Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 21.437
Filtrar
1.
Nat Commun ; 15(1): 1347, 2024 Feb 14.
Artigo em Inglês | MEDLINE | ID: mdl-38355644

RESUMO

Accurate identification and localization of multiple abnormalities are crucial steps in the interpretation of chest X-rays (CXRs); however, the lack of a large CXR dataset with bounding boxes severely constrains accurate localization research based on deep learning. We created a large CXR dataset named CXR-AL14, containing 165,988 CXRs and 253,844 bounding boxes. On the basis of this dataset, a deep-learning-based framework was developed to identify and localize 14 common abnormalities and calculate the cardiothoracic ratio (CTR) simultaneously. The mean average precision values obtained by the model for 14 abnormalities reached 0.572-0.631 with an intersection-over-union threshold of 0.5, and the intraclass correlation coefficient of the CTR algorithm exceeded 0.95 on the held-out, multicentre and prospective test datasets. This framework shows an excellent performance, good generalization ability and strong clinical applicability, which is superior to senior radiologists and suitable for routine clinical settings.


Assuntos
Anormalidades Múltiplas , Aprendizado Profundo , Humanos , Estudos Prospectivos , Raios X , Cardiomegalia/diagnóstico por imagem
2.
Microsurgery ; 44(2): e31143, 2024 Feb.
Artigo em Inglês | MEDLINE | ID: mdl-38343010

RESUMO

INTRODUCTION: Institutional protocols often mandate the use of x-rays when a microneedle is lost intraoperatively. Although x-rays can reliably show a macroneedle, the benefit of x-rays in detecting microneedles in human tissues has not been established as available data on this topic are investigated in anthropometric models. The current study aims to evaluate whether x-rays can reliably detect retained microneedles in a human cadaveric model. We hypothesize that microneedles would be detected at a significantly lower rate than macroneedles by x-ray in human tissues. MATERIALS AND METHODS: Needles ranging from 4-0 to 10-0 were placed randomly throughout a cadaveric hand and foot. Each tissue sample was x-rayed using a Fexitron X-Ray machine, taking both anteroposterior and lateral views. A total of six x-ray images were then evaluated by 11 radiologists, independently. The radiologists circled over the area where they visualized a needle. The accuracy of detecting macroneedles (size 4-0 to 7-0) was compared with that of microneedles (size 8-0 to 10-0) using a chi-square test. RESULTS: The overall detection rate for the microneedles was significantly lower than the detection rate for macroneedles (13.5% vs 88.8%, p < .01). When subcategorized between the hand and the foot, the detection rate for microneedles was also significantly lower than the rate for macroneedles (hand: 7.6% for microneedles, 93.2% for macroneedles, p < .01; foot: 19.5% for microneedles, 84.4% for macroneedles, p < .01). The detection rate, in general, significantly decreased as the sizes of needles became smaller (7-0:70.5%, 8-0:18.2%, 9-0:16.7%, 10-0:2.3%, p < .01). CONCLUSION: X-rays, while useful in detecting macroneedles, had a significantly lower rate of detecting microneedles in a cadaveric model. The routine use of x-rays for a lost microneedle may not be beneficial. Further investigation with fresh tissue and similar intraoperative x-ray systems is warranted to corroborate and support these findings.


Assuntos
Sistemas de Liberação de Medicamentos , Agulhas , Humanos , Sistemas de Liberação de Medicamentos/métodos , Raios X , Cadáver
3.
Sci Rep ; 14(1): 2690, 2024 02 01.
Artigo em Inglês | MEDLINE | ID: mdl-38302556

RESUMO

Deep learning technology can effectively assist physicians in diagnosing chest radiographs. Conventional domain adaptation methods suffer from inaccurate lesion region localization, large errors in feature extraction, and a large number of model parameters. To address these problems, we propose a novel domain-adaptive method WDDM to achieve abnormal identification of chest radiographic images by combining Wasserstein distance and difference measures. Specifically, our method uses BiFormer as a multi-scale feature extractor to extract deep feature representations of data samples, which focuses more on discriminant features than convolutional neural networks and Swin Transformer. In addition, based on the loss minimization of Wasserstein distance and contrast domain differences, the source domain samples closest to the target domain are selected to achieve similarity and dissimilarity across domains. Experimental results show that compared with the non-transfer method that directly uses the network trained in the source domain to classify the target domain, our method has an average AUC increase of 14.8% and above. In short, our method achieves higher classification accuracy and better generalization performance.


