Search

Diabetic Retinopathy (DR), the most prevalent eye condition in diabetic patients, can be brought on by diabetes. It may result in reduced eyesight or possibly complete blindness. Early detection can help prevent eye damage, but it can be challenging because symptoms might not show up right away. To grade diabetic retinopathy and identify all stages of DR, many predictive models based on machine learning and deep learning have been proposed. However, the majority of the diabetic retinopathy grading models now in use cannot identify all stages. For DR grading, the most used metrics, including as accuracy, f1-score, precision, recall, and AUC-ROC score, are unreliable. Since they do not consider the level of disagreement between target and predicted labels and the ordinality of target labels, which helps classify all DR stages. To address these issues, we developed the customized EfficientNet b3. The Quadratic Weighted Kappa (QWK) was chosen as the primary evaluation metric since the distinct stages of diabetic retinopathy are ordinal. The degree of discrepancy between the actual and expected labels is also taken into consideration. We were able to achieve a higher QWK score of 0.87 by recognizing all the DR stages after aggregating the predictions from the efficientnet-b3 model, which was trained for 30 and 60 epochs, respectively. This result is state-of-the-art. DR grading must identify each stage in order to provide patients with the greatest care available. Our Confusion matrix of ensembled efficientnetb3 shows, which is state-of-the-art, that all DR stages are identified with higher probabilities.

Backgrounds: The rapid development of comprehensive genomic profiling facilitates the identification of rare epidermal growth factor receptor (EGFR) mutations; however, the clinical implications of these newly identified rare mutations were largely unknown. Methods: A lung adenocarcinoma patient with leptomeningeal metastasis (LM) who was treated with various EGFR tyrosine kinase inhibitors (TKIs) was included in the study. Next-generation sequencing of 139 cancer-relevant genes was performed on the serial cerebrospinal fluid (CSF) specimens collected during the treatment. Simulated structural analysis and in vitro drug sensitivity were conducted to elucidate novel EGFR TKI resistant mechanisms. Results: The patient was initially positive for EGFR 19-exon deletion (19-Del), and upon drug resistance to first-generation EGFR inhibitors and osimertinib, a previously unreported EGFR-SEPT14 (E25:S7) fusion was identified. The patient achieved a promising response after treating intrathecal pemetrexed plus second-generation EGFR inhibitors. Structural analysis suggested that the fusion protein and wild-type EGFR had similar secondary structures but distinct tertiary structures. In vitro studies demonstrated that EGFR-SEPT14 fusion itself was not oncogenic; however, it could function together with EGFR 19-Del to confer drug resistance to first-generation inhibitors and osimertinib while remaining sensitive to some second- and third-generation EGFR inhibitors. Conclusions: Overall, this is the first study reporting EGFR fusions can serve as an acquired resistant mechanism to EGRR inhibitors in lung adenocarcinoma, which is potentially through heterodimerization with classic EGFR mutations, and the identification of its differential sensitivity to certain EGFR TKIs could facilitate personalized treatments to patients with rare EGFR alterations.

