Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 43.818
Filtrar
1.
IEEE Trans Image Process ; 30: 7636-7648, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34469297

RESUMO

Convolutional neural networks are capable of extracting powerful representations for face recognition. However, they tend to suffer from poor generalization due to imbalanced data distributions where a small number of classes are over-represented (e.g. frontal or non-occluded faces) and some of the remaining rarely appear (e.g. profile or heavily occluded faces). This is the reason why the performance is dramatically degraded in minority classes. For example, this issue is serious for recognizing masked faces in the scenario of ongoing pandemic of the COVID-19. In this work, we propose an Attention Augmented Network, called AAN-Face, to handle this issue. First, an attention erasing (AE) scheme is proposed to randomly erase units in attention maps. This well prepares models towards occlusions or pose variations. Second, an attention center loss (ACL) is proposed to learn a center for each attention map, so that the same attention map focuses on the same facial part. Consequently, discriminative facial regions are emphasized, while useless or noisy ones are suppressed. Third, the AE and the ACL are incorporated to form the AAN-Face. Since the discriminative parts are randomly removed by the AE, the ACL is encouraged to learn different attention centers, leading to the localization of diverse and complementary facial parts. Comprehensive experiments on various test datasets, especially on masked faces, demonstrate that our AAN-Face models outperform the state-of-the-art methods, showing the importance and effectiveness.


Assuntos
Reconhecimento Facial Automatizado/métodos , Face/anatomia & histologia , Processamento de Imagem Assistida por Computador/métodos , Redes Neurais de Computação , COVID-19 , Humanos , Máscaras
2.
Nat Commun ; 12(1): 4712, 2021 08 05.
Artigo em Inglês | MEDLINE | ID: mdl-34354073

RESUMO

Single-pixel holography (SPH) is capable of generating holographic images with rich spatial information by employing only a single-pixel detector. Thanks to the relatively low dark-noise production, high sensitivity, large bandwidth, and cheap price of single-pixel detectors in comparison to pixel-array detectors, SPH is becoming an attractive imaging modality at wavelengths where pixel-array detectors are not available or prohibitively expensive. In this work, we develop a high-throughput single-pixel compressive holography with a space-bandwidth-time product (SBP-T) of 41,667 pixels/s, realized by enabling phase stepping naturally in time and abandoning the need for phase-encoded illumination. This holographic system is scalable to provide either a large field of view (~83 mm2) or a high resolution (5.80 µm × 4.31 µm). In particular, high-resolution holographic images of biological tissues are presented, exhibiting rich contrast in both amplitude and phase. This work is an important step towards multi-spectrum imaging using a single-pixel detector in biophotonics.


Assuntos
Holografia/métodos , Animais , Encéfalo/anatomia & histologia , Compressão de Dados/métodos , Compressão de Dados/estatística & dados numéricos , Feminino , Holografia/instrumentação , Holografia/estatística & dados numéricos , Processamento de Imagem Assistida por Computador/métodos , Processamento de Imagem Assistida por Computador/estatística & dados numéricos , Camundongos , Camundongos Endogâmicos C57BL , Camundongos Nus , Dispositivos Ópticos , Imagem Óptica/instrumentação , Imagem Óptica/métodos , Imagem Óptica/estatística & dados numéricos , Fenômenos Ópticos , Cauda/anatomia & histologia
3.
Sci Rep ; 11(1): 16201, 2021 08 10.
Artigo em Inglês | MEDLINE | ID: mdl-34376765

RESUMO

Optical spectroscopic techniques have been commonly used to detect the presence of biofilm-forming pathogens (bacteria and fungi) in the agro-food industry. Recently, near-infrared (NIR) spectroscopy revealed that it is also possible to detect the presence of viruses in animal and vegetal tissues. Here we report a platform based on visible and NIR (VNIR) hyperspectral imaging for non-contact, reagent free detection and quantification of laboratory-engineered viral particles in fluid samples (liquid droplets and dry residue) using both partial least square-discriminant analysis and artificial feed-forward neural networks. The detection was successfully achieved in preparations of phosphate buffered solution and artificial saliva, with an equivalent pixel volume of 4 nL and lowest concentration of 800 TU·[Formula: see text]L-1. This method constitutes an innovative approach that could be potentially used at point of care for rapid mass screening of viral infectious diseases and monitoring of the SARS-CoV-2 pandemic.


