Search

Abstract Astragalus membranaceus (Fisch.) Bge (AM) is a medicinal herb plant belonging to the Leguminosae family. In this study, we present a chromosome-scale genome assembly of AM, aiming to enhance the molecular biology and functional studies of Astragali Radix. The genome size of AM is about 1.43 Gb, with a contig N50 value of 1.67 Mb. A total of 98.16% of the assembly anchored to 9 pseudochromosomes using Hi-C technology. The assembly completeness was estimated to be 97.27% using BUSCO with the long terminal repeat assembly index (LAI) of 16.22 and quality value (QV) of 48.58. Additionally, the genome contained 67.98% repetitive sequences. Genome annotation predicted 29,914 protein-coding genes, including 73 genes involved in the flavonoid biosynthetic pathway and 2,048 transcription factors. The high-quality genome assembly and gene annotation resources will greatly facilitate future functional genomic studies in Leguminosae species.

Abstract Background Cancer stem cells (CSCs) are a small population of cells in tumor tissues that can drive tumor initiation and promote tumor progression. A small number of previous studies indirectly mentioned the role of F-box and WD repeat domain-containing 7 (FBXW7) as a tumor suppressor in Triple-negative breast cancer (TNBC). However, few studies have focused on the function of FBXW7 in cancer stemness in TNBC and the related mechanism. Methods We detected FBXW7 by immunohistochemistry (IHC) in 80 TNBC patients. FBXW7 knockdown and overexpression in MD-MBA-231 and HCC1937 cell models were constructed. The effect of FBXW7 on malignant phenotype and stemness was assessed by colony assays, flow cytometry, transwell assays, western blot, and sphere formation assays. Immunoprecipitation-Mass Spectrometry (IP-MS) and ubiquitination experiments were used to find and verify potential downstream substrate proteins of FBXW7. Animal experiments were constructed to examine the effect of FBXW7 on tumorigenic potential and cancer stemness of TNBC cells in vivo. Results The results showed that FBXW7 was expressed at low levels in TNBC tissues and positively correlated with prognosis of TNBC patients. In vitro, FBXW7 significantly inhibited colony formation, cell cycle progression, cell migration, EMT process, cancer stemness and promotes apoptosis. Further experiments confirmed that chromodomain-helicase-DNA-binding protein 4 (CHD4) is a novel downstream target of FBXW7 and is downregulated by FBXW7 via proteasomal degradation. Moreover, CHD4 could promote the nuclear translocation of β-catenin and reverse the inhibitory effect of FBXW7 on β-catenin, and ultimately activate the Wnt/β-catenin pathway. Rescue experiments confirmed that the FBXW7-CHD4-Wnt/β-catenin axis was involved in regulating the maintenance of CSC in TNBC cells. In animal experiments, FBXW7 reduced CSC marker expression and suppressed TNBC cell tumorigenesis in vivo. Conclusions Taken together, these results highlight that FBXW7 degrades CHD4 protein through ubiquitination, thereby blocking the activation of the Wnt/β-catenin pathway to inhibit the stemness of TNBC cells. Thus, targeting FBXW7 may be a promising strategy for therapeutic intervention against TNBC.

In this paper, we introduce a high-altitude multi-drone multi-target (HAMDMT) tracking method called STCA, which aims to collaboratively track similar targets that are easily confused. We approach this challenge by categorizing the HAMDMT tracking into two principal tasks: Single-Drone Tracking and Cross-Drone Association. Single-Drone Tracking employs positional and appearance data vectors to overcome the challenges arising from similar target appearances within the field of view of a single drone. The Cross-Drone Association employs image-matching technology (LightGlue) to ascertain the topological relationships between images captured by disparate drones, thereby accurately determining the associations between targets across multiple drones. In Cross-Drone Association, we enhanced LightGlue into a more efficacious method, designated T-LightGlue, for cross-drone target tracking. This approach markedly accelerates the tracking process while reducing indicator dropout. To narrow down the range of targets involved in the cross-drone association, we develop a Common View Area Model based on the four vertices of the image. Considering to mitigate the occlusion encountered by high-altitude drones, we design a Local-Matching Model that assigns the same ID to the mutually nearest pair of targets from different drones after mapping the centroids of the targets across drones. The MDMT dataset is the only one captured by a high-altitude drone and contains a substantial number of similar vehicles. In the MDMT dataset, the STCA achieves the highest MOTA in Single-Drone Tracking, with the IDF1 system achieving the second-highest performance and the MDA system achieving the highest performance in Cross-Drone Association.