Papillary thyroid carcinoma (PTC) is one of the fastest growing endocrine system malignant carcinomas detected over the past decade. Unfortunately, more than 25% of PTC patients are characterized by their aggressiveness and subsequent metastasis; these characteristics usually indicate poor prognosis. Recently, increasing evidence has suggested that solute carrier (SLC) transporters may play a pivotal role in the initiation, invasion and metastasis of human carcinoma. However, the expression and clinicopathological significance of SLC transporters in patients with PTC remains undetermined. In this study, we aimed to elucidate how the differential expression of SLC transporters affects clinicopathological features, as well as determine the possible regulatory signaling pathways involved. Three differentially expressed SLC transporters were screened from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) database using a bioinformatics approach. The results indicated that high SLC34A2 and low SLC4A4 protein expression exhibited a higher percentage of capsular invasion and extra-thyroid metastasis in patients. Logistic regression analysis showed that high SLC34A2 expression in tumors was identified as an independent risk factor for capsular invasion [odds ratio (OR)=11.400, 95% confidence interval (CI)=1.733-74.995, P=0.011] and extra-thyroid metastasis (OR=4.920, 95%CI=1.234-19.623, P=0.024), while low SLC4A4 expression in tumors was only identified as independent risk factors for extra-thyroid metastasis (OR=8.568, 95%CI =1.186-61.906, P=0.033). Specifically, for tumors with capsular invasion and extra-thyroid metastasis, the protein expression staining of SLC34A2 was markedly enhanced in the cytoplasm of follicular epithelial cells, contrastingly, SLC4A4 expression was notably weakened in the cytomembrane and nucleus. Intriguingly, both high SLC34A2 and low SLC4A4 protein expression were significantly linked to a high urinary iodine concentration in patients with PTC. Mechanistically, compared with adjacent normal thyroids, p-ERK was significantly up-regulated by 17.8% in the invading tumor; p-ERK, p-JNK, and p-P38 were markedly up-regulated by 29.2%, 67.1%, and 38.9% for metastatic tumors, respectively. Importantly, SLC4A4 negatively correlated with p-JNK (r=-0.696, P= 0.004) and p-P38 (r=-0.534, P=0.049). In conclusion, we suggest that up-regulated SLC34A2 (mainly in the cytoplasm) and down-regulated SLC4A4 (mainly in the cytomembrane and nucleus), which might be attributed to excess iodine intake, were closely linked to extra-thyroid metastasis in PTCs. Furthermore, this effect of SLC4A4 may be through the activation of JNK/P38 MAPK signaling pathway. Future in vivo and in vitro gain- or loss-of-function experiments are needed to verify these findings and further elucidate the deeper molecular mechanisms.

The aim of the present study was to elucidate the association between excessive chronic iodine exposure and the risk of developing papillary thyroid carcinoma (PTC). The demographic information and pathological characteristics of patients with thyroid nodules were retrieved from medical records at The Second Hospital of Shandong University. A fasting urine specimen was collected, and creatinine and urinary iodine concentration (UIC) were determined. The water iodine data from the domicile districts of these patients were collated from published reports. The results revealed that almost half of the patients with PTC (44.3%) also exhibited a high UIC (≥300 µg/l). Multivariate analysis revealed that the adjusted odds ratio for high UIC was 3.987 (95% CI: 1.355-11.736) and the adjusted area under the receiver operating characteristic curve was 0.776 (95% CI: 0.687-0.864), which was associated with PTC risk in patients with thyroid nodules. Integrated ecological assessment of chronic iodine exposures demonstrated that >80% (81.4%) of the patients with PTC who also exhibited a high UIC were from historically non-iodine-deficient regions, and 66.7% of patients with PTC who resided in historically iodine-excessive regions were characterized by high UICs. Importantly, a high UIC was significantly associated with capsular invasion and extrathyroid metastasis (P

The gas hydrate deposits at seabed fall into a special category of gas hydrate resources. This study investigates the technical feasibility of harvesting natural gas from seabed hydrates using a new thermal method called Moving Riser Method (MRM). A mathematical model for heat transfer along the injection pipe in the MRM system was developed. Heat transfer from the injected hot water to the gas hydrate deposit was analyzed. This study concludes that with today’s pipe insulation technology water temperature drops only a few degrees from sea surface level to the seafloor level in an insulated 800 m deep vertical pipe. The injected water at seafloor level will be hot enough to dissociate gas hydrate at a commercial rate with an affordable gas consumption rate. The gas production to gas combustion ratio (PCR) is greater than 4. The PCR increases slightly with gas combustion rate. Even the gas production ship stays at the same location for over 40 h with continuous injection of hot water, the water-hydrate boundary will still be within 0.9 m of the hot water injection point. Therefore it is possible to use a gas collector of reasonable size (e.g., 2 m in diameter) to gather all dissociated gas from the hydrate deposit. Result of this investigation shows that harvesting natural gas from gas hydrate at seabed with the MRM is technically viable, economically feasible, and environmentally safe.