Assuntos
Processamento de Imagem Assistida por Computador/métodos , Infecções por Lentivirus/diagnóstico , Técnicas de Diagnóstico Molecular/métodos , Espectroscopia de Luz Próxima ao Infravermelho/métodos , Células HEK293 , Humanos , Processamento de Imagem Assistida por Computador/normas , Lentivirus/isolamento & purificação , Lentivirus/patogenicidade , Infecções por Lentivirus/virologia , Técnicas de Diagnóstico Molecular/normas , Sistemas Automatizados de Assistência Junto ao Leito , Saliva/virologia , Sensibilidade e Especificidade , Espectroscopia de Luz Próxima ao Infravermelho/normas
5.
BMC Plant Biol ; 21(1): 398, 2021 Aug 25.
Artigo em Inglês | MEDLINE | ID: mdl-34433428

RESUMO

BACKGROUND: The root distribution in the soil is one of the elements that comprise the root system architecture (RSA). In monocots, RSA comprises radicle and crown roots, each of which can be basically represented by a single curve with lateral root branches or approximated using a polyline. Moreover, RSA vectorization (polyline conversion) is useful for RSA phenotyping. However, a robust software that can enable RSA vectorization while using noisy three-dimensional (3D) volumes is unavailable. RESULTS: We developed RSAtrace3D, which is a robust 3D RSA vectorization software for monocot RSA phenotyping. It manages the single root (radicle or crown root) as a polyline (a vector), and the set of the polylines represents the entire RSA. RSAtrace3D vectorizes root segments between the two ends of a single root. By utilizing several base points on the root, RSAtrace3D suits noisy images if it is difficult to vectorize it using only two end nodes of the root. Additionally, by employing a simple tracking algorithm that uses the center of gravity (COG) of the root voxels to determine the tracking direction, RSAtrace3D efficiently vectorizes the roots. Thus, RSAtrace3D represents the single root shape more precisely than straight lines or spline curves. As a case study, rice (Oryza sativa) RSA was vectorized from X-ray computed tomography (CT) images, and RSA traits were calculated. In addition, varietal differences in RSA traits were observed. The vector data were 32,000 times more compact than raw X-ray CT images. Therefore, this makes it easier to share data and perform re-analyses. For example, using data from previously conducted studies. For monocot plants, the vectorization and phenotyping algorithm are extendable and suitable for numerous applications. CONCLUSIONS: RSAtrace3D is an RSA vectorization software for 3D RSA phenotyping for monocots. Owing to the high expandability of the RSA vectorization and phenotyping algorithm, RSAtrace3D can be applied not only to rice in X-ray CT images but also to other monocots in various 3D images. Since this software is written in Python language, it can be easily modified and will be extensively applied by researchers in this field.


Assuntos
Oryza/anatomia & histologia , Oryza/crescimento & desenvolvimento , Fenótipo , Raízes de Plantas/anatomia & histologia , Raízes de Plantas/crescimento & desenvolvimento , Software , Algoritmos , Produtos Agrícolas/anatomia & histologia , Produtos Agrícolas/crescimento & desenvolvimento , Processamento de Imagem Assistida por Computador/métodos , Imageamento Tridimensional/métodos
6.
PLoS One ; 16(8): e0255886, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34388187

