Deep Reconstruction Transfer Convolutional Neural Network for Rolling Bearing Fault Diagnosis.

Deep transfer learning has been widely used to improve the versatility of models. In the problem of cross-domain fault diagnosis in rolling bearings, most models require that the given data have a similar distribution, which limits the diagnostic effect and generalization of the model. This paper proposes a deep reconstruction transfer convolutional neural network (DRTCNN), which satisfies the domain adaptability of the model under cross-domain conditions. Firstly, the model uses a deep reconstruction convolutional automatic encoder for feature extraction and data reconstruction. Through sharing parameters and unsupervised training, the structural information of target domain samples is effectively used to extract domain-invariant features. Secondly, a new subdomain alignment loss function is introduced to align the subdomain distribution of the source domain and the target domain, which can improve the classification accuracy by reducing the intra-class distance and increasing the inter-class distance. In addition, a label smoothing algorithm considering the credibility of the sample is introduced to train the model classifier to avoid the impact of wrong labels on the training process. Three datasets are used to verify the versatility of the model, and the results show that the model has a high accuracy and stability.

Lane Detection Algorithm in Curves Based on Multi-Sensor Fusion.

Identifying lane markings is a key technology in assisted driving and autonomous driving. The traditional sliding window lane detection algorithm has good detection performance in straight lanes and curves with small curvature, but its detection and tracking performance is poor in curves with larger curvature. Large curvature curves are common scenes in traffic roads. Therefore, in response to the problem of poor lane detection performance of traditional sliding window lane detection algorithms in large curvature curves, this article improves the traditional sliding window algorithm and proposes a sliding window lane detection calculation method, which integrates steering wheel angle sensors and binocular cameras. When a vehicle first enters a bend, the curvature of the bend is not significant. Traditional sliding window algorithms can effectively detect the lane line of the bend and provide angle input to the steering wheel, enabling the vehicle to travel along the lane line. However, as the curvature of the curve increases, traditional sliding window lane detection algorithms cannot track lane lines well. Considering that the steering wheel angle of the car does not change much during the adjacent sampling time of the video, the steering wheel angle of the previous frame can be used as input for the lane detection algorithm of the next frame. By using the steering wheel angle information, the search center of each sliding window can be predicted. If the number of white pixels within the rectangular range centered around the search center is greater than the threshold, the average of the horizontal coordinate values of these white pixels will be used as the horizontal coordinate value of the sliding window center. Otherwise, the search center will be used as the center of the sliding window. A binocular camera is used to assist in locating the position of the first sliding window. The simulation and experimental results show that compared with traditional sliding window lane detection algorithms, the improved algorithm can better recognize and track lane lines with large curvature in bends.

Blind and low-complexity modulation format identification based on signal envelope flatness for autonomous digital coherent receivers.

Modulation format identification (MFI) is a critical technology for autonomous digital coherent receivers in next-generation elastic optical networks. A novel and simple MFI scheme, to the best of our knowledge, based on signal envelope flatness is proposed without requiring any training or other prior information. After amplitude normalization and partition, the incoming polarization division multiplexed (PDM) signals can be classified into quadrature phase shift keying (QPSK), 8 quadrature amplitude modulation (QAM), 16QAM, and 64QAM signals according to envelope flatnesses R1, R2, and R3 of signals in different amplitude ranges. The feasibility of the proposed MFI scheme is first verified via numerical simulations with 28 GBaud PDM-QPSK/-8QAM/-16QAM/-64QAM signals. Only by using 4000 symbols can the proposed MFI scheme achieve a 100% correct identification rate for the four modulation formats over a wide optical signal-to-noise ratio (OSNR) range. Proof-of-concept experiments among 28 GBaud PDM-QPSK/-8QAM/-16QAM systems under back-to-back and long-haul fiber transmission links are implemented to further demonstrate the effectiveness of the proposed MFI scheme. The experimental results show that the proposed MFI scheme can obtain a 100% correct identification rate when the OSNR value of each modulation format is higher than the threshold corresponding to 7% FEC and is resilient towards fiber nonlinearities. More importantly, the proposed MFI scheme can significantly reduce computational complexity.

