Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 72.054
Filtrar
1.
Zhong Nan Da Xue Xue Bao Yi Xue Ban ; 45(8): 980-987, 2020 Aug 28.
Artigo em Inglês, Chinês | MEDLINE | ID: mdl-33053542

RESUMO

OBJECTIVES: To evaluate the image quality of magnetic resonance 3D-sampling perfection with application-optimized contrasts by using different filp angle evolutions (3D-SPACE) sequence combined with different fusion methods in image fusion technology and the application value of fusion technology in preoperative evaluation of pituitary macroadenoma. METHODS: We collected MRI data of 43 patients with pituitary macroadenoma confirmed by Xiangya Hospital of Central South University, including collecting conventional MRI scan+enhancement, 3D-SPACE T2WI, and 3D-SPACE T1WI+C image data. 3D-SPACE sequence fusion was used in 6 combinations of 3D-SPACE T2WI/3D-SPACE T1WI+C, which were normal phase+normal phase, reverse phase+normal phase, normal phase+reverse phase, reverse phase+reverse phase, and normal phase pseudo color+normal phase, normal phase+normal phase pseudo color. Two senior radiologists used semi-quantitative methods to evaluate and compare the image quality of different combinations to obtain the best fusion mode. According to the degree of tumor invasion of the optic chiasma, oculomotor nerve, and cavernous sinus vessels, the MRI enhancement, 3D-SPACE T2WI, 3D-SPACE T1WI+C, and 2 3D-SPACE sequence fusion images were performed according to a three-level score system. Taking the intraoperative observation as the gold standard, Fisher probability exact method was used to compare different sequences to show the difference between the degree of invasion of the pituitary macroadenoma to the surrounding tissue and the intraoperative results. RESULTS: The Kruskal-Wallis H rank sum test showed that among the 6 image fusion modes, 3D-SPACE T1WI+C normal phase pseudo-color and 3D-SPACE T2WI normal phase fusion images had the best quality (P<0.05). No significance was observed among the 4 groups in the evaluation of consistency between the intraoperative findings and the graphically displayed extent of tumor invasion into oculomotor nerve (both level I, II, and III, P>0.05). The 3D-SPACE T2WI/3D-SPACE T1WI+C fusion images and the 3D-SPACE T2WI images showed better performance in the evaluation of consistency between the intraoperative findings and the graphically displayed extent of tumor invasion into optic chiasma (level II and III) than that in other two kinds of imaging data of the MRI enhancement group and 3D-SPACE T1WI+C group (P<0.01, P<0.05, respectively), with no significance being observed in level I. The 3D-SPACE T2WI/3D-SPACE T1WI+C fusion images and the 3D-SPACE T1WI+C images showed better performance in the evaluation of consistency between the intraoperative findings and the graphically displayed extent of tumor invasion into cavernous sinus vessels (level II) than that in other two kinds of imaging data of the MRI enhancement group and 3D-SPACE T2WI group (P<0.01, P<0.05, respectively), with no significance being observed in level I and III. CONCLUSIONS: Magnetic resonance 3D-SPACE sequence combined with image fusion technology is better than conventional magnetic resonance sequence for showing pituitary macroadenoma invasion of skull base vascular nerves. The invasion is better than the 3D-SPACE sequence alone, showing that the relationship between tumor and cavernous sinus vascular grade II better than the 3D-SPACE sequence alone. It has good application prospects for preoperative risk assessment and surgical plan.


Assuntos
Aumento da Imagem , Neoplasias Hipofisárias , Humanos , Imageamento Tridimensional , Imagem por Ressonância Magnética , Espectroscopia de Ressonância Magnética , Neoplasias Hipofisárias/diagnóstico por imagem , Neoplasias Hipofisárias/cirurgia
2.
Stomatologiia (Mosk) ; 99(5): 38-45, 2020.
Artigo em Russo | MEDLINE | ID: mdl-33034175