Oil companies have realized that refracturing old wells is the more budget-friendly way to achieve cash flow than drilling and completing new wells in the down turn of the energy industry. But it has been seen that the refracturing technology does not always work equally well in all reservoir conditions. It is desirable to have a quantitative method for assessment of well refracturing design and operations. An analytical transient flow model was derived in this study to estimate the orthogonality between the old and new hydraulic fractures after refracturing. Pressure transient data from seven wells refractured in an oilfield in the southwestern China was analyzed with the model. The model implies that orthogonal fractures created in well refracturing treatments induce orthogonal reservoir linear flow in the stimulated reservoir volume. This fracture configuration can be identified by the unit slope of pressure derivative in the log-log diagnosis plot. The angle between fractures initiated from wellbore can be estimated using pressure transient data analysis technique with a log-log diagnostic plot, with extreme situations that merged or parallel fractures will show a half slope and orthogonal fractures will show a unit slope. Case studies of data from 7 wells show that orthogonal fractures were created in 3 wells, merging or parallel fractures were created in 1 well, and fractures with non-right angles were created in 3 wells. The actual well productivity data for the 7 wells are basically consistent with the expected trend implied by the estimated fracture configurations with an exception of one well. The reason is not clear and needs further investigations. More production data from refractured wells are required to further validate the technique.

Sustained Casing Pressure (SCP) has been recognized as a major problem in multi-fractured shale gas wells. The problem is attributed to the failure of cement sheath during the hydraulic fracturing process. Currently available mechanics models predict well integrity conditions with high error. We believe the reason is that the debonding at the casing-cement interface and plastic deformation of cement do not create significant axial channels for gas to migrate to the surface but the radial cracking of cement sheath is responsible for the gas migration resulting in the SCP. Although some mathematical models can predict radial cracking of cement sheath, they have limitations in applications because they do not account for the weakened cement sheath and reduced support from formation rock due to the low efficiency of cement placement in the annulus. An analytical model was derived in this work to predict the Maximum Permissible Fracturing Pressure (MPFP) due to radial cracking with consideration of the weakened cement sheath and the reduced support from formation rock. The results are in good agreement with observations in the Fuling shale gas field, China. Sensitivity analyses show that improving cement strength is not an effective means of increasing MPFP. The MPFP should be enhanced by increasing the thickness of the cement sheath using large wellbore size and improving the cement placement efficiency. This work also shows that cement sheath failure will occur before casing failure occurs in the hydraulic fracturing process. Therefore integrity of cement sheath should be considered in the hydraulic pressure design.

Production prediction in petroleum industry plays a significant role in designing the strategy of the exploration and development. However, due to the complex and uncertain underground formations, it is difficult to obtain accurate forecasts. Most of the recent neural network studies on predicting the production rate in reservoirs are limited in optimizing the network models' parameters and reducing the significant computational cost during the training process. To address these issues, we develop a novel echo-state network (ESN) with attention mechanism to forecast the well performance based on production data. Firstly, in the training stage of the ESN model, we used the attention mechanism to extract the relevant features from contextual production rate; secondly, a dimensionality reduction procedure of the reservoir is carried out; thirdly, the generalization capability of the ESN model will be increased significantly by using the regularization constraints; finally, the generic algorithm will be employed in the validation process to optimize the hyperparameters of network. Five cases were carried out to validate the prediction performance of the presented approach. The comparisons with existing deep learning methods and decline curve analysis (DCA) models were implemented. The obtained results indicate that our built model is significantly superior to other deep learning and DCA approaches currently available in the literature. The approach presented in this paper is not only developed to forecast the short-term well performance, but also the production rate reconstructed and predicted using the new method can be employed to estimate missing flow history.

Deep learning technology can fit the nonlinear relations between different logging sequences. It solves the prediction problems that cannot be effectively disposed by traditional physical or empirical models. However, these deep learning models lack uncertainty analysis, which affects the popularity of logging prediction models in the petroleum engineering application. In this study, we investigate the nonlinear well logging prediction and uncertainty analysis methods based on recurrent neural network (RNN) with attention mechanism and Bayesian theory. A codec structure model based on gated recurrent unit (GRU) neural network and attention mechanism is established for data prediction. Integrating Bayesian theory into GRU neural network, a GRU Bayesian framework is presented to capture model uncertainty and data uncertainty. The importance of different logging sequences to the predicted goal at the same depth and a certain depth range is considered by using the attention mechanism, which improves the prediction accuracy and reduces the uncertainty of the prediction. Compressional waves sonic log (DTC) data of carbonate reservoir is predicted by using the existing logging attributes (density, spontaneous potential, natural gamma ray and deep investigate double lateral resistivity). Compared with the traditional RNN models, the accuracy of the constructed model has increased by 7.95% (R2 = 93.07%) and the error decreased by 1.7166% (RMSE = 2.9261%) in the field data. More importantly, the proposed method can quantitatively analyze the uncertainties of the predicted results, which effectively heightens the application of the logging prediction model in the petroleum engineering field.