Abstract Underwater image enhancement for image processing and underwater robotic vision have recently attracted much academic attention. However, in most existing methods, underwater image enhancement is completed with a simple assumption: the attenuation coefficients are unified across the color channels. This assumption leads to unstable and visually unpleasing enhancement results. Moreover, these methods cannot be successfully applied to explore relatively independent transmissions from multiple color channels with complimentary feature information. To address these challenges, a novel channel‐wise transmission estimation network (CTEN) is proposed, which aims to pioneer the exploration of the transmission difference across the color channels in an underwater scene. Specifically, a color‐specific correction module is proposed to automatically quantify the transmission ability of multiple color channels in the underwater environment. Furthermore, a channel‐wise transmission estimation module is designed to simultaneously explore the relative independence of multi‐color channels and estimate the medium transmissions for each color channel, which represents the attenuation degree of different color radiances after reflecting in the water. Then, a novel residual strategy is introduced to integrate these two modules to complete the underwater enhancement. Using the model, the authors are able to provide an answer as to why channel‐wise transmission estimation are better than single transmission estimation and establish a generalization theory to show the effect of the independent transmission estimation model for each color channel. Experiments on several underwater image datasets verify the superiority of the proposed CTEN model.

There exist various methods for transferring knowledge between neural networks, such as parameter transfer, feature sharing, and knowledge distillation. However, these methods are typically applied when transferring knowledge between networks of equal size or from larger networks to smaller ones. Currently, there is a lack of methods for transferring knowledge from shallower networks to deeper ones, which is crucial in real-world scenarios such as system upgrades where network size increases for better performance. End-to-end training is the commonly used method for network training. However, in this training strategy, the deeper network cannot inherit the knowledge from the existing shallower network. As a result, not only is the flexibility of the network limited but there is also a significant waste of computing power and time. Therefore, it is imperative to develop new methods that enable the transfer of knowledge from shallower to deeper networks. To address the aforementioned issue, we propose an depth incremental learning strategy (DILS). It starts from a shallower net and deepens the net gradually by inserting new layers each time until reaching requested performance. We also derive an analytical method and a network approximation method for training new added parameters to guarantee the new deeper net can inherit the knowledge learned by the old shallower net. It enables knowledge transfer from smaller to larger networks and provides good initialization of layers in the larger network to stabilize the performance of large models and accelerate their training process. Its reasonability can be guaranteed by information projection theory and is verified by a series of synthetic and real-data experiments.

PurposeTo propose a deep learning network with subregion partition for predicting metastatic origins and EGFR/HER2 status in patients with brain metastasis.MethodsWe retrospectively enrolled 140 patients with clinico-pathologically confirmed brain metastasis originated from primary NSCLC (n = 60), breast cancer (BC, n = 60) and other tumor types (n = 20). All patients underwent contrast-enhanced brain MRI scans. The brain metastasis was subdivided into phenotypically consistent subregions using patient-level and population-level clustering. A residual network with a global average pooling layer (RN-GAP) was proposed to calculate deep learning-based features. Features from each subregion were selected with least absolute shrinkage and selection operator (LASSO) to build logistic regression models (LRs) for predicting primary tumor types (LR-NSCLC for the NSCLC origin and LR-BC for the BC origin), EGFR mutation status (LR-EGFR) and HER2 status (LR-HER2).ResultsThe brain metastasis can be partitioned into a marginal subregion (S1) and an inner subregion (S2) in the MRI image. The developed models showed good predictive performance in the training (AUCs, LR-NSCLC vs. LR-BC vs. LR-EGFR vs. LR-HER2, 0.860 vs. 0.909 vs. 0.850 vs. 0.900) and validation (AUCs, LR-NSCLC vs. LR-BC vs. LR-EGFR vs. LR-HER2, 0.819 vs. 0.872 vs. 0.750 vs. 0.830) set.ConclusionOur proposed deep learning network with subregion partitions can accurately predict metastatic origins and EGFR/HER2 status of brain metastasis, and hence may have the potential to be non-invasive and preoperative new markers for guiding personalized treatment plans in patients with brain metastasis.