Assuntos
Fontes de Energia Elétrica , Tórax , Raios X , Tórax/diagnóstico por imagem , Generalização Psicológica , Redes Neurais de Computação
4.
Eur Radiol Exp ; 8(1): 20, 2024 Feb 02.
Artigo em Inglês | MEDLINE | ID: mdl-38302850

RESUMO

BACKGROUND: Artificial intelligence (AI) seems promising in diagnosing pneumonia on chest x-rays (CXR), but deep learning (DL) algorithms have primarily been compared with radiologists, whose diagnosis can be not completely accurate. Therefore, we evaluated the accuracy of DL in diagnosing pneumonia on CXR using a more robust reference diagnosis. METHODS: We trained a DL convolutional neural network model to diagnose pneumonia and evaluated its accuracy in two prospective pneumonia cohorts including 430 patients, for whom the reference diagnosis was determined a posteriori by a multidisciplinary expert panel using multimodal data. The performance of the DL model was compared with that of senior radiologists and emergency physicians reviewing CXRs and that of radiologists reviewing computed tomography (CT) performed concomitantly. RESULTS: Radiologists and DL showed a similar accuracy on CXR for both cohorts (p ≥ 0.269): cohort 1, radiologist 1 75.5% (95% confidence interval 69.1-80.9), radiologist 2 71.0% (64.4-76.8), DL 71.0% (64.4-76.8); cohort 2, radiologist 70.9% (64.7-76.4), DL 72.6% (66.5-78.0). The accuracy of radiologists and DL was significantly higher (p ≤ 0.022) than that of emergency physicians (cohort 1 64.0% [57.1-70.3], cohort 2 63.0% [55.6-69.0]). Accuracy was significantly higher for CT (cohort 1 79.0% [72.8-84.1], cohort 2 89.6% [84.9-92.9]) than for CXR readers including radiologists, clinicians, and DL (all p-values < 0.001). CONCLUSIONS: When compared with a robust reference diagnosis, the performance of AI models to identify pneumonia on CXRs was inferior than previously reported but similar to that of radiologists and better than that of emergency physicians. RELEVANCE STATEMENT: The clinical relevance of AI models for pneumonia diagnosis may have been overestimated. AI models should be benchmarked against robust reference multimodal diagnosis to avoid overestimating its performance. TRIAL REGISTRATION: NCT02467192 , and NCT01574066 . KEY POINT: • We evaluated an openly-access convolutional neural network (CNN) model to diagnose pneumonia on CXRs. • CNN was validated against a strong multimodal reference diagnosis. • In our study, the CNN performance (area under the receiver operating characteristics curve 0.74) was lower than that previously reported when validated against radiologists' diagnosis (0.99 in a recent meta-analysis). • The CNN performance was significantly higher than emergency physicians' (p ≤ 0.022) and comparable to that of board-certified radiologists (p ≥ 0.269).


Assuntos
Aprendizado Profundo , Pneumonia , Humanos , Estudos Prospectivos , Inteligência Artificial , Raios X , Pneumonia/diagnóstico por imagem
5.
BMJ Open ; 14(2): e080034, 2024 02 05.
Artigo em Inglês | MEDLINE | ID: mdl-38316593