RESUMO

THE AIM: To evaluate versality and accuracy of computer navigation in orthognathic surgery, analyzing the position of osteotomized bone fragments on virtual and postoperative 3D models. MATERIAL AND METHODS: During our study we operated 27 patients with different asymmetric deformations of facial skeleton (13 patients with Class III Angle, 11 patients with Class II Angle and 3 patients with hemifacial microsomia). In 7 clinical cases optical navigation stations BrainLab 18070 Kick («BrainLab¼, Germany) and Stryker CranialMap CMF Version 2.0 («Stryker¼, USA) were used for preoperative virtual planning. In other clinical cases (20 patients) preoperative planning performed with using of 3D-cephafolometric programs Dolphin Imagin¼ and Blender 2.79. Intraoperative control of osteotomized bone fragments performed with using of optical navigation stations BrainLab 18070 Kick («BrainLab¼, Germany) and Stryker CranialMap CMF Version 2.0 («Stryker¼, USA). RESULTS: Mean surgical time was 181 minutes (150-210 min). Mean time of registration procedure was 5 minutes (3-8 min). Mean target registration error (TRE) was 0.9±0.18 mm. Absolute difference values between actual and virtual movements of maxilla was from 0.72 to 1.12 mm in vertical, from 0.56 to 0.94 mm in sagittal (COP) and from 0.39 to 0.58 mm in transversal (MSP) planes. CONCLUSION: Intraoperative control of maxilla-mandibular complex with using of computer navigation in orthognathic surgery allows to simplify bone fragments positioning, reduce surgery time, obtain a satisfactory aesthetic treatment result with occlusion restoration.


Assuntos
Cirurgia Ortognática , Procedimentos Cirúrgicos Ortognáticos , Cirurgia Assistida por Computador , Humanos , Imageamento Tridimensional , Mandíbula
3.
Zhonghua Kou Qiang Yi Xue Za Zhi ; 55(10): 765-771, 2020 Oct 09.
Artigo em Chinês | MEDLINE | ID: mdl-33045789

RESUMO

Objective: To analyze the anatomical morphological characteristics of the trabecular bone of human mandibular condyle by observing the distribution of volume of interests (VOI). Methods: The micro-CT images of a right condyle specimen of a 61-year-old adult male was analyzed in this study. The cylindrical VOI with both diameter and height of 2 mm were arranged, according to a certain pattern, as many as possible at various levels within the trabecular bone of the condyle. Each VOI had no intersection area. The selected VOI were divided into 5 parts: medial part, middle part, lateral part, anterior part and posterior part, with 6 layers from top to bottom. And the distribution of the overall anatomical morphological characteristics of three-dimensional (3D) images of the trabecular bone of the condyle was analyzed by using seven morphological parameters of each VOI, i.e. bone mineral density (BMD), bone volume/total volume (BV/TV), bone surface area/bone volume (BS/BV), trabecular thickness (Tb.Th), trabecular bone number (Tb.N), trabecular spacing (Tb.Sp) and trabecular bone pattern factor (Tb.Pf). Results: In the present study, totally 34 VOI were selected from the condyle specimen.The morphological distribution of the VOI was uneven in the 3D structure of the trabecular bone of the human condyle. BMD, BV/TV, Tb.N and Tb.Th were much higher at the middle part, while showed the smallest at the medial part. The anterior part demonstrated much higher parameters than the posterior part at the first, second, fifth and sixth layers, respectively, however, the posterior part showed much higher parameters than the anterior part at the third and fourth layers, respectively. The BMD was much higher at the first [(332.66±97.11) mg/cm3] and sixth [(255.79±45.68) mg/cm3] layers, while the lowest at the second layer [(255.79±41.06) mg/cm3]. The BV/TV and Tb.N were much higher at the first layer, with the lowest at the fifth layer. The Tb.Th at the first layer [(0.11±0.03) mm] was much higher than the others, which were similarly lower. The BS/BV, Tb.Sp and Tb.Pf were lower at the first layer and much higher at the medial and lateral parts, while were lower at the middle and anterior parts. The posterior part demonstrated much higher BS/BV, Tb.Sp and Tb.Pf than the anterior part at the first, fifth and sixth layers, respectively. However, the anterior part showed much higher scores than the posterior part at the third and fourth layers, respectively. The ANOVA results showed that the 7 morphological parameters of VOI were not statistically significant amongst the 6 layers (P>0.05). However, the 6 out of the 7 parameters, i.e. BV/TV, BS/BV, Tb.Th, Tb.N, Tb.Sp and Tb.Pf, were statistically significant amongst the five parts (P<0.05), while the only parameter of BMD was not statistically significant (P>0.05). Conclusions: The anatomical distribution characteristics of the trabecular bone of condyle were analyzed by using 3D imaging measurement based on the VOI. The results showed uneven distributions and indicated that the method of dividing the trabecular bone of mandibular condyle into VOI sets, which accorded with its specific anatomical characteristics, was feasible for further reference.