This study presents a novel technique to develop offshore gas hydrates using geothermal energy. The technique involves configuring a pair of y-shaped wellbores, which transport the geothermal energy from the geothermal zone to the hydrate-bearing zone using a working fluid (i.e., seawater) to facilitate the dissociation of gas from the gas hydrates. One of the wellbores is a heat absorber located in the geothermal zone, which warms up the seawater. The heated seawater flows to another wellbore called a heat dissipator located in the hydrate-bearing zone, which heats the gas hydrates. A mathematical model is established to obtain the temperature profiles of the working fluid along the gas recovery wellbore. The mathematical model is an integrated part of the technique to optimize wellbore configurations and operational parameters such as the injection rate of the working fluid. The results suggest that the proposed technique is technically feasible. It is found through sensitivity analysis that the configuration of cement (thermal conductivity and thickness) and the properties of the working fluid (heat capacity and injection rate) are the major factors that affect the heat transfer performance of the process. Higher overall thermal conductivity of the wellbore assembly at the heat absorber and heat dissipator, and lower overall heat conductivity at the vertical wellbore section can increase the heat transfer efficiency of the process. Lower heat capacity and higher injection rates of the working fluid are seen to increase the heat transfer efficiency. The results of this study provide a promising alternative technique for the economical and sustainable extraction of gas hydrate. The derived mathematical model can serve as a fundamental pilot tool for the design, deployment, and operation of the y-shaped well couples.

The objective of this study was to develop an accurate method for predicting hydraulic unit types in a heterogeneous carbonate reservoir. There is a significant practical potential in the use of the flow unit characterization. Identification of flow units in inhomogeneous carbonate reservoir presents a great challenge to geologists and engineers. A new method for dividing the flow units was proposed in this study based on the joint of wavelet transform (WT) and least squares support vector machine (LSSVM) within the most productive carbonate reservoir of the Minghuazhen Formation in Region A, Block X in the Petrochina Dagang oilfield. Petrophysical properties derived from core data and logging from 21 representative wells were analyzed. The flow units were classified as five types based on the flow zone index (FZI) approach. The WT and LSSVM were jointly used for learning and training each unit. The well logs were broken down into high and low frequency data using WT. Sensitivity analysis of parameters of training samples to select the largest impact was performed with C5.0 decision tree to obtain a WT-trained set. A predictive model was then established by training LSSVM model. The final trained model with the identification rule and criterion for the classification of flow units was used for identifying flow units in the cored and non-cored intervals of reservoir. The result from this study is consistent with core data and is more accurate than that from the previous investigations. It is concluded that using the combination of the WT and LSSVM improved the accuracy of classification of flow units in the Minghuazhen Formation.

Well logging is a significant method of formation description and resource assessment in exploration and development of oil, natural gas, minerals, groundwater, and sub-surface thermal energy, as well as geotechnical engineering and environmental research. However, the shortage problem of well logging data always exists because well logs can only be measured through a drilling process involving costly and time-consuming field trials. To address this issue, bidirectional long short-term memory (BiLSTM), attention mechanism, and convolutional neural network (CNN) were coupled to build hybrid neural networks for predicting missing well logs. The proposed architecture is a structure of two branches. One branch uses CNN to capture the spatial properties of well logs, and the other one conducts the feature selections by utilizing two-layer BiLSTM with attention mechanism. The spatio-temporal correlations from two branches are merged to forecast the target well logs. The performance of the proposed method is evaluated within a highly heterogeneous reservoir at the Gangdong oilfield in China. In our experiments, six models were trained and used for generating synthetic well logs including compensated neutron logs (CNL), acoustic (AC), spontaneous potential (SP), gamma-ray (GR), density (DEN), and formation resistivity (RT). Moreover, traditional machine learning models, CNN, BiLSTM, and other deep learning benchmark models were developed to compare with the presented models. Results show that the proposed method achieves higher prediction accuracy because it takes into account the spatio-temporal information of well logs.