RESUMO

INTRODUCTION: Cystic fibrosis (CF) is a life-limiting autosomal recessive genetic condition. It is caused by mutations in the gene that encodes for a chloride and bicarbonate conducting transmembrane channel. X-ray velocimetry (XV) is a novel form of X-ray imaging that can generate lung ventilation data through the breathing cycle. XV technology has been validated in multiple animal models, including the ß-ENaC mouse model of CF lung disease. It has since been assessed in early-phase clinical trials in adult human subjects; however, there is a paucity of data in the paediatric cohort, including in CF. The aim of this pilot study was to investigate the feasibility of performing a single-centre cohort study in paediatric patients with CF and in those with normal lungs to demonstrate the appropriateness of proceeding with further studies of XV in these cohorts. METHODS AND ANALYSIS: This is a cross-sectional, single-centre, pilot study. It will recruit children aged 3-18 years to have XV lung imaging performed, as well as paired pulmonary function testing. The study will aim to recruit 20 children without CF with normal lungs and 20 children with CF. The primary outcome will be the feasibility of recruiting children and performing XV testing. Secondary outcomes will include comparisons between XV and current assessments of pulmonary function and structure. ETHICS AND DISSEMINATION: This project has ethical approval granted by The Women's and Children's Hospital Human Research Ethics Committee (HREC ID 2021/HRE00396). Findings will be disseminated through peer-reviewed publication and conferences. TRIAL REGISTRATION NUMBER: ACTRN12623000109606.


Assuntos
Fibrose Cística , Adulto , Animais , Camundongos , Criança , Humanos , Feminino , Fibrose Cística/diagnóstico por imagem , Fibrose Cística/complicações , Projetos Piloto , Raios X , Estudos de Coortes , Estudos Transversais , Pulmão/diagnóstico por imagem
6.
Int J Mol Sci ; 25(3)2024 Jan 26.
Artigo em Inglês | MEDLINE | ID: mdl-38338829

RESUMO

Molecular Dynamics simulations study material structure and dynamics at the atomic level. X-ray and neutron scattering experiments probe exactly the same time- and length scales as the simulations. In order to benchmark simulations against measured scattering data, a program is required that computes scattering patterns from simulations with good single-core performance and support for parallelization. In this work, the existing program Sassena is used as a potent solution to this requirement for a range of scattering methods, covering pico- to nanosecond dynamics, as well as the structure from some Ångströms to hundreds of nanometers. In the case of nanometer-level structures, the finite size of the simulation box, which is referred to as the finite size effect, has to be factored into the computations for which a method is described and implemented into Sassena. Additionally, the single-core and parallelization performance of Sassena is investigated, and several improvements are introduced.


Assuntos
Benchmarking , Simulação de Dinâmica Molecular , Raios X , Radiografia , Nêutrons , Difração de Nêutrons/métodos , Espalhamento a Baixo Ângulo , Difração de Raios X
8.
PeerJ ; 12: e16767, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-38313011

RESUMO

Paired petrography and acid maceration has shown that preferential silicification of shelly faunas can bias recovery based on taxon and body size. Here, silicified fossils from the Upper Ordovician Edinburg Formation, Strasburg Junction, Virginia, USA, were analyzed using X-ray tomographic microscopy (µCT) in conjunction with recovered residues from acid maceration of the same materials to further examine sources of potential bias. Results reveal that very small (<~1 mm) fossils are poorly resolved in µCT when scanning at lower resolutions (~30 µm), underestimating abundance of taxa including ostracods and bryozoans. Acid maceration, meanwhile, fails to recover poorly silicified fossils prone to disarticulation and/or fragmentation during digestion. Tests for patterns of breakage, however, indicate no significant size or taxonomic bias during extraction. Comparisons of individual fossils from 3-D fossil renders and maceration residues reveal patterns of fragmentation that are taxon-specific and allow the differentiation of biostratinomic and preparational breakage. Multivariate ordinations and cluster analyses of µCT and residue data in general produce concordant results but indicate that the variation in taxonomic composition of our samples is compromised by the resolvability of small size classes in µCT imaging, limiting the utility of this method for addressing paleoecological questions in these specific samples. We suggest that comparability of results will depend strongly on the sample size, taphonomic history, textural, and compositional characteristics of the samples in question, as well as µCT scan parameters. Additionally, applying these methods to different deposits will test the general applicability of the conclusions drawn on the relative strengths and weaknesses of the methods.


