RESUMO
Achieving real-time and high-accuracy 3D reconstruction of dynamic scenes is a fundamental challenge in many fields, including online monitoring, augmented reality, and so on. On one hand, traditional methods, such as Fourier transform profilometry (FTP) and phase-shifting profilometry (PSP), are struggling to balance measuring efficiency and accuracy. On the other hand, deep learning-based approaches, which offer the potential for improved accuracy, are hindered by large parameter amounts and complex structures less amenable to real-time requirements. To solve this problem, we proposed a network architecture search (NAS)-based method for real-time processing and 3D measurement of dynamic scenes with rate equivalent to single-shot. A NAS-optimized lightweight neural network was designed for efficient phase demodulation, while an improved dual-frequency strategy was employed coordinately for flexible absolute phase unwrapping. The experiment results demonstrate that our method can effectively perform 3D reconstruction with a reconstruction speed of 58fps, and realize high-accuracy measurement of dynamic scenes based on deep learning for what we believe to be the first time with the average RMS error of about 0.08 mm.
RESUMO
High-quality imaging with reduced optical complexity has been extensively investigated owing to its promising future in academic and industrial research. However, the practical performance of most imaging systems has encountered a bottleneck posed by optics rather than electronics. Here, we propose a digital lens (DL) to compensate for the chromatic aberration induced by physical optical elements, while the residual wavelength-independent degradation is tackled through a self-designed neural network. By transforming physical aberration correction to an algorithm-based computational imaging task, the proposed DL enables our framework to reduce optical complexity and achieve achromatic imaging in the analog domain. Real experiments have been conducted with an off-the-shelf single lens and recovered images show up to 14.62â dB higher peak signal-to-noise ratio (PSNR) than the original chromatic input. Furthermore, we run a comprehensive ablation study to highlight the contribution of embedding the proposed DL, which shows a 4.83â dB PSNR improvement compared with the methods without DL. Technically, the proposed method can be an alternative for future applications that require both simple optics and high-fidelity visualization.
RESUMO
Large DOF (depth-of-field) with high SNR (signal-noise-ratio) imaging is a crucial technique for applications from security monitoring to medical diagnostics. However, traditional optical design for large DOF requires a reduction in aperture size, and hence with a decrease in light throughput and SNR. In this paper, we report a computational imaging system integrating dual-aperture optics with a physics-informed dual-encoder neural network to realize prominent DOF extension. Boosted by human vision mechanism and optical imaging law, the dual-aperture imaging system is consisted of a small-aperture NIR camera to provide sharp edge and a large-aperture VIS camera to provide faithful color. To solve the imaging inverse problem in NIR-VIS fusion with different apertures, a specific network with parallel double encoders and the multi-scale fusion module is proposed to adaptively extract and learn the useful features, which contributes to preventing color deviation while preserving delicate scene textures. The proposed imaging framework is flexible and can be designed in different protos with varied optical elements for different applications. We provide theory for system design, demonstrate a prototype device, establish a real-scene dataset containing 3000 images, perform elaborate ablation studies and conduct peer comparative experiments. The experimental results demonstrate that our method effectively produces high-fidelity with larger DOF range than input raw images about 3 times. Without complex optical design and strict practical limitations, this novel, intelligent and integratable system is promising for variable vision applications such as smartphone photography, computational measurement, and medical imaging.
RESUMO
Hyperspectral imaging is being extensively investigated owing to its promising future in critical applications such as medical diagnostics, sensing, and surveillance. However, current techniques are complex with multiple alignment-sensitive components and spatiospectral parameters predetermined by manufacturers. In this paper, we demonstrate an end-to-end snapshot hyperspectral imaging technique and build a physics-informed dual attention neural network with multimodal learning. By modeling the 3D spectral cube reconstruction procedure and solving that compressive-imaging inverse problem, the hyperspectral volume can be directly recovered from only one scene RGB image. Spectra features and camera spectral sensitivity are jointly leveraged to retrieve the multiplexed spatiospectral correlations and realize hyperspectral imaging. With the help of integrated attention mechanism, useful information supplied by disparate modal components is adaptively learned and aggregated to make our network flexible for variable imaging systems. Results show that the proposed method is ultra-faster than the traditional scanning method, and 3.4 times more precise than the existing hyperspectral imaging convolutional neural network. We provide theory for network design, demonstrate training process, and present experimental results with high accuracy. Without bulky benchtop setups and strict experimental limitations, this simple and effective method offers great potential for future spectral imaging applications such as pathological digital stain, computational imaging and virtual/augmented reality display, etc.