Abstract Parkinson's disease (PD) is currently an ongoing challenge in daily clinical medicine. To reduce diagnosis time and arduousness and even assess PD levels, a temporal pyramid attention‐based spatiotemporal (PAST) fusion model for diagnosis of PD is produced by using gait data from ground reaction forces. This model is innovative in two aspects. First, by using the temporal pyramid attention module, multiscale temporal attention is obtained from raw sequences. Second, 1D convolutional neural network and bidirectional long short‐term memory layers are used together to learn spatial fusion features from multiple channels in the spatial domain to obtain multichannel, multiscale fusion features. Experiments are performed on the PhysioBank data set, and the results show that the proposed PAST model outperforms other state‐of‐the‐art methods on classification results. This model can assist in the diagnosis and treatment of PD by using gait data.

An important indicator of cervical cancer diagnosis is to calculate the proportion of diseased cells and cancer cells, so it is necessary to segment cells and judge the cell status. The existing methods are difficult to deal with the segmentation of overlapping cells. In order to solve this problem, we put forward such a hypothesis by reading a large number of literature, that is, image segmentation and edge measurement tasks have unity in high-level features. To prove this hypothesis, in this paper, we focus on the complementary between overlapping cervical cell edge information and cell object information to get higher accuracy cell edge detection results. Specifically, we present a joint multi-task learning framework for overlapping cell edge detection by the mask guidance pyramid network. The main component of the framework is the Mask Guidance Module (MGM), which integrates two tasks and stores the shared latent semantics to interact in the two tasks. For semantic edge detection, we propose the novel Refinement Aggregated Module (RAM) fusion to promote semantic edges. Finally, to improve the edge pixel accuracy, the edge consistency constraint loss function is introduced to our model training. Our extensive experiments have proved that our method outperforms other edge detection efforts.

The in situ measurement of thick oil slick thickness (>0.5 mm) and area in real time in order to estimate the volume of an oil spill is very important for determining the oil spill response strategy and evaluating the oil spill disposal efficiency. In this article, a method is proposed to assess the volume of oil slicks by simultaneously measuring the thick oil slick thickness and area using ultrasonic inspection and image processing methods, respectively. A remotely operated vehicle (ROV), integrating two ultrasonic immersion transducers, was implemented as a platform to receive ultrasonic reflections from an oil slick. The oil slick thickness was determined by multiplying the speed of sound by the ultrasonic traveling time within the oil slick, which was calculated using the cross-correlation method. Images of the oil slick were captured by an optical camera using an airborne drone. The oil slick area was calculated by conducting image processing on images of the oil slick using the proposed image processing algorithms. Multiple measurements were performed to verify the proposed method in the laboratory experiments. The results show that the thickness, area and volume of a thick oil slick can be accurately measured with the proposed method. The method could potentially be used as an applicable tool for measuring the volume of an oil slick during an oil spill response.

Abstract Megakaryocytic leukemia 1 (MKL1) is a key transcription factor involved in non-small cell lung cancer (NSCLC) growth and metastasis. Yet, its downstream target genes, especially long non-coding RNA (lncRNA) targets, are poorly investigated. In this study, we employed lncRNA array technology to identify differentially expressed lncRNAs in NSCLC cells with or without overexpression of MKL1. Candidate lncRNAs were further explored for their clinical significance and function in NSCLC. The results showed that MKL1 promoted the expression of lncRNA SNHG18 in NSCLC cells. SNHG18 upregulation in NSCLC specimens correlated with lymph node metastasis and reduced overall survival of NSCLC patients. SNHG18 expression served as an independent prognostic factor for NSCLC. Knockdown of SNHG18 blocked MKL1-induced growth and invasion of NSCLC cells in vitro. Animal studies validated the requirement for SNHG18 in NSCLC growth and metastasis. Moreover, overexpression of SNHG18 promoted NSCLC cell proliferation and invasion. Mechanically, SNHG18 exerted its prometastatic effects on NSCLC cells through repression of miR-211-5p and induction of BRD4. Clinical evidence indicated that SNHG18 expression was negatively correlated with miR-211-5p expression in NSCLC tissues. Altogether, SNHG18 acts as a lncRNA mediator of MKL1 in NSCLC. SNHG18 facilitates NSCLC growth and metastasis by modulating the miR-211-5p/BRD4 axis. Therefore, SNHG18 may be a potential therapeutic target for the treatment of NSCLC.