Assuntos
Fósseis , Microscopia , Raios X , Virginia
9.
Biomed Res ; 45(1): 25-31, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-38325843

RESUMO

The cell cycle dependence of radiosensitivity has yet to be fully determined, as it is technically difficult to achieve a high degree of cell cycle synchronization in cultured cell systems and accurately detect the cell cycle phase of individual cells simultaneously. We used human cervical carcinoma HeLa cells expressing fluorescent ubiquitination-based cell cycle indicators (FUCCI), and employed the mitotic harvesting method that is one of the cell cycle synchronization methods. The imaging analysis confirmed that the cell cycle is highly synchronized after mitotic cell harvesting until 18-20 h of the doubling time has elapsed. Also, flow cytometry analysis revealed that the S and G2 phases peak at approximately 12 and 14-16 h, respectively, after mitotic harvesting. In addition, the clonogenic assay showed the changes in surviving fractions following exposure to X-rays according to the progress through the cell cycle. These results indicate that HeLa-FUCCI cells become radioresistant in the G1 phase, become radiosensitive in the early S phase, rapidly become radioresistant in the late S phase, and become radiosensitive again in the G2 phase. Our findings may contribute to the further development of combinations of radiation and cell cycle-specific anticancer agents.


Assuntos
Células HeLa , Humanos , Raios X , Sobrevivência Celular , Microscopia de Fluorescência , Ciclo Celular , Ubiquitinação
10.
Commun Biol ; 7(1): 157, 2024 Feb 07.
Artigo em Inglês | MEDLINE | ID: mdl-38326549

RESUMO

The characterization of the vibrations of the middle ear ossicles during sound transmission is a focal point in clinical research. However, the small size of the structures, their micrometer-scale movement, and the deep-seated position of the middle ear within the temporal bone make these types of measurements extremely challenging. In this work, dynamic synchrotron-based X-ray phase-contrast microtomography is used on acoustically stimulated intact human ears, allowing for the three-dimensional visualization of entire human eardrums and ossicular chains in motion. A post-gating algorithm is used to temporally resolve the fast micromotions at 128 Hz, coupled with a high-throughput pipeline to process the large tomographic datasets. Seven ex-vivo fresh-frozen human temporal bones in healthy conditions are studied, and the rigid body motions of the ossicles are quantitatively delineated. Clinically relevant regions of the ossicular chain are tracked in 3D, and the amplitudes of their displacement are computed for two acoustic stimuli.


Assuntos
Imageamento Tridimensional , Síncrotrons , Humanos , Raios X , Orelha Média/diagnóstico por imagem , Ossículos da Orelha/diagnóstico por imagem
11.
Sci Rep ; 14(1): 3341, 2024 02 09.
Artigo em Inglês | MEDLINE | ID: mdl-38336974

RESUMO

Accurate annotation of vertebral bodies is crucial for automating the analysis of spinal X-ray images. However, manual annotation of these structures is a laborious and costly process due to their complex nature, including small sizes and varying shapes. To address this challenge and expedite the annotation process, we propose an ensemble pipeline called VertXNet. This pipeline currently combines two segmentation mechanisms, semantic segmentation using U-Net, and instance segmentation using Mask R-CNN, to automatically segment and label vertebral bodies in lateral cervical and lumbar spinal X-ray images. VertXNet enhances its effectiveness by adopting a rule-based strategy (termed the ensemble rule) for effectively combining segmentation outcomes from U-Net and Mask R-CNN. It determines vertebral body labels by recognizing specific reference vertebral instances, such as cervical vertebra 2 ('C2') in cervical spine X-rays and sacral vertebra 1 ('S1') in lumbar spine X-rays. Those references are commonly relatively easy to identify at the edge of the spine. To assess the performance of our proposed pipeline, we conducted evaluations on three spinal X-ray datasets, including two in-house datasets and one publicly available dataset. The ground truth annotations were provided by radiologists for comparison. Our experimental results have shown that the proposed pipeline outperformed two state-of-the-art (SOTA) segmentation models on our test dataset with a mean Dice of 0.90, vs. a mean Dice of 0.73 for Mask R-CNN and 0.72 for U-Net. We also demonstrated that VertXNet is a modular pipeline that enables using other SOTA model, like nnU-Net to further improve its performance. Furthermore, to evaluate the generalization ability of VertXNet on spinal X-rays, we directly tested the pre-trained pipeline on two additional datasets. A consistently strong performance was observed, with mean Dice coefficients of 0.89 and 0.88, respectively. In summary, VertXNet demonstrated significantly improved performance in vertebral body segmentation and labeling for spinal X-ray imaging. Its robustness and generalization were presented through the evaluation of both in-house clinical trial data and publicly available datasets.