Assuntos
Imageamento Hiperespectral , Redes Neurais de ComputaçãoRESUMO
Spectral sensitivity, as one of the most important parameters of a digital camera, is playing a key role in many computer vision applications. In this paper, a confidence voting convolutional neural network (CVNet) is proposed to rebuild the spectral sensitivity function, modeled as the sum of weighted basis functions. By evaluating useful information supplied by different image segments, disparate confidence is calculated to automatically learn basis functions' weights, only using one image captured by the object camera. Three types of basis functions are made up and employed in the network, including Fourier basis function (FBF), singular value decomposition basis function (SVDBF), and radial basis function (RBF). Results show that the accuracy of the proposed method with FBF, SVDBF, and RBF is 97.92%, 98.69%, and 99.01%, respectively. We provide theory for network design, build a dataset, demonstrate training process, and present experimental results with high precision. Without bulky benchtop setups and strict experimental limitations, this proposed simple and effective method could be an alternative in the future for spectral sensitivity function estimation.
RESUMO
Large depth-of-field (DOF) imaging with a high resolution is useful for applications ranging from robot vision to bio-imaging. However, it is challenging to construct an optical system with both a high resolution and large DOF. The common solution is to design relatively complex optical systems, but the setup of such systems is often bulky and expensive. In this paper, we propose a novel, compact, and low-cost method for large-DOF imaging. The core concept is to (1) design an aspherical lens with a depth-invariant point spread function to enable uniform image blurring over the whole depth range and (2) construct a deep learning network to reconstruct images with high fidelity computationally. The raw images captured by the aspherical lens are deblurred by the trained network, which enables large-DOF imaging at a smaller F number. Experimental results demonstrate that our end-to-end computational imager can achieve enhanced imaging performance. It can reduce loss by up to 46.5% compared to inherited raw images. With the capabilities of high-resolution and large-DOF imaging, the proposed method is promising for applications such as microscopic pathological diagnosis, virtual/augmented reality displays, and smartphone photography.
RESUMO
Spectral sensitivity is largely related to sensor imaging, which has drawn widespread attention in computer vision. Accurate estimation becomes increasingly urgent because manufacturers rarely disclose it. In this Letter, we present a novel, compact, inexpensive, and real-time computational system for snapshot spectral sensitivity estimation. A multi-scale camera based on the multi-scale convolutional neural network is first proposed, to the best of our knowledge, to automatically extract multiplexing features of an input image by multiscale deep learning, which is vital to solving the inverse problem in sensitivity estimation. Our network is flexible and can be designed with different convolutional kernel sizes for a given application. We build a dataset with 10,500 raw images and generate an excellent pre-trained model. Commercial cameras are adopted to test model validity; the results show that our system can achieve estimation accuracy as high as 91.35%. We provide a method for system design, propose a deep learning network, build a dataset, demonstrate training process, and present experimental results with high precision. This simple and effective method provides an accurate approach for precise estimation of spectral sensitivity and is suitable for computational applications such as pathological digital stain, virtual/augmented reality display, and high-quality image acquisition.
RESUMO
Phase-shifting profilometry (PSP) based on the binary defocusing technique has been widely used due to its high-speed capability. However, the required adjustment in projector defocus by traditional method is inaccurate, inflexible, and associated with fringe pitch. Instead of manual defocusing adjustment, a passive defocus of the binary patterns based on deep learning is proposed in this paper. Learning the corresponding binary patterns with a specifically designed convolutional neural network, high-quality three-step sinusoidal patterns can be generated. Experimental results demonstrate that the proposed method could reduce phase error by 80%-90% for different fringe pitches without projector defocus and outperform the traditional method by providing more accurate and robust results within a large measuring depth.