AimsThis study aimed to observe the efficacy and safety of anlotinib in the treatment of small cell lung cancer (SCLC) in the real world, as first-line maintenance therapy, second-line, and above.MethodsClinical data of 109 patients with SCLC treated with anlotinib and hospitalized at The First Affiliated Hospital of Zhengzhou University from June 2018 to June 2020 were retrospectively analyzed. Analysis of short-term efficacy and survival was performed, with p

Accurate labeling is essential for supervised deep learning methods. However, it is almost impossible to accurately and manually annotate thousands of images, which results in many labeling errors for most datasets. We proposes a local label point correction (LLPC) method to improve annotation quality for edge detection and image segmentation tasks. Our algorithm contains three steps: gradient-guided point correction, point interpolation, and local point smoothing. We correct the labels of object contours by moving the annotated points to the pixel gradient peaks. This can improve the edge localization accuracy, but it also causes unsmooth contours due to the interference of image noise. Therefore, we design a point smoothing method based on local linear fitting to smooth the corrected edge. To verify the effectiveness of our LLPC, we construct a largest overlapping cervical cell edge detection dataset (CCEDD) with higher precision label corrected by our label correction method. Our LLPC only needs to set three parameters, but yields 30–40% average precision improvement on multiple networks. The qualitative and quantitative experimental results show that our LLPC can improve the quality of manual labels and the accuracy of overlapping cell edge detection. We hope that our study will give a strong boost to the development of the label correction for edge detection and image segmentation. We will release the dataset and code at: https://github.com/nachifur/LLPC.

Platinum-based cytotoxic chemotherapy is considered the standard treatment for advanced gastric cancer (GC). However, cisplatin chemoresistance often occurs with the mechanisms being not well clarified, which results in the cancer recurrence and poor survival. Ginsenoside Rg3, isolated from the Chinese Herb Panax Ginseng, is recognized as an anti-cancer agent. Herein, we aimed to reveal whether Ginsenoside Rg3 alleviates cisplatin resistance and sensitizes GC cells to cisplatin-induced apoptosis, and draw out the underlying molecular mechanism in cisplatin-resistant GC cells. The lower expression of miR-429 was found in AGSR-CDDP cells; it was also in association with cisplatin-resistance in GC cells and expression of which was restored following Ginsenoside Rg3 treatment. We also demonstrated that miR-429 made a contribution toward chemosensitivity in GC cells partly through SOX2 regulation. SOX2 was found to contribute to developing platinum resistance and was an authentic target for miR-429 in AGSR-CDDP cells. Importantly, enforced expression of SOX2 with a pcDNA3-SOX2 construct lacking the 3′-UTR miRNA binding site diminished the cytotoxic effects of miR-429 in AGSR-CDDP cells. We demonstrated that Ginsenoside Rg3 enhanced chemosensitivity in AGSR-CDDP GC cells, at least in part, through up-regulating miR-429, thereby targeting SOX2 and modulating downstream PI3K/AKT/mTOR signaling. Ginsenoside Rg3 was also found to regulate apoptosis-related genes via miR-429 in cisplatin-resistant GC cells. Ginsenoside Rg3 treatment significantly suppressed the migration rate of AGSR-CDDP GC cells, while following transfection with anti-miR-429, the anti-migratory effects of Ginsenoside Rg3 was partially abolished. This data suggested that Ginsenoside Rg3 may impede the chemoresistance and migration of GC cells mainly mediated through miR-429. We concluded that miR-429-regulated SOX2 expression was one of the main mechanisms by which Ginsenoside Rg3 dramatically promoted its anticancer effects on cisplatin-resistant GC cells. We also underscored a supporting model in which miR-429 adjusted PI3K/AKT/mTOR signaling by regulating SOX2 in cisplatin-resistant GC cells.