Assuntos
Tomografia Computadorizada por Raios X , Corpo Vertebral , Tomografia Computadorizada por Raios X/métodos , Raios X , Radiografia , Vértebras Cervicais/diagnóstico por imagem , Processamento de Imagem Assistida por Computador/métodos
12.
Biomed Instrum Technol ; 58(1): 18-24, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-38324284

RESUMO

Trends toward the use of irradiator parameter release (also called machine-based release) put pressure on equipment manufacturers to guarantee accuracy and reliability of monitored process parameters. In the specific case of X-ray processing, relevance of these monitored parameters is questionable due to the additional difficulty coming from the fact that the X-ray converter does not have associated parameters or a monitored feedback mechanism. To bridge this gap, this article presents a novel method to verify in real-time consistency of certain X-ray field properties. It covers the description of an X-ray flux monitor and its experimental characterization. The proposed detector can be used as a control and monitoring tool in addition to the conventional "passive" dosimetry per ISO 11137-1 and ISO 11137-3. It can detect photon flux deviation on the order of magnitude of 1%. Its performance would allow real-time monitoring of each pallet being processed and ensure that the correct X-ray beam is directed to the product. Further, the known response of the detector to a product can serve as a validation that the correct product is in front of the beam. Moreover, a detector of this type could contribute to moving from the current dosimetric release to irradiator parameter release. Compared with current practices, benefits would include an increased number of control points used to verify process conformity, real-time information on the radiation field (process output validation), limited manual handling of dosimeters, and verification that the product treated is the same as the performance qualification dose-mapped product.


Assuntos
Esterilização , Raios X , Reprodutibilidade dos Testes
13.
Biomed Instrum Technol ; 58(1): 7-17, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-38324282

RESUMO

Synthetic organic polymers commonly are used in the construction of healthcare product and medical device components. Medical devices often are sterilized to ensure that they are free from viable microorganisms. A common technique to achieve this is using ionizing radiation, usually gamma. A trend exists in industrial sterilization to supplement gamma with alternative accelerator technologies (e.g., X-ray). In the current work, studies were performed to characterize polymer modifications caused by gamma and X-ray sterilization processes and to assess the comparative equivalency. The studies were developed to evaluate two key process parameters: dose and dose rate. Three commonly used polymers were selected: high-density polyethylene, low-density polyethylene, and polypropylene. Four grades of each family were chosen. The dose assessment involved sample exposures to both gamma and X-ray irradiation at two dose levels (30 and 55 kGy). All other processing conditions, including dose rate, were controlled at standard processing levels akin to each sterilization technology. The dose rate assessment expanded on each dose level by introducing two additional dose rate parameters. Subsequent laboratory testing used techniques to characterize physico-chemical properties of the polymers to ascertain equivalency across test groups. Initial results indicated positive levels of equivalency between gamma and X-ray irradiation.


Assuntos
Instalações de Saúde , Indústrias , Raios X , Polietileno , Polímeros , Atenção à Saúde
14.
Phys Med Biol ; 69(4)2024 Feb 13.
Artigo em Inglês | MEDLINE | ID: mdl-38347732