RESUMO
PURPOSE: This study aimed to investigate the correlation between mSEPT9 and tumor burden as well as the role of mSEPT9 in monitoring colorectal cancer (CRC) patients. METHODS: A total of 309 patients were recruited and received mSEPT9 detection in this retrospective study. Clinicopathologic characteristics were collected, including age, gender, differentiation, gene mutation, stage, and tumor markers. The correlation between mSEPT9 and clinical tumor burden was analyzed. A relative mSEPT9 value was determined using the ΔΔCt method. RESULTS: The overall positivity rate of mSEPT9 was 39.8% in CRC patients. mSEPT9 status was significantly associated with disease status and tumor markers (CEA and CA19-9). The mSEPT9 positivity rates were 15.6%, 50.0%, 64.4%, and 70.0% for P0M0, P1M0, P0M1, and P1M1 patients, respectively (p < 0.001). Among 137 CRC patients who received mSEPT9 assay before surgery, the pre-operation mSEPT9 positivity rate increased significantly from stage I to stage IV (Stage I vs. II vs. III vs. IV 25% vs. 59.1% vs. 57.1% vs. 70.0%, respectively). Consecutive blood samples were obtained from 26 patients during therapy. The patients with increased mSEPT9 levels showed a higher progression rate. CONCLUSIONS: mSEPT9 was a biomarker reflecting tumor burden, and serial detections of mSEPT9 could be a promising strategy for disease monitoring in CRC patients.
Assuntos
Neoplasias Colorretais , Septinas/sangue , Carga Tumoral , Adulto , Idoso , Idoso de 80 Anos ou mais , Biomarcadores Tumorais/sangue , Neoplasias Colorretais/sangue , Neoplasias Colorretais/epidemiologia , Neoplasias Colorretais/patologia , Feminino , Humanos , Masculino , Pessoa de Meia-IdadeRESUMO
AIM: To explore the prognostic value of the pretreatment platelet (PLT) count in patients undergoing transcatheter arterial chemoembolization (TACE) with hepatocellular carcinoma (HCC). MATERIALS & METHODS: We prospectively analyzed 317 hepatitis B virus-related HCC patients undergoing TACE. Time to progression (TTP) was selected to evaluate the clinical significance of PLT level in HCC patients. RESULTS: PLT was the only parameter showing statistical significance of all the clinical characteristics between two distinct tumor response groups. After ruling out cirrhosis as a potential major confounding factor, the conclusion was further established. Higher pretreatment PLT level, portal vessel invasion and higher stratification of α-fetoprotein level were independently associated with longer TTP. The prognostic score model combining the three risk factors revealed that higher risk scores might mean shorter TTP. CONCLUSION: The pretreatment PLT level is a potentially useful biomarker to predict the prognostic outcomes in HCC patients undergoing TACE and deserves to be further explored in subsequent works.
Assuntos
Plaquetas , Carcinoma Hepatocelular/terapia , Quimioembolização Terapêutica/métodos , Neoplasias Hepáticas/terapia , Adulto , Idoso , Idoso de 80 Anos ou mais , Carcinoma Hepatocelular/sangue , Carcinoma Hepatocelular/patologia , Carcinoma Hepatocelular/virologia , Progressão da Doença , Feminino , Seguimentos , Vírus da Hepatite B/isolamento & purificação , Humanos , Neoplasias Hepáticas/sangue , Neoplasias Hepáticas/patologia , Neoplasias Hepáticas/virologia , Masculino , Pessoa de Meia-Idade , Contagem de Plaquetas , Prognóstico , Estudos Prospectivos , Estudos Retrospectivos , Fatores de Tempo , Resultado do TratamentoRESUMO
Object visualization is influenced by the spectral distribution of an illuminant impinging upon it. In this paper, we proposed a color entropy evaluation method to provide the optimal illumination that best helps surgeons distinguish tissue features. The target-specific optimal illumination was obtained by maximizing the color entropy value of our sample tissue, whose spectral reflectance was measured using multispectral imaging. Sample images captured under optimal light were compared with that under commercial white light emitting diodes (3000K, 4000K and 5500K). Results showed images under the optimized illuminant had better visual performance such as more subtle details exhibited.