Panax ginseng has therapeutic effects on various inflammation-related diseases. Ginsenoside Rb3 (GRb3), a natural compound with anti-inflammatory and immunomodulatory properties, is one of the main active panaxadiol extracted from Panax ginseng. We explored whether GRb3 inhibited LPS-mediated inflammation through TLR4/NF-κB/MAPK signaling in macrophages. GRb3 attenuated NO and PGE2 production by attenuating iNOS and COX2 expression. GRb3 also suppressed pro-inflammatory cytokines levels, including IL-1β, IL-6, and TNF-α. Moreover, GRb3 administration significantly suppressed NF-κB (p65) nuclear translocation and the phosphorylation levels of p65, IκBα, JNK, p38, and ERK dose-dependently. Molecular docking demonstrated that GRb3 could dock onto the hydrophobic binding site of TLR4/MD2 complex, with a binding energy of −8.79 kcal/mol. Molecular dynamics (MD) displayed stable TLR4-MD2-GRb3 modeling. GRb3 dose-dependently inhibited LPS binding to cell membranes and blocked TLR4 expression. Surface plasmon resonance imaging (SPRi) revealed that GRb3 had an excellent binding affinity to TLR4/MD2 complex. Notably, resatorvid (TAK242), a selective TLR4 inhibitor, did not increase the repressive influence of GRb3 in RAW264.7 macrophages. Moreover, TLR4 overexpression partially reversed the repressive roles of GRb3 on the NF-κB/MAPK pathway and inflammatory mediators. Collectively, our study strongly indicated that GRb3 attenuated LPS-mediated inflammation through direct inhibition of TLR4 signaling. A novel insight into the underlying mechanism of anti-inflammatory effects of GRb3 in macrophages was confirmed.

The Guanylate binding proteins (GBPs) are a family of large GTPases and the most studied GBP family member is the guanylate binding protein 1 (GBP1). Earlier studies revealed that GBP1 expression was inflammatory cytokines-inducible, and most of the studies focused on inflammation diseases. Increasing number of cancer studies began to reveal its biological role in cancers recently, although with contradictory findings in literature. It was discovered from our earlier prostate cancer cell line models studies that when prostate cancer cells treated with either ethidium bromide or a cell cycle inhibitor flavopiridol for a long-term, the treatment-survived tumor cells experienced metabolic reprogramming toward Warburg effect pathways with greater aggressive features, and one common finding from these cells was the upregulation of GBP1. In this study, possible role of GBP1 in two independent prostate cancer lines by application of CRISR/Cas9 gene knockout (KO) technology was investigated. The GBP1 gene KO DU145 and PC3 prostate cancer cells were significantly less aggressive in vitro, with less proliferation, migration, wound healing, and colony formation capabilities, in addition to a significantly lower level of mitochondrial oxidative phosphorylation and glycolysis. At the same time, such GBP1 KO cells were significantly more sensitive to chemotherapeutic reagents. Xenograft experiments verified a significantly slower tumor growth of the GBP1 KO cells in nude mouse model. Furthermore, GBP1 protein expression in clinical prostate cancer sample revealed its aggressive clinical feature correlation and shorter overall survival association. Collectively, our results indicate a pro-survival or oncogenic role of GBP1 in prostate cancer.

Shadow features such as colour ratio, texture, and chromaticity have proved to be quite effective in shadow detection. Many shadow detection methods have been proposed on the basis of different features. However, previous works for shadow detection mainly focus on designing an effective classifier for existing shadow features, but pay less attention on the analysis of shadow features themselves. The majority of studies simply report the final shadow detection results rather than make an evaluation on each feature. Readers often do not know which features are more effective or whether these shadow features are complementary. The following problems are still unsolved: the robustness of each feature, which feature plays the most important role in a detection method, and what is the best performance that current features can reach. The purpose of this study is to answer these questions, and the authors hope that this study can offer guidance for future shadow detection algorithms via the evaluation of frequently used shadow features. Several useful and interesting conclusions are obtained after conducting extensive comparison experiments on a large dataset.

Abstract Background Hereditary diffuse gastric carcinoma (HDGC) accounts for 1–3% of all gastric carcinomas. Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors in the gastrointestinal (GI) tract but they comprise fewer than 1% of all GI malignancies. Small-cell carcinoma (SmCC) is a rare histological type of esophageal carcinoma, accounting for 0.4% to 2.8% of all esophageal tumors. Co-occurrence of SmCC with esophageal tumors caused by squamous carcinoma is also very uncommon. Although multiple primary malignancies are no longer rare in clinical practice, the simultaneous appearance of HDGC, GIST, esophageal small cell and squamous carcinoma in situ is extremely rare and very few cases have been reported. Case presentation We present a case of a 53 year-old woman with synchronous occurrence of four malignancies including HDGC, GIST, esophageal small cell- and local squamous carcinoma in situ. A total gastrectomy with D2 lymph node dissection and postoperative adjuvant chemotherapy with oxaliplatin and paclitaxel liposome were performed. After a 1-year follow-up, this patient was still in good condition with no evidence of recurrence. Conclusion This is the unique case that describes the co-existence of the aforementioned four types of neoplasm. This case demonstrates that a diagnosis of gastric cancer does not preclude the presence of other malignancies and every case should be thoroughly analyzed to avoid missing other problems, which may worsen the prognosis.