RESUMO

Objective. Chest x-ray image representation and learning is an important problem in computer-aided diagnostic area. Existing methods usually adopt CNN or Transformers for feature representation learning and focus on learning effective representations for chest x-ray images. Although good performance can be obtained, however, these works are still limited mainly due to the ignorance of mining the correlations of channels and pay little attention on the local context-aware feature representation of chest x-ray image.Approach. To address these problems, in this paper, we propose a novel spatial-channel high-order attention model (SCHA) for chest x-ray image representation and diagnosis. The proposed network architecture mainly contains three modules, i.e. CEBN, SHAM and CHAM. To be specific, firstly, we introduce a context-enhanced backbone network by employing multi-head self-attention to extract initial features for the input chest x-ray images. Then, we develop a novel SCHA which contains both spatial and channel high-order attention learning branches. For the spatial branch, we develop a novel local biased self-attention mechanism which can capture both local and long-range global dependences of positions to learn rich context-aware representation. For the channel branch, we employ Brownian Distance Covariance to encode the correlation information of channels and regard it as the image representation. Finally, the two learning branches are integrated together for the final multi-label diagnosis classification and prediction.Main results. Experiments on the commonly used datasets including ChestX-ray14 and CheXpert demonstrate that our proposed SCHA approach can obtain better performance when comparing many related approaches.Significance. This study obtains a more discriminative method for chest x-ray classification and provides a technique for computer-aided diagnosis.


Assuntos
Diagnóstico por Computador , Tórax , Raios X , Radiografia
16.
Commun Biol ; 7(1): 147, 2024 Feb 02.
Artigo em Inglês | MEDLINE | ID: mdl-38307988

RESUMO

Structural insights into the photoactivated adenylate cyclases can be used to develop new ways of controlling cellular cyclic adenosine monophosphate (cAMP) levels for optogenetic and other applications. In this work, we use an integrative approach that combines biophysical and structural biology methods to provide insight on the interaction of adenosine triphosphate (ATP) with the dark-adapted state of the photoactivated adenylate cyclase from the cyanobacterium Oscillatoria acuminata (OaPAC). A moderate affinity of the nucleotide for the enzyme was calculated and the thermodynamic parameters of the interaction have been obtained. Stopped-flow fluorescence spectroscopy and small-angle solution scattering have revealed significant conformational changes in the enzyme, presumably in the adenylate cyclase (AC) domain during the allosteric mechanism of ATP binding to OaPAC with small and large-scale movements observed to the best of our knowledge for the first time in the enzyme in solution upon ATP binding. These results are in line with previously reported drastic conformational changes taking place in several class III AC domains upon nucleotide binding.


Assuntos
Trifosfato de Adenosina , Adenilil Ciclases , Adenilil Ciclases/genética , Adenilil Ciclases/química , Adenilil Ciclases/metabolismo , Trifosfato de Adenosina/metabolismo , Espectrometria de Fluorescência , Raios X , Conformação Molecular
17.
BMC Med Imaging ; 24(1): 6, 2024 01 02.
Artigo em Inglês | MEDLINE | ID: mdl-38166579

RESUMO

In this paper, we propose an attention-enhanced architecture for improved pneumonia detection in chest X-ray images. A unique attention mechanism is integrated with ResNet to highlight salient features crucial for pneumonia detection. Rigorous evaluation demonstrates that our attention mechanism significantly enhances pneumonia detection accuracy, achieving a satisfactory result of 96% accuracy. To address the issue of imbalanced training samples, we integrate an enhanced focal loss into our architecture. This approach assigns higher weights to minority classes during training, effectively mitigating data imbalance. Our model's performance significantly improves, surpassing that of traditional approaches such as the pretrained ResNet-50 model. Our attention-enhanced architecture thus presents a powerful solution for pneumonia detection in chest X-ray images, achieving an accuracy of 98%. By integrating enhanced focal loss, our approach effectively addresses imbalanced training sample. Comparative analysis underscores the positive impact of our model's spatial and channel attention modules. Overall, our study advances pneumonia detection in medical imaging and underscores the potential of attention-enhanced architectures for improved diagnostic accuracy and patient outcomes. Our findings offer valuable insights into image diagnosis and pneumonia prevention, contributing to future research in medical imaging and machine learning.