RESUMO
OBJECTIVE: To establish an indirect reference interval based on the test results of alanine aminotransferase stored in a laboratory information system. METHODS: All alanine aminotransferase results were included for outpatients and physical examinations that were stored in the laboratory information system of Zhongshan Hospital during 2014. The original data were transformed using a Box-Cox transformation to obtain an approximate normal distribution. Outliers were identified and omitted using the Chauvenet and Tukey methods. The indirect reference intervals were obtained by simultaneously applying nonparametric and Hoffmann methods. The reference change value was selected to determine the statistical significance of the observed differences between the calculated and published reference intervals. RESULTS: The indirect reference intervals for alanine aminotransferase of all groups were 12 to 41 U/L (male, outpatient), 12 to 48 U/L (male, physical examination), 9 to 32 U/L (female, outpatient), and 8 to 35 U/L (female, physical examination), respectively. The absolute differences when compared with the direct results were all smaller than the reference change value of alanine aminotransferase. CONCLUSION: The Box-Cox transformation combined with the Hoffmann and Tukey methods is a simple and reliable technique that should be promoted and used by clinical laboratories.
Assuntos
Exame Físico , Alanina Transaminase , Biomarcadores , Feminino , Hospitais , Humanos , Masculino , Pacientes Ambulatoriais , Valores de ReferênciaRESUMO
Ischemic stroke (IS) is a complex neurological disease that can lead to severe disability or even death. Understanding the molecular mechanisms involved in the occurrence and progression of IS is of great significance for developing effective treatment strategies. In this context, the role of neddylation refers to the potential impact of neddylation on various cellular processes, which may contribute to the pathogenesis and outcome of IS. First, differential analysis was conducted on the GSE16561 dataset from the GEO database to identify 350 differentially expressed genes (DEGs) between the IS and Control groups. By intersecting the differential genes with neddylation-related genes, 11 neddylation-related DEGs were obtained. Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analyses showed that the DEGs were mainly enriched in hematopoietic cell lineage and neutrophil degranulation, while the neddylation-related DEGs were mainly enriched in apoptosis and post-translational protein modification. Further Lasso-Cox and random forest analyses were performed on the 11 neddylation-related DEGs, identifying key genes SRPK1, BIRC2, and KLHL3. Additionally, validation of the key genes was carried out using the GSE58294 dataset and clinical patients. Finally, the correlation between the key genes and ferroptosis and cuproptosis was analyzed, and a ceRNA network was constructed. Our study helps to elucidate the complex role of neddylation in the mechanism of ischemic stroke, providing potential opportunities for the development of therapeutic interventions.
RESUMO
BACKGROUND: We have recently developed a new Coronary Artery Tree description and Lesion EvaluaTion (CatLet) angiographic scoring system. Our preliminary studies have demonstrated its superiority over the the Synergy between percutaneous coronary intervention (PCI) with Taxus and Cardiac Surgery (SYNTAX) score with respect to outcome predictions for acute myocardial infarction (AMI) patients. The current study hypothesized that the residual CatLet (rCatLet) score predicts clinical outcomes for AMI patients and that a combination with the three clinical variables (CVs)-age, creatinine, and ejection fraction, will enhance its predicting values. METHODS: The rCatLet score was calculated retrospectively in 308 consecutively enrolled patients with AMI. Primary endpoint, major adverse cardiac or cerebrovascular events (MACCE) including all-cause mortality, non-fatal AMI, transient ischemic attack/stroke, and ischemia-driven repeat revascularization, was stratified according to rCatLet score tertiles: rCatLet_low ≤3, rCatLet_mid 4-11, and rCatLet_top ≥12, respectively. Cross-validation confirmed a reasonably good agreement between the observed and predicted risks. RESULTS: Of 308 patients analyzed, the rates of MACCE, all-cause death, and cardiac death were 20.8%, 18.2%, and 15.3%, respectively. Kaplan-Meier curves for all endpoints showed increasing outcome events with the increasing tertiles of the rCatLet score, with P values <0.001 on trend test. For MACCE, all-cause death, and cardiac death, the area under the curves (AUCs) of the rCatLet score were 0.70 (95% confidence intervals [CI]: 0.63-0.78), 0.69 (95% CI: 0.61-0.77), and 0.71 (95% CI: 0.63-0.79), respectively; the AUCs of the CVs-adjusted rCatLet score models were 0.83 (95% CI: 0.78-0.89), 0.87 (95% CI: 0.82-0.92), and 0.89 (95% CI: 0.84-0.94), respectively. The performance of CVs-adjusted rCatLet score was significantly better than the stand-alone rCatLet score in terms of outcome predictions. CONCLUSION: The rCatLet score has a predicting value for clinical outcomes for AMI patients and the incorporation of the three CVs into the rCatLet score will enhance its predicting ability. TRIAL REGISTRATION: http://www.chictr.org.cn , ChiCTR-POC-17013536.