Recovering correct or at least realistic colors of underwater scenes is a challenging issue for image processing due to the unknown imaging conditions including the optical water type, scene location, illumination, and camera settings. With the assumption that the illumination of the scene is uniform, a chromatic adaptation-based color correction technology is proposed in this paper to remove the color cast using a single underwater image without any other information. First, the underwater RGB image is first linearized to make its pixel values proportional to the light intensities arrived at the pixels. Second, the illumination is estimated in a uniform chromatic space based on the white-patch hypothesis. Third, the chromatic adaptation transform is implemented in the device-independent XYZ color space. Qualitative and quantitative evaluations both show that the proposed method outperforms the other test methods in terms of color restoration, especially for the images with severe color cast. The proposed method is simple yet effective and robust, which is helpful in obtaining the in-air images of underwater scenes.

Purpose: This study aims at sharing our experience as how to obtain and identify axial view image of the acetabular anterior column in patients. Methods: Pelvic computed tomography data of six normal adults were used to reconstruct three-dimensional (3D) models. The transparency of each 3D model was downgraded at the view perpendicular to the cross section of the anterior column axis to simulate the anterior column axial view image. Fluoroscopy was performed in all patients to obtain the anterior column axial view image in the operating room. Each fluoroscopic image was compared with the corresponding simulation image to analyze potential anatomic landmarks that were helpful to identify the translucent area (projection of the screw path) in the patients. Results and Conclusions: To obtain ideal anterior column axial fluoroscopic image, the patient should be positioned supine with the leg of “abnormal side” straight and contralateral side flexion, abduction, and external rotation; the C-arm machine should be placed at the caudal end of the operation table with the C-arm fluoroscopic intensifier first positioned at the pelvic lateral view and then tilted approximately 30° toward the “abnormal side” and rotated approximately 45° toward the caudal end of the operation table. To identify the translucent area on the anterior column axial view fluoroscopic image obtained from the patient, the greater sciatic notch, the true pelvis edge, and the acetabulum should be identified first and the translucent area is located in the area surrounded by these three anatomic landmarks.

Notch signaling involves the processes that govern cell proliferation, cell fate decision, cell differentiation and stem cell maintenance. Due to its fundamental role in stem cells, it has been speculated during the recent years that Notch family may have critical functions in cancer stem cells or cancer cells with a stem cell phenotype, therefore playing an important role in the process of oncogenesis. In this study, expression of Notch family in KYSE70, KYSE140 and KYSE450 squamous esophageal cancer cell lines and virus transformed squamous esophageal epithelial cell line Het-1A was examined by quantitative RT-PCR. Compared to the Het-1A cells, higher levels of Nocth1 and Notch3 expression in the cancer cell lines were identified. Due to the finding that NOTCH3 mainly mediates squamous cell differentiation, NOTCH1 expression was further studied in these cell lines. By Western blot analyses, the KYSE70 cell line which derived from a poorly differentiated tumor highly expressed Notch1, and the Notch1 expression in this cell line was hypoxia inducible, while the KYSE450 cell line which derived from a well differentiated tumor was always negative for Notch1, even in hypoxia. Additional studies demonstrated that the KYSE70 cell line was more 5-FU resistant than the KYSE450 cell line and such 5-FU resistance is correlated to Notch1 expression verified by Notch1 knockdown experiments. In clinical samples, Notch1 protein expression was detected in the basal cells of human esophagus epithelia, and its expression in squamous cell carcinomas was significantly associated with higher pathological grade and shorter overall survival. We conclude that Notch1 expression is associated with cell aggressiveness and 5-FU drug resistance in human esophageal squamous cell carcinoma cell lines in vitro and is significantly associated with a poor survival in human esophageal squamous cell carcinomas.