The natural gas in gas hydrates is considered as a major resource of future clean energy. Producing natural gas from gas hydrates at seafloor presents a great challenge to the engineers due to low efficiency in heat transfer processes. A numerical model was developed in this study to investigate the temperature profiles along the hose assembly in the MRM for harvesting the natural gas from hydrates at the seafloor. Sensitivity analyses were carried out to find major factors affecting the delivering water temperature to the gas hydrate deposits. The result shows that decreasing the temperature of the injected water at surface will cause drop in the delivery temperature of water in the pipe. The drop in the surface temperature is transferred almost by the same amount to the drop in the delivering temperature with the pipe insulation used. The temperature of the water delivered to the gas hydrates drops slowly and linearly with water depth at high water injection rate, and sharply and non-linearly with water depth at low water injection rate. The delivery water temperature is strongly controlled by the pipe insulation but insensitive to it when the thermal conductivity of the insulation material is less than 0.0245 W/m-oC. The temperature in the riser annulus is not essentially affected by any of the invested parameters. A comparison between the numerical model and analytical model under the condition of homogeneous mixture flow was performed. The result shows a good agreement between them, which verified the numerical model.

Calbindin 1 (CALB1), a constituent CaIn this study, CALB1 was found to be overexpressed in NSCLC tissues through the immunohistochemistry assay. Higher CALB1 expression levels were significantly associated with the tumor-node-metastasis (TNM) stage. Moreover, higher expression of CALB1 predicts poor survival in NSCLC patients using the Kaplan-Meier plotter online analysis. In addition, miR-454-3p was predicted to target CALB1 using a software algorithm, validated by the luciferase assay, and analyzed by quantitative polymerase chain reaction and Western blot. The authors further found that miR-454-3p was downregulated in NSCLC tissues and cell lines. Downregulation of CALB1 or upregulation of miR-454-3p significantly suppressed NSCLC cell proliferation and induced cell apoptosis as shown by CCK-8 and flow cytometry analysis, respectively.Our results suggest that CALB1 is a direct target of miR-454-3p and downregulation of CALB1 by miR-454-3p suppressed NSCLC cell functions, which may shed light on its potential application in NSCLC therapy.

Objective. Despite gaining popularity, bilateral axillo-breast approach robotic thyroidectomy (BABA RT) remains controversial. We performed a meta-analysis to evaluate the safety and effectiveness of BABA RT compared with open thyroidectomy (OT) in thyroid disease. Methods. A literature search was conducted using various databases, including PubMed, Web of Science, and Cochrane Library, up to February 2018. Outcomes of interest included patient characteristics, surgical outcomes, adverse events, complications, and surgical completeness. Results. A total of 11 publications including 2733 patients (1070 in the BABA RT and 1663 in the open group) were finally selected for the meta-analysis. BABA RT was associated with an equivalent complication rate, including transient and permanent hypocalcemia, transient and permanent recurrent laryngeal nerve palsy, bleeding, chyle leakage, and seroma, as well as surgical outcomes including tumor size, length of hospital stay, total drain amount, and pain score. BABA RT was also associated with an equivalent surgical completeness, including total dose of radioactive iodine, proportion of stimulated thyroglobulin (sTg)

Background Transoral thyroidectomy is a kind of "natural orifice transluminal endoscopic surgery (NOTES)" which is now being performed in increasing frequency. However, the safety and feasibility have not been concluded yet. Materials and methods A systemic literature search was performed in Pubmed, Cochrane, and Embase databases to identify all studies written in English and published up to April 2017. The keywords used were "transoral endoscopic," "transoral robotic," "oral vestibular endoscopic," and "oral vestibular robotic" combined with "thyroidectomy" or "thyroid surgery." Results Ten articles containing 211 cases matched the review criteria. The weighted average operative time was 119.9 minutes with an average intraoperative blood loss of 35.5 mL while the weighted average length of hospital stay was 4.0 days. The overall conversion rate to open surgery was 1.9%. An overall incidence rate of temporary hypoparathyroidism was 7.1%, temporary recurrent laryngeal nerve injury was 4.3%, whereas of mental nerve palsy was 4.3%. Conclusions According to those reviewed literatures, we can conclude that transoral thyroidectomy is safe and feasible in well-selected patients and offers good perioperative and postoperative outcomes.