Assuntos
Doença da Artéria Coronariana , Infarto do Miocárdio , Intervenção Coronária Percutânea , Humanos , Doença da Artéria Coronariana/complicações , Morte , Infarto do Miocárdio/cirurgia , Infarto do Miocárdio/etiologia , Estudos Retrospectivos , Medição de Risco , Fatores de Risco , Resultado do TratamentoRESUMO
Circular RNA (circRNA) is a novel emerging type of endogenous regulatory non-coding RNA molecules with a covalent closed-loop configuration, which exerts important functions in multiple biological processes. CircRNAs are known to regulate gene expression as functional regulators interacting with miRNAs by sponge, which have been reported to regulate skeletal muscle development. Nevertheless, the information of circRNAs involved in regulating muscle growth and development in fish is largely unknown. Here, we first identified 312 and 511 circRNAs in skeletal muscle of juvenile and adult largemouth bass (LMB) using RNA sequencing, respectively. The differentially expressed circRNAs (DE-circRNAs) analysis showed that there are 44 DE-circRNAs at two different skeletal muscle growth stages. Six circRNAs were chosen randomly and their relative expression levels in juvenile and adult LMB were confirmed by real-time PCR, indicating that these circRNAs were existed authenticity. In addition, Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis showed that these hose genes (their linear mRNAs) of DE-circRNAs were mainly enriched in the regulation of actin cytoskeleton signaling pathways. The circRNA-miRNA interaction regulatory networks indicated that one circRNA can regulate one or more miRNA. For instance, more than 30 miRNAs were regulated by two circRNAs (circRNA389 and circRNA399). Of them, the muscle-related miRNAs including the let-7 family, miR-133 and miR-26 and so on were found acting as miRNAs sponge regulated by circRNAs, indicating the roles of circRNAs in regulating muscle growth-related genes expression. Overall, these findings will not only broaden our understanding of circRNAs regulation mechanisms underlying muscle growth and development in LMB but also provides a novel clue for further functional research in carnivorous fish genetic breeding.
Assuntos
Bass , MicroRNAs , Animais , Bass/genética , Bass/metabolismo , Perfilação da Expressão Gênica , MicroRNAs/genética , MicroRNAs/metabolismo , Músculo Esquelético/metabolismo , RNA Circular/genética , RNA Mensageiro/genéticaRESUMO
Long non-coding RNAs (lncRNAs) are a class of transcriptional RNA molecules, which play critical roles in diverse biological processes. However, little is known about the overall expression pattern and roles of lncRNAs in skeletal muscle of largemouth bass (LMB). Here, we constructed two skeletal muscle RNA libraries to find lncRNAs that may involve in the regulation of skeletal muscle development between juvenile and adult LMB. A total of 16,147 lncRNAs and 4611 differentially expressed lncRNAs were identified. Among these identified lncRNAs, 10 lncRNAs were randomly selected to confirm their expression by real-time qPCR both in libraries, which were consistent with the RNA sequencing results. The target mRNAs of lncRNAs were predicted for GO enrichment analysis. Results showed that these targets associated with growth and development of muscle, such as skeletal muscle fiber development, myoblast proliferation and differentiation. Importantly, correlation analysis of lncRNA-miRNA-mRNA regulatory network revealed that several lncRNAs targeted miRNAs which are closely involved in the regulation of muscle development. It is the first time to identify a number of lncRNA that correlate with skeletal muscle development in LMB. Our results not only provide a comprehensive expression profile of muscle lncRNAs in this species, but also provide a theoretical basis for further elaborating genetic regulation mechanism of muscle growth and development, and pave the way for the future molecular assisted breeding in carnivorous fishes.