RESUMO

BACKGROUND: The COVID-19 pandemic has exposed the vulnerability of healthcare services worldwide, especially in underdeveloped countries. There is a clear need to develop novel computer-assisted diagnosis tools to provide rapid and cost-effective screening in places where massive traditional testing is not feasible. Lung ultrasound is a portable, easy to disinfect, low cost and non-invasive tool that can be used to identify lung diseases. Computer-assisted analysis of lung ultrasound imagery is a relatively recent approach that has shown great potential for diagnosing pulmonary conditions, being a viable alternative for screening and diagnosing COVID-19. OBJECTIVE: To evaluate and compare the performance of deep-learning techniques for detecting COVID-19 infections from lung ultrasound imagery. METHODS: We adapted different pre-trained deep learning architectures, including VGG19, InceptionV3, Xception, and ResNet50. We used the publicly available POCUS dataset comprising 3326 lung ultrasound frames of healthy, COVID-19, and pneumonia patients for training and fine-tuning. We conducted two experiments considering three classes (COVID-19, pneumonia, and healthy) and two classes (COVID-19 versus pneumonia and COVID-19 versus non-COVID-19) of predictive models. The obtained results were also compared with the POCOVID-net model. For performance evaluation, we calculated per-class classification metrics (Precision, Recall, and F1-score) and overall metrics (Accuracy, Balanced Accuracy, and Area Under the Receiver Operating Characteristic Curve). Lastly, we performed a statistical analysis of performance results using ANOVA and Friedman tests followed by post-hoc analysis using the Wilcoxon signed-rank test with the Holm's step-down correction. RESULTS: InceptionV3 network achieved the best average accuracy (89.1%), balanced accuracy (89.3%), and area under the receiver operating curve (97.1%) for COVID-19 detection from bacterial pneumonia and healthy lung ultrasound data. The ANOVA and Friedman tests found statistically significant performance differences between models for accuracy, balanced accuracy and area under the receiver operating curve. Post-hoc analysis showed statistically significant differences between the performance obtained with the InceptionV3-based model and POCOVID-net, VGG19-, and ResNet50-based models. No statistically significant differences were found in the performance obtained with InceptionV3- and Xception-based models. CONCLUSIONS: Deep learning techniques for computer-assisted analysis of lung ultrasound imagery provide a promising avenue for COVID-19 screening and diagnosis. Particularly, we found that the InceptionV3 network provides the most promising predictive results from all AI-based techniques evaluated in this work. InceptionV3- and Xception-based models can be used to further develop a viable computer-assisted screening tool for COVID-19 based on ultrasound imagery.


Assuntos
COVID-19/diagnóstico por imagem , Aprendizado Profundo , Processamento de Imagem Assistida por Computador/métodos , Pulmão/diagnóstico por imagem , Ultrassonografia/métodos , Humanos
7.
Theranostics ; 11(16): 8027-8042, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34335978

RESUMO

Prostate cancer (PCa) is one of the most frequently diagnosed malignancies of men in the world. Due to a variety of treatment options in different risk groups, proper diagnostic and risk stratification is pivotal in treatment of PCa. The development of precise medical imaging procedures simultaneously to improvements in big data analysis has led to the establishment of radiomics - a computer-based method of extracting and analyzing image features quantitatively. This approach bears the potential to assess and improve PCa detection, tissue characterization and clinical outcome prediction. This article gives an overview on the current aspects of methodology and systematically reviews available literature on radiomics in PCa patients, showing its potential for personalized therapy approaches. The qualitative synthesis includes all imaging modalities and focuses on validated studies, putting forward future directions.


Assuntos
Diagnóstico por Imagem/tendências , Processamento de Imagem Assistida por Computador/métodos , Neoplasias da Próstata/diagnóstico por imagem , Humanos , Masculino , Medicina de Precisão/métodos , Medicina de Precisão/tendências
9.
Zool Res ; 42(4): 492-501, 2021 Jul 18.
Artigo em Inglês | MEDLINE | ID: mdl-34235898

RESUMO

Fish morphological phenotypes are important resources in artificial breeding, functional gene mapping, and population-based studies in aquaculture and ecology. Traditional morphological measurement of phenotypes is rather expensive in terms of time and labor. More importantly, manual measurement is highly dependent on operational experience, which can lead to subjective phenotyping results. Here, we developed 3DPhenoFish software to extract fish morphological phenotypes from three-dimensional (3D) point cloud data. Algorithms for background elimination, coordinate normalization, image segmentation, key point recognition, and phenotype extraction were developed and integrated into an intuitive user interface. Furthermore, 18 key points and traditional 2D morphological traits, along with 3D phenotypes, including area and volume, can be automatically obtained in a visualized manner. Intuitive fine-tuning of key points and customized definitions of phenotypes are also allowed in the software. Using 3DPhenoFish, we performed high-throughput phenotyping for four endemic Schizothoracinae species, including Schizopygopsis younghusbandi, Oxygymnocypris stewartii, Ptychobarbus dipogon, and Schizothorax oconnori. Results indicated that the morphological phenotypes from 3DPhenoFish exhibited high linear correlation (>0.94) with manual measurements and offered informative traits to discriminate samples of different species and even for different populations of the same species. In summary, we developed an efficient, accurate, and customizable tool, 3DPhenoFish, to extract morphological phenotypes from point cloud data, which should help overcome traditional challenges in manual measurements. 3DPhenoFish can be used for research on morphological phenotypes in fish, including functional gene mapping, artificial selection, and conservation studies. 3DPhenoFish is an open-source software and can be downloaded for free at https://github.com/lyh24k/3DPhenoFish/tree/master.


