be LL homozygous is the region on Chr 17 specified in a congenic strain, as evidenced by a total genome scan. The results showed that BPs of C17S.L1 and C17S.L2 were lower ( P < 0.04) than that of DSS rats. In contrast, BPs of C17S.L3, C17S.L4, and C17S.L5 were not different ( P > 0.6) from that of DSS rats. Consequently, a BP QTL must be located in an interval of ∼15 cM shared between C17S.L1 and C17S.L2 and unique to them both, as opposed to C17S.L3, C17S.L4, and C17S.L5. The present study illustrates the importance of thorough chromosome coverage, the necessity for a genome-wide screening, and the use of “negative” controls in physically mapping a QTL by congenic strains.

Objective Our previous linkage analyses showed that there was likely a quantitative trait locus (QTL) for blood pressure (BP) on chromosome 8 (Chr 8) in the strain comparison between the Dahl salt-sensitive (S) and the Lewis (LEW) rats. The current work is to delineate the chromosome interval harboring this QTL by using congenic strains with different chromosome substitutions. Methods Two congenic strains were produced by replacing different segments of the S rats with the homologous segments of the LEW rats. A genome-wide marker screening was utilized to accelerate this process. The two strains generated are designated as C8S.L1 and C8S.L2, respectively. BPs of the rats were measured by telemetry. Results C8S.L1 showed a BP lower than that of S rats. In contrast, C8S.L2 did not have chromosome overlaps with C8S.L1, but unexpectedly, exhibited a BP-raising effect, higher than that of S rats. Conclusion There are at least two QTLs present in a section of Chr 8 that possess opposite BP effects. The current congenic work reveals not only the presence of QTLs, but the complexity of QTLs on BP. The novel congenic strain with hypertension more severe than S provides a new model for studies in elucidating physiological mechanisms controlling BP.

Pursuing fully a suggestion from linkage analysis that there might be a quantitative trait locus (QTL) for blood pressure (BP) in a chromosome (Chr) 2 region of the Dahl salt-sensitive rat (DSS), four congenic strains were made by replacing various fragments of DSS Chr 2 with those of Lewis (LEW). Consequently, a BP QTL was localized to a segment of around 3 cM or near 3 Mb on Chr 2 by comparative congenics. The BP-augmenting alleles of this QTL originated from the LEW rat, a normotensive strain compared with DSS. The dissection of a QTL with such a paradoxical effect illustrated the power of congenics in unearthing a gene hidden in the context of the whole animal system, presumably by interactions with other genes. The locus for the angiotensin II receptor AT-1B ( Agtr1b) is not supported as a candidate gene for the QTL because a congenic strain harboring it did not have an effect on BP. There are ∼19 known and unknown genes present in the QTL interval. Among them, no standout candidate genes are reputed to affect BP. Thus the QTL will likely represent a novel gene for BP regulation.