Production of natural gas from unconventional gas-hydrate reservoirs faces several challenges. One of them is the well control issue due to the natural gas released from gas hydrates during well drilling. It is very important for drilling engineers to know if the temperature of drilling fluid in the borehole is lower than the critical temperature that causes hydrate dissociation. However, there is no method to predict the fluid temperature with consideration of the effects of Joule-Thomson cooling and drilling cuttings entrained at the bottom hole. This paper fills the gap. An analytical model was developed in this study for predicting temperature profile in drilling gas-hydrate deposits. Result from the analytical model was verified by a numerical model developed using finite difference method. A case study indicates a good consistency between model-implications and field observations. Sensitivity analyses with the model show that the bottom-hole temperature in gas-hydrate drilling is dominated by the temperature and flow rate of the injected drilling fluid. The temperature of drilling fluid in the annulus can become greater than the geo-temperature at the same depth at high fluid flow rate. The Joule-Thomson cooling effect below the drill bit nozzles rapidly diminishes in a short interval above the bottom hole due to the heating effect of geo-thermal gradient. The rate of penetration of drill bit has negligible effect on the fluid temperature profile due to the low percentage of heat flow contributed by the entrained drill cuttings. This paper provides drilling engineers a rigorous method for predicting wellbore temperature profile during drilling gas-hydrates reservoirs.

It is generally recognized that oil influx from pay zones creates high-temperature fluid streams in Underbalanced drilling (UBD) and increases drilling cost due to bore hole complications. However there is lack of an accurate and simple method to predict the fluid temperature that is necessary for taking actions to prevent down hole complications. This study fills the gap. A simple analytical method was developed to replace sophisticated mathematics involving numerical solution procedure. The analytical model includes closed form equations for predicting the fluid temperature profiles inside the drill string and in the annulus. The analytical model considers formation oil influx, entrained cuttings and Joule-Thomason cooling effect at the drill bit. A field case study and a sensitivity analysis with the analytical model were conducted. The field case study with the analytical model indicates that the model can match temperature measurement perfectly by tuning three parameters namely temperature drop at bit, formation oil influx rate, and thermal conductivity of cement sheath. Result of sensitivity analysis shows that the temperature profiles in UBD is very sensitive to the temperature drop at bit and sensitive to the formation oil influx rate and thermal conductivity of cement sheath, especially in the low range of influx rate and thermal conductivity. More clean data sets are required to further validate the newly developed analytical model. Once fully validated, this analytical model can be used in the petroleum industry to improve Underbalanced drilling performance by reducing drilling complications and thus cost of operations.

Cholangiocarcinoma is a devastating malignancy that is notoriously difficult to diagnose and is associated with a high mortality. Despite extensive efforts to improve the diagnosis and treatment of this neoplasm, limited progress has been made. Integrin β6 is a subtype of integrin that is expressed exclusively on the surfaces of epithelial cells and is associated with a variety of tumors. In the present study, we investigated the expression and roles of integrin β6 in cholangiocarcinoma. β6 upregulation in cholangiocarcinoma was correlated with lymph node metastasis and distant metastasis. Moreover, integrin β6 was identified as a biomarker for the diagnosis of cholangiocarcinoma and an indicator of lymph node metastasis. Integrin β6 significantly promoted the proliferation, migration and invasion of cholangiocarcinoma cells. Furthermore, integrin β6 increased Rac1-GTPase, resulting in the upregulation of metalloproteinase-9 (MMP9) and F-actin polymerization. Taken together, our results indicate that integrin β6 promotes tumor invasiveness in a Rac1-dependent manner and is a potential biomarker for tumor metastasis. Integrin β6 may help to improve the diagnostic accuracy and targeting β6 may be a novel strategy for the treatment of cholangiocarcinoma.

In this paper, we address the problem of data aggregation scheduling problem to minimize the latency in wireless sensor networks (WSNs). An efficient distributed synchronous aggregation scheduling method is proposed to structure a collision-free schedule for data aggregation in WSNs. By using a connected dominating set (CDS) as an aggregation tree, we implement the synchronous aggregation scheduling. We prove that the latency of the aggregation schedule generated by our algorithm is at most 4 15 16 R    time-slots where R is the network radius and  is the maximum node degree in the communication graph of the original network. Analysis and simulation results show the validity and superiority of the algorithm.