Assuntos
Peixes/anatomia & histologia , Processamento de Imagem Assistida por Computador/métodos , Imageamento Tridimensional/veterinária , Software , Animais , Peixes/classificação , Imageamento Tridimensional/métodos , Reprodutibilidade dos Testes , Especificidade da Espécie
10.
Nucleic Acids Res ; 49(13): 7292-7297, 2021 07 21.
Artigo em Inglês | MEDLINE | ID: mdl-34197605

RESUMO

Detection of diffraction-limited spots in single-molecule microscopy images is traditionally performed with mathematical operators designed for idealized spots. This process requires manual tuning of parameters that is time-consuming and not always reliable. We have developed deepBlink, a neural network-based method to detect and localize spots automatically. We demonstrate that deepBlink outperforms other state-of-the-art methods across six publicly available datasets containing synthetic and experimental data.


Assuntos
Processamento de Imagem Assistida por Computador/métodos , Redes Neurais de Computação , Software , Microscopia
11.
Lancet Digit Health ; 3(8): e486-e495, 2021 08.
Artigo em Inglês | MEDLINE | ID: mdl-34325853

RESUMO

BACKGROUND: Medical artificial intelligence (AI) has entered the clinical implementation phase, although real-world performance of deep-learning systems (DLSs) for screening fundus disease remains unsatisfactory. Our study aimed to train a clinically applicable DLS for fundus diseases using data derived from the real world, and externally test the model using fundus photographs collected prospectively from the settings in which the model would most likely be adopted. METHODS: In this national real-world evidence study, we trained a DLS, the Comprehensive AI Retinal Expert (CARE) system, to identify the 14 most common retinal abnormalities using 207 228 colour fundus photographs derived from 16 clinical settings with different disease distributions. CARE was internally validated using 21 867 photographs and externally tested using 18 136 photographs prospectively collected from 35 real-world settings across China where CARE might be adopted, including eight tertiary hospitals, six community hospitals, and 21 physical examination centres. The performance of CARE was further compared with that of 16 ophthalmologists and tested using datasets with non-Chinese ethnicities and previously unused camera types. This study was registered with ClinicalTrials.gov, NCT04213430, and is currently closed. FINDINGS: The area under the receiver operating characteristic curve (AUC) in the internal validation set was 0·955 (SD 0·046). AUC values in the external test set were 0·965 (0·035) in tertiary hospitals, 0·983 (0·031) in community hospitals, and 0·953 (0·042) in physical examination centres. The performance of CARE was similar to that of ophthalmologists. Large variations in sensitivity were observed among the ophthalmologists in different regions and with varying experience. The system retained strong identification performance when tested using the non-Chinese dataset (AUC 0·960, 95% CI 0·957-0·964 in referable diabetic retinopathy). INTERPRETATION: Our DLS (CARE) showed satisfactory performance for screening multiple retinal abnormalities in real-world settings using prospectively collected fundus photographs, and so could allow the system to be implemented and adopted for clinical care. FUNDING: This study was funded by the National Key R&D Programme of China, the Science and Technology Planning Projects of Guangdong Province, the National Natural Science Foundation of China, the Natural Science Foundation of Guangdong Province, and the Fundamental Research Funds for the Central Universities. TRANSLATION: For the Chinese translation of the abstract see Supplementary Materials section.


Assuntos
Aprendizado Profundo , Sistemas Especialistas , Processamento de Imagem Assistida por Computador/métodos , Programas de Rastreamento/métodos , Modelos Biológicos , Retina , Doenças Retinianas/diagnóstico , Área Sob a Curva , Inteligência Artificial , Tecnologia Biomédica , China , Retinopatia Diabética/diagnóstico , Fundo de Olho , Humanos , Oftalmologistas , Fotografação , Curva ROC
12.
Methods Mol Biol ; 2350: 267-287, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34331291