Long Noncoding RNA LINC01703 Exacerbates the Malignant Properties of Non-Small Cell Lung Cancer by Upregulating MACC1 in a MicroRNA-605-3p-Mediated Manner.

Long intergenic nonprotein-coding RNA 1703 (LINC01703) has diagnostic significance in lung adenocarcinoma. However, its specific roles in non-small cell lung cancer (NSCLC) and downstream mechanisms have not been investigated. In the current study, we characterized the role of LINC01703 in NSCLC malignancy and elucidated its detailed mechanism of action. LINC01703 expression was measured by qRT-PCR. The regulatory effects of LINC01703 on the malignancy of NSCLC cells were assessed by multiple functional experiments. The targeted interaction was confirmed by RNA immunoprecipitation and luciferase reporter assays. Herein, overexpression of LINC01703 in NSCLC was indicated in the TCGA database and further proven in our cohort. Functional studies revealed that knocking down LINC01703 repressed cell proliferation, colony formation, migration, and invasion in vitro, which was accompanied by the induction of apoptosis. The tumor growth of LINC01703-silenced cells was also inhibited in vivo. Mechanistic analyses revealed that LINC01703 functioned as a competing endogenous RNA for microRNA-605-3p (miR-605-3p) in NSCLC cells, which thereby upregulated the miR-605-3p target metastasis associated with colon cancer 1 (MACC1). Rescue experiments highlighted that the regulatory actions of LINC01703 ablation on NSCLC cells were abolished in response to miR-605-3p downregulation or MACC1 overexpression. In conclusion, LINC01703 enhanced the aggressiveness of NSCLC cells by altering miR-605-3p/MACC1. Our work suggests the therapeutic potential of LINC01703/miR-605-3p/MACC1 in NSCLC.

Coronary artery fistula and lung adenocarcinoma: a case report and literature review.

Coronary artery fistula (CAF) is a rare condition, whilst lung cancer is one of the most common malignant tumors worldwide. We came cross an interesting case with both diseases. To the best of our knowledge, this is the first case report pertaining to a patient with a coexisting CAF and lung adenocarcinoma. The patient was a 67-year-old woman who was admitted to our hospital for evaluation of persistent cough. Through the examination she was diagnosed coronary artery fistula and lung adenocarcinoma. Both diseases were successfully treated in a single operation (artery ligation and pulmonary lobectomy). The post-operative period was uneventful. At 3-month follow-up, there were no signs of blood shunting or cancer recurrence. There is no standard guidelines to treat both diseases. We want to seek out a solution to the problem. In this patient, we successfully performed artery ligation and pulmonary lobectomy in a single operation without any complications. We believe the treatment of patients with CAFs should be individualized. But, there is still a lot of shortcomings in our research. First of all, we have no enough cases to support our approach. What's more, the long-term effects of the operation are not certain. Last but not least, we have no proof in genetics with both diseases.

Mutation screening of mismatch repair gene Mlh3 in familial esophageal cancer.

AIM: To shed light on the possible role of mismatch repair gene Mlh3 in familial esophageal cancer (FEC). METHODS: A total of 66 members from 10 families suggestive of a genetic predisposition to hereditary esophageal cancer were screened for germline mutations in Mlh3 with denaturing high performance liquid chromatography (DHPLC), a newly developed method of comparative sequencing based on heteroduplex detection. For all samples exhibiting abnormal DHPLC profiles, sequence changes were evaluated by cycle sequencing. For any mutation in family members, we conducted a segregation study to compare its prevalence in sporadic esophageal cancer patients and normal controls. RESULTS: Exons of Mlh3 in all samples were successfully examined. Overall, 4 missense mutations and 3 polymorphisms were identified in 4 families. Mlh3 missense mutations in families 9 and 10 might be pathogenic, but had a reduced penetrance. While in families 1 and 7, there was no sufficient evidence supporting the monogenic explanations of esophageal cancers in families. The mutations were found in 33% of high-risk families and 50% of low-risk families. CONCLUSION: Mlh3 is a high risk gene with a reduced penetrance in some families. However, it acts as a low risk gene for esophageal cancer in most families. Mutations of Mlh3 may work together with other genes in an accumulated manner and result in an increased risk of esophageal tumor. DHPLC is a robust and sensitive technique for screening gene mutations.