Assuntos
Densidade Óssea , Imageamento Tridimensional , Adulto , Humanos , Masculino , Côndilo Mandibular/diagnóstico por imagem , Pessoa de Meia-Idade , Microtomografia por Raio-X
4.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1848-1852, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018360

RESUMO

Gastric endoscopy is a standard clinical process that enables medical practitioners to diagnose various lesions inside a patient's stomach. If any lesion is found, it is very important to perceive the location of the lesion relative to the global view of the stomach. Our previous research showed that this could be addressed by reconstructing the whole stomach shape from chromoendoscopic images using a structure-from-motion (SfM) pipeline, in which indigo carmine (IC) blue dye-sprayed images were used to increase feature matches for SfM by enhancing stomach surface's textures. However, spraying the IC dye to the whole stomach requires additional time, labor, and cost, which is not desirable for patients and practitioners. In this paper, we propose an alternative way to achieve whole stomach 3D reconstruction without the need of the IC dye by generating virtual IC-sprayed (VIC) images based on image-to-image style translation trained on unpaired real no-IC and IC-sprayed images. We have specifically investigated the effect of input and output color channel selection for generating the VIC images and found that translating no-IC green-channel images to IC-sprayed red-channel images gives the best SfM reconstruction result.


Assuntos
Procedimentos Cirúrgicos do Sistema Digestório , Imageamento Tridimensional , Carmim , Humanos , Índigo Carmim , Estômago/diagnóstico por imagem
5.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2039-2042, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018405

RESUMO

Scoliosis is a 3D spinal deformation where the spine takes a lateral curvature, which generates an angle in a coronal plane. For periodic detection of scoliosis, safe and economic imaging modality is needed as continuous exposure to radiative imaging may cause cancer. 3D ultrasound imaging is a cost-effective and radiation-free imaging modality which gives volume projection image. Identification of mid-spine line using manual, semi-automatic and automatic methods have been published. Still, there are some difficulties like variations in human measurement, slow processing of data associated with them. In this paper, we propose an unsupervised ground truth generation and automatic spine curvature segmentation using U- Net. This approach of the application of Convolutional Neural Network on ultrasound spine image, to perform automatic detection of scoliosis, is a novel one.


Assuntos
Imageamento Tridimensional , Escoliose , Humanos , Redes Neurais de Computação , Escoliose/diagnóstico por imagem , Coluna Vertebral/diagnóstico por imagem , Ultrassonografia
6.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2043-2046, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018406

RESUMO

Segmentation of carotid vessel wall is required in vessel wall volume (VWV) and local vessel-wall-plus-plaque thickness (VWT) quantification of the carotid artery. Manual segmentation of the vessel wall is time-consuming and prone to interobserver variability. In this paper, we proposed a convolutional neural network (CNN) to segment the common carotid artery (CCA) from 3D carotid ultrasound images. The proposed CNN involves three U-Nets that segmented the 3D ultrasound (3DUS) images in the axial, lateral and frontal orientations. The segmentation maps generated by three U-Nets were consolidated by a novel segmentation average network (SAN) we proposed in this paper. The experimental results show that the proposed CNN improved the segmentation accuracies. Compared to only using U-Net alone, the proposed CNN improved the Dice similarity coefficient (DSC) for vessel wall segmentation from 64.8% to 67.5%, the sensitivity from 63.8% to 70.5%, and the area under receiver operator characteristic curve (AUC) from 0.89 to 0.94.


Assuntos
Artérias Carótidas , Imageamento Tridimensional , Artérias Carótidas/diagnóstico por imagem , Artéria Carótida Primitiva/diagnóstico por imagem , Ultrassonografia , Ultrassonografia Doppler
7.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2105-2108, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018421

RESUMO

The 3D ultrasound reconstruction technology has led to a rapid development of ultrasound spine imaging in recent decades. However, the current imaging apparatus is bulky and not portable. The objective of this study is to develop a new compact and wireless system to offer the real-time visualized spine images during data acquisition. A portable and WI-FI based ultrasound scanner and a compact EM tracking system were assembled to acquire ultrasound transverse frames with location information which could be reconstructed into 3D spine image volume in real-time. The validation was implemented on the 2D coronal images of vertebra phantoms, and the in vivo data acquisition and reconstruction were demonstrated on volunteers. The result showed that the new system could provide reconstructed spine images in real time and the average errors of the reconstructed images were about 1mm (approximate to image pixel size).


Assuntos
Algoritmos , Coluna Vertebral , Humanos , Imageamento Tridimensional , Projetos Piloto , Coluna Vertebral/diagnóstico por imagem , Ultrassonografia
8.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2408-2411, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018492

RESUMO

The scope of this paper is to present a new carotid vessel segmentation algorithm implementing the U-net based convolutional neural network architecture. With carotid atherosclerosis being the major cause of stroke in Europe, new methods that can provide more accurate image segmentation of the carotid arterial tree and plaque tissue can help improve early diagnosis, prevention and treatment of carotid disease. Herein, we present a novel methodology combining the U-net model and morphological active contours in an iterative framework that accurately segments the carotid lumen and outer wall. The method automatically produces a 3D meshed model of the carotid bifurcation and smaller branches, using multispectral MR image series obtained from two clinical centres of the TAXINOMISIS study. As indicated by a validation study, the algorithm succeeds high accuracy (99.1% for lumen area and 92.6% for the perimeter) for lumen segmentation. The proposed algorithm will be used in the TAXINOMISIS study to obtain more accurate 3D vessel models for improved computational fluid dynamics simulations and the development of models of atherosclerotic plaque progression.