RESUMO

The UltraPlex method for multiplexed two-dimensional fluorescent immunohistochemistry is described, in which hapten tags conjugated to primary antibodies facilitate multiplexed imaging of four or more antigens per tissue section at once. Anti-hapten secondary antibodies labeled with fluorophores provide amplified signal for detection, which is accomplished using a standard fluorescent microscope or digital slide scanner. The protocol is rapid and straightforward and utilizes conventionally prepared tissue samples. The resulting staining is highly sensitive and specific, enabling high-resolution imaging of multiple cellular subtypes within tissue samples. Tumor cells and tumor-infiltrating lymphocytes are presented as examples. Multiple 4-plex-stained tissue samples can be digitally overlaid to create 8-plex (or more) high-content images, enabling visualization of distribution of complex cellular subtypes across tissues.


Assuntos
Imunofluorescência , Haptenos , Imuno-Histoquímica/métodos , Biomarcadores , Biomarcadores Tumorais , Análise de Dados , Humanos , Processamento de Imagem Assistida por Computador/métodos , Neoplasias/metabolismo , Neoplasias/patologia , Coloração e Rotulagem
13.
Nat Methods ; 18(8): 953-958, 2021 08.
Artigo em Inglês | MEDLINE | ID: mdl-34312564

RESUMO

Unbiased quantitative analysis of macroscopic biological samples demands fast imaging systems capable of maintaining high resolution across large volumes. Here we introduce RAPID (rapid autofocusing via pupil-split image phase detection), a real-time autofocus method applicable in every widefield-based microscope. RAPID-enabled light-sheet microscopy reliably reconstructs intact, cleared mouse brains with subcellular resolution, and allowed us to characterize the three-dimensional (3D) spatial clustering of somatostatin-positive neurons in the whole encephalon, including densely labeled areas. Furthermore, it enabled 3D morphological analysis of microglia across the entire brain. Beyond light-sheet microscopy, we demonstrate that RAPID maintains high image quality in various settings, from in vivo fluorescence imaging to 3D tracking of fast-moving organisms. RAPID thus provides a flexible autofocus solution that is suitable for traditional automated microscopy tasks as well as for quantitative analysis of large biological specimens.


Assuntos
Encéfalo/anatomia & histologia , Processamento de Imagem Assistida por Computador/métodos , Imageamento Tridimensional/métodos , Microglia/citologia , Microscopia de Fluorescência/métodos , Animais , Masculino , Camundongos
14.
Nat Methods ; 18(7): 799-805, 2021 07.
Artigo em Inglês | MEDLINE | ID: mdl-34226721

RESUMO

A growing appreciation of the importance of cellular metabolism and revelations concerning the extent of cell-cell heterogeneity demand metabolic characterization of individual cells. We present SpaceM, an open-source method for in situ single-cell metabolomics that detects >100 metabolites from >1,000 individual cells per hour, together with a fluorescence-based readout and retention of morpho-spatial features. We validated SpaceM by predicting the cell types of cocultured human epithelial cells and mouse fibroblasts. We used SpaceM to show that stimulating human hepatocytes with fatty acids leads to the emergence of two coexisting subpopulations outlined by distinct cellular metabolic states. Inducing inflammation with the cytokine interleukin-17A perturbs the balance of these states in a process dependent on NF-κB signaling. The metabolic state markers were reproduced in a murine model of nonalcoholic steatohepatitis. We anticipate SpaceM to be broadly applicable for investigations of diverse cellular models and to democratize single-cell metabolomics.


Assuntos
Processamento de Imagem Assistida por Computador/métodos , Metabolômica/métodos , Análise de Célula Única/métodos , Espectrometria de Massas por Ionização e Dessorção a Laser Assistida por Matriz/métodos , Animais , Técnicas de Cocultura , Células Epiteliais , Ácidos Graxos/farmacologia , Hepatócitos/efeitos dos fármacos , Hepatócitos/metabolismo , Humanos , Inflamação/metabolismo , Interleucina-17/metabolismo , Masculino , Camundongos , Camundongos Endogâmicos C57BL , NF-kappa B/metabolismo , Células NIH 3T3 , Hepatopatia Gordurosa não Alcoólica/metabolismo , Hepatopatia Gordurosa não Alcoólica/patologia , Transdução de Sinais , Estresse Fisiológico
15.
Methods Mol Biol ; 2350: 239-251, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34331289

RESUMO

Lifetime multiplexed imaging refers to the simultaneous labeling of different structures with fluorescent probes that present identical photoluminescence spectra and distinct fluorescence lifetimes. This technique allows extracting quantitative information from multichannel in vivo fluorescence imaging. In vivo lifetime multiplexed imaging requires fluorophores with excitation and emission bands in the near-infrared (NIR) and tunable fluorescence lifetimes, plus an imaging system capable of time-resolved image acquisition and analysis.