Assuntos
Pneumonia , Tórax , Humanos , Raios X , Aprendizado de Máquina , Pneumonia/diagnóstico por imagem
18.
BMC Med Imaging ; 24(1): 1, 2024 01 02.
Artigo em Inglês | MEDLINE | ID: mdl-38166813

RESUMO

Deep learning is a highly significant technology in clinical treatment and diagnostics nowadays. Convolutional Neural Network (CNN) is a new idea in deep learning that is being used in the area of computer vision. The COVID-19 detection is the subject of our medical study. Researchers attempted to increase the detection accuracy but at the cost of high model complexity. In this paper, we desire to achieve better accuracy with little training space and time so that this model easily deployed in edge devices. In this paper, a new CNN design is proposed that has three stages: pre-processing, which removes the black padding on the side initially; convolution, which employs filter banks; and feature extraction, which makes use of deep convolutional layers with skip connections. In order to train the model, chest X-ray images are partitioned into three sets: learning(0.7), validation(0.1), and testing(0.2). The models are then evaluated using the test and training data. The LMNet, CoroNet, CVDNet, and Deep GRU-CNN models are the other four models used in the same experiment. The propose model achieved 99.47% & 98.91% accuracy on training and testing respectively. Additionally, it achieved 97.54%, 98.19%, 99.49%, and 97.86% scores for precision, recall, specificity, and f1-score respectively. The proposed model obtained nearly equivalent accuracy and other similar metrics when compared with other models but greatly reduced the model complexity. Moreover, it is found that proposed model is less prone to over fitting as compared to other models.


Assuntos
COVID-19 , Humanos , COVID-19/diagnóstico por imagem , Raios X , Tórax , Redes Neurais de Computação
19.
Sci Data ; 11(1): 20, 2024 Jan 03.
Artigo em Inglês | MEDLINE | ID: mdl-38172163

RESUMO

X-ray coronary angiography is the most common tool for the diagnosis and treatment of coronary artery disease. It involves the injection of contrast agents into coronary vessels using a catheter to highlight the coronary vessel structure. Typically, multiple 2D X-ray projections are recorded from different angles to improve visualization. Recent advances in the development of deep-learning-based tools promise significant improvement in diagnosing and treating coronary artery disease. However, the limited public availability of annotated X-ray coronary angiography image datasets presents a challenge for objective assessment and comparison of existing tools and the development of novel methods. To address this challenge, we introduce a novel ARCADE dataset with 2 objectives: coronary vessel classification and stenosis detection. Each objective contains 1500 expert-labeled X-ray coronary angiography images representing: i) coronary artery segments; and ii) the locations of stenotic plaques. These datasets will serve as a benchmark for developing new methods and assessing existing approaches for the automated diagnosis and risk assessment of coronary artery disease.


Assuntos
Doença da Artéria Coronariana , Humanos , Cateteres , Meios de Contraste , Angiografia Coronária/métodos , Doença da Artéria Coronariana/diagnóstico por imagem , Raios X
20.
Sci Rep ; 14(1): 822, 2024 01 08.
Artigo em Inglês | MEDLINE | ID: mdl-38191885

RESUMO

A first irradiation platform capable of delivering 10 MV X-ray beams at ultra-high dose rates (UHDR) has been developed and characterized for FLASH radiobiological research at TRIUMF. Delivery of both UHDR (FLASH mode) and low dose-rate conventional (CONV mode) irradiations was demonstrated using a common source and experimental setup. Dose rates were calculated using film dosimetry and a non-intercepting beam monitoring device; mean values for a 100 µA pulse (peak) current were nominally 82.6 and 4.40 × 10-2 Gy/s for UHDR and CONV modes, respectively. The field size for which > 40 Gy/s could be achieved exceeded 1 cm down to a depth of 4.1 cm, suitable for total lung irradiations in mouse models. The calculated delivery metrics were used to inform subsequent pre-clinical treatments. Four groups of 6 healthy male C57Bl/6J mice were treated using thoracic irradiations to target doses of either 15 or 30 Gy using both FLASH and CONV modes. Administration of UHDR X-ray irradiation to healthy mouse models was demonstrated for the first time at the clinically-relevant beam energy of 10 MV.


Assuntos
Benchmarking , Radiometria , Masculino , Animais , Camundongos , Raios X , Radiografia , Modelos Animais de Doenças , Camundongos Endogâmicos C57BL
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA
...