Assuntos
Bass , MicroRNAs , RNA Longo não Codificante , Animais , Bass/genética , Bass/metabolismo , MicroRNAs/genética , MicroRNAs/metabolismo , Desenvolvimento Muscular/genética , Músculo Esquelético/metabolismo , RNA Longo não Codificante/genética , RNA Mensageiro/genéticaRESUMO
OBJECTIVE: To investigate the predictive value of PTEN and echocardiography in the treatment of heart failure with trimetazidine combined with metoprolol. METHODS: A total of 100 patients with coronary heart disease and HF who admitted to our hospital from August 2018 to August 2020 were enrolled into research. All patients received routine treatment according to the guidelines and were treated with trimetazidine and metoprolol for a total course of 6 months. Echocardiographic parameters and PTEN levels were measured at baseline and after treatment. The patients were divided into groups according to the quartile of basic PTEN level, and the total effective rates were compared. The echocardiographic parameters of patients with different prognosis were analyzed. Bivariate correlation analysis was used to evaluate the correlation between PTEN, echocardiography and treatment effect. RESULTS: Compared with that before treatment, the level of PTEN increased significantly after treatment (P < 0.01). According to the quartile of basic PTEN level, the total effective rate of patients with different levels of basic PTEN was was statistically different (P < 0.01). There was a linear correlation between the level of basic PTEN and the treatment effect, and the total effective rate of patients with high level of basic PTEN was higher than that of patients with low level of PTEN. Compared with before treatment, LVEF, SV, E/A and lvfs increased significantly after treatment (P < 0.01). There was a correlation between the basic echocardiographic parameters and the treatment effect of patients. The basic echocardiographic parameters of patients with poor prognosis were worse than those with good prognosis. PTEN expression in patients' serum was only positively correlated with E/A, but not with LVFE, SV and LVFS (P < 0.01). CONCLUSION: PTEN and echocardiographic parameters serve as a good method to evaluate the short-term therapeutic effect of trimetazidine combined with metoprolol in patients with heart failure.
RESUMO
BACKGROUND: Patient-based real-time quality control (PBRTQC) has gained attention because of its potential to detect analytical errors in situations wherein internal quality control is less effective. Multiple PBRTQC algorithms have been proposed. However, there is a lack of comprehensive comparison of the performance of PBRTQC algorithms on different types of analytical errors. Thus, a comparative study was conducted. METHODS: The performance of six different PBRTQC algorithms was evaluated on three types of analytical errors using 906,552 test results for outpatient serum sodium, chloride, alanine aminotransferase, and creatinine at the Department of Laboratory Medicine at Zhongshan Hospital, Fudan University in 2019. The performance results were compared and assessed. RESULTS: The moving average, moving median, exponentially weighted moving average, and moving quartiles performed similarly for effectively detecting constant errors (CE) and proportional errors (PE) but not random errors (RE). The moving sum of positive patients and moving standard deviation could detect RE for serum sodium and chlorides but performed poorly on detecting the CE and PE. CONCLUSIONS: This study demonstrated the importance of assessing the potential source of error of a particular analyte and the corresponding type of analytical error before choosing a quality control algorithm for implementation.
Assuntos
Algoritmos , Laboratórios , Humanos , Controle de QualidadeRESUMO
Surgical light is important for helping the surgeon easily identify specific tissues during an operation. We propose a spectral reflectance comparison model to optimize the light-emitting diode light spectrum in the operating room. An entropy evaluation method, meant specifically for surgical situations, was developed to evaluate images of biological samples. White light was mixed to achieve an optimal spectrum, and images of different tissues under the light were captured and analyzed. Results showed that images obtained with light with an optimal spectrum had a higher contrast than those obtained with a commercial white light of different color temperatures. Optimized surgical light obtained using this simple and effective method could replace the traditional surgical illumination systems.