Assuntos
Nanopartículas , Imagem Óptica/métodos , Espectroscopia de Luz Próxima ao Infravermelho/métodos , Algoritmos , Animais , Feminino , Corantes Fluorescentes/química , Processamento de Imagem Assistida por Computador/métodos , Camundongos , Nanopartículas/química , Imagem Óptica/instrumentação , Espectroscopia de Luz Próxima ao Infravermelho/instrumentação
16.
Methods Mol Biol ; 2350: 299-312, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34331293

RESUMO

Infrared (IR) and Raman spectroscopies are being increasingly employed for the label-free analysis of biochemical samples. Both are vibrational imaging techniques, but they provide complementary information about the chemical composition of the sample, and thus the integration of Raman and IR images leads to a comprehensive understanding of the samples. Here, we summarize the steps needed for performing multiplexed infrared and Raman imaging, identifying and overcoming the two main challenges: first, the technical difficulties caused by the incompatibilities of the techniques and, second, the necessity to extract the information from the large number of vibrational variables found in both IR and Raman spectra.


Assuntos
Espectroscopia de Infravermelho com Transformada de Fourier/métodos , Análise Espectral Raman/métodos , Animais , Humanos , Processamento de Imagem Assistida por Computador/métodos , Imuno-Histoquímica , Software
17.
Methods Mol Biol ; 2350: 313-329, 2021.
Artigo em Inglês | MEDLINE | ID: mdl-34331294

RESUMO

We describe a multiplexed imaging mass spectrometry approach especially suitable for fibrosis research. Fibrosis is a process characterized by excessive extracellular matrix (ECM) secretion. Buildup of ECM impairs tissue and organ function to promote further progression of disease. It is an ongoing analytical challenge to access ECM for diagnosis and therapeutic treatment of fibrosis. Recently, we reported the use of the enzyme collagenase type III to target the ECM proteome in thin histological tissue sections of fibrotic diseases including hepatocellular carcinoma, breast cancer, prostate cancer, lung cancer and aortic valve stenosis. Detection of collagenase type III peptides by matrix-assisted laser desorption/ionization (MALDI) imaging mass spectrometry (IMS) allows localization of ECM peptide sequences to specific regions of fibrosis. We have further identified that the ECM proteome accessed by collagenase type III has on average 3.7 sites per protein that may be differentially N-glycosylated. N-glycosylation is a major posttranslational modification of the ECM proteome, influencing protein folding, secretion, and organization. Understanding both N-glycosylation signaling and regulation of ECM expression significantly informs on ECM signaling in fibrosis.


Assuntos
Biomarcadores , Matriz Extracelular/metabolismo , Histocitoquímica/métodos , Espectrometria de Massas/métodos , Polissacarídeos/metabolismo , Fibrose/metabolismo , Fibrose/patologia , Glicosilação , Processamento de Imagem Assistida por Computador/métodos , Imuno-Histoquímica/métodos , Peptídeos/metabolismo , Processamento de Proteína Pós-Traducional , Proteômica/métodos , Pesquisa , Espectrometria de Massas por Ionização e Dessorção a Laser Assistida por Matriz/métodos , Fluxo de Trabalho
18.
Curr Opin Ophthalmol ; 32(5): 459-467, 2021 Sep 01.
Artigo em Inglês | MEDLINE | ID: mdl-34324454

RESUMO

PURPOSE OF REVIEW: The development of deep learning (DL) systems requires a large amount of data, which may be limited by costs, protection of patient information and low prevalence of some conditions. Recent developments in artificial intelligence techniques have provided an innovative alternative to this challenge via the synthesis of biomedical images within a DL framework known as generative adversarial networks (GANs). This paper aims to introduce how GANs can be deployed for image synthesis in ophthalmology and to discuss the potential applications of GANs-produced images. RECENT FINDINGS: Image synthesis is the most relevant function of GANs to the medical field, and it has been widely used for generating 'new' medical images of various modalities. In ophthalmology, GANs have mainly been utilized for augmenting classification and predictive tasks, by synthesizing fundus images and optical coherence tomography images with and without pathologies such as age-related macular degeneration and diabetic retinopathy. Despite their ability to generate high-resolution images, the development of GANs remains data intensive, and there is a lack of consensus on how best to evaluate the outputs produced by GANs. SUMMARY: Although the problem of artificial biomedical data generation is of great interest, image synthesis by GANs represents an innovation with yet unclear relevance for ophthalmology.


Assuntos
Aprendizado Profundo , Processamento de Imagem Assistida por Computador , Redes Neurais de Computação , Oftalmologia , Inteligência Artificial , Humanos , Processamento de Imagem Assistida por Computador/métodos
19.
Molecules ; 26(12)2021 Jun 15.
Artigo em Inglês | MEDLINE | ID: mdl-34203597

RESUMO

We used Raman micro-spectroscopy technique to analyze the molecular changes associated with oral squamous cell carcinoma (SCC) cells in the form of frozen tissue. Previously, Raman micro-spectroscopy technique on human tissue was mainly based on spectral analysis, but we worked on imaging of molecular structure. In this study, we evaluated the distribution of four components at the cell level (about 10 µm) to describe the changes in protein and molecular structures of protein belonging to malignant tissue. We analyzed ten oral SCC samples of five patients without special pretreatments of the use of formaldehyde. We obtained cell level images of the oral SCC cells at various components (peak at 935 cm-1: proline and valine, 1004 cm-1: phenylalanine, 1223 cm-1: nucleic acids, and 1650 cm-1: amide I). These mapping images of SCC cells showed the distribution of nucleic acids in the nuclear areas; meanwhile, proline and valine, phenylalanine, and amide I were detected in the cytoplasm areas of the SCC cells. Furthermore, the peak of amide I in the cancer area shifts to the higher wavenumber side, which indicates the α-helix component may decrease in its relative amounts of protein in the ß-sheet or random coil conformation. Imaging of SCC cells with Raman micro-spectroscopy technique indicated that such a new observation of cancer cells is useful for analyzing the detailed distribution of various molecular conformation within SCC cells.


Assuntos
Análise Espectral Raman/métodos , Carcinoma de Células Escamosas de Cabeça e Pescoço/diagnóstico por imagem , Carcinoma de Células Escamosas/diagnóstico por imagem , Carcinoma de Células Escamosas/patologia , Núcleo Celular/metabolismo , Citoplasma/metabolismo , Diagnóstico por Imagem/métodos , Humanos , Processamento de Imagem Assistida por Computador/métodos , Japão , Conformação Molecular , Neoplasias Bucais/diagnóstico por imagem , Neoplasias Bucais/metabolismo , Carcinoma de Células Escamosas de Cabeça e Pescoço/metabolismo , Carcinoma de Células Escamosas de Cabeça e Pescoço/patologia
20.
Nat Commun ; 12(1): 4340, 2021 07 15.
Artigo em Inglês | MEDLINE | ID: mdl-34267207

RESUMO

Scattering in biological tissues is a major barrier for in vivo optical imaging of all but the most superficial structures. Progress toward overcoming the distortions caused by scattering in turbid media has been made by shaping the excitation wavefront to redirect power into a single point in the imaging plane. However, fast, non-invasive determination of the required wavefront compensation remains challenging. Here, we introduce a quickly converging algorithm for non-invasive scattering compensation, termed DASH, in which holographic phase stepping interferometry enables new phase information to be updated after each measurement. This leads to rapid improvement of the wavefront correction, forming a focus after just one measurement iteration and achieving an order of magnitude higher signal enhancement at this stage than the previous state-of-the-art. Using DASH, we demonstrate two-photon fluorescence imaging of microglia cells in highly turbid mouse hippocampal tissue down to a depth of 530 µm.


Assuntos
Algoritmos , Hipocampo/diagnóstico por imagem , Processamento de Imagem Assistida por Computador/métodos , Microscopia de Fluorescência por Excitação Multifotônica/métodos , Animais , Hipocampo/citologia , Holografia , Camundongos , Microglia , Microscopia de Fluorescência por Excitação Multifotônica/instrumentação , Pontos Quânticos , Espalhamento de Radiação
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA
...