Assuntos
Aprendizado Profundo , Imageamento Tridimensional , Artérias Carótidas/diagnóstico por imagem , Europa (Continente) , Imagem por Ressonância Magnética
9.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2634-2637, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018547

RESUMO

In this paper, we propose a novel video-based remote heart rate (HR) estimation method based on 3D facial landmarks. The key contributions in our method are twofold: (i) We introduce 3D facial landmarks detection to the video-based HR estimation and (ii) we propose a novel face patch visibility check manner based on the face patch normal in the 3D space. We experimentally demonstrate that, compared with baseline methods using 2D facial landmarks, our proposed method using 3D facial landmarks improves the robustness of HR estimation to head rotations and partial face occlusion. We also demonstrate that our visibility check is effective for selecting sufficiently visible face patches, contributing to the improvement of HR estimation accuracy.


Assuntos
Face , Imageamento Tridimensional , Frequência Cardíaca
10.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1576-1579, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018294

RESUMO

Quantitative Coronary Angiography (QCA) is an important tool in the study of coronary artery disease. Validation of this technique is crucial for their ongoing development and refinement although it is difficult due to several factors such as potential sources of error. The present work aims to a further validation of a new semi-automated method for three-dimensional (3D) reconstruction of coronary bifurcations arteries based on X-Ray Coronary Angiographies (CA). In a dataset of 40 patients (79 angiographic views), we used the aforementioned method to reconstruct them in 3D space. The validation was based on the comparison of these 3D models with the true silhouette of 2D models annotated by an expert using specific metrics. The obtained results indicate a good accuracy for the most parameters (≥ 90 %). Comparison with similar works shows that our new method is a promising tool for the 3D reconstruction of coronary bifurcations and for application in everyday clinical use.


Assuntos
Doença da Artéria Coronariana , Imageamento Tridimensional , Angiografia Coronária , Doença da Artéria Coronariana/diagnóstico por imagem , Coração , Humanos
11.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2096-2100, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018419

RESUMO

X-ray imaging is currently the gold standard for the assessment of spinal deformities. The purpose of this study is to evaluate a freehand 3D ultrasound system for volumetric reconstruction of the spine. A setup consisting of an ultrasound scanner with a linear transducer, an electromagnetic measuring system and a workstation was used. We conducted 64 acquisitions of US images of 8 adults in a natural standing position, and we tested three setups: 1) Subjects are constrained to be close to a wall, 2) Subjects are unconstrained, and 3) Subjects are constrained to performing fast and slow acquisitions. The spinous processes were manually selected from the volume reconstruction from tracked ultrasound images to generate a 3D point-based model depicting the centerline of the spine. The results suggested that a freehand 3D ultrasound system can be suitable for representing the spine. Volumetric reconstructions can be computed and landmarking can be performed to model the surface of the spine in the 3D space. These reconstructions promise to generate computer-based descriptors to analyze the shape of the spine in the 3D space.Clinical Relevance- We provide clinicians with a protocol that could be integrated in clinical setups for the assessment and monitoring of AIS, based on US image acquisitions, which constitutes a radiation-free technology.


Assuntos
Imageamento Tridimensional , Coluna Vertebral , Adulto , Fenômenos Eletromagnéticos , Humanos , Radiografia , Coluna Vertebral/diagnóstico por imagem , Ultrassonografia
12.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2109-2112, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018422

RESUMO

Quantification of ovarian and follicular volume and follicle count are performed in clinical practice for diagnosis and management in assisted reproduction. Ovarian volume and Antral Follicle Count (AFC) are typically tracked over the ovulation cycle. Volumetric analysis of ovary and follicle is manual and largely operator dependent. In this manuscript, we have proposed a deep-learning method for automatic simultaneous segmentation of ovary and follicles in 3D Transvaginal Ultrasound (TVUS), namely S-Net. The proposed loss function restricts false detection of follicles outside the ovary. Additionally, we have used multi-layered loss to provide deep supervision for training the network. S-Net is optimized for inference time and memory while utilizing 3D context in the 2D deep-learning network. 66 3D TVUS volumes (13,200 2D image slices) were acquired from 66 subjects in this Institutional Review Board (IRB) approved study. The segmentation framework provides approximately 92% and 87% average DICE overlap with the ground truth annotations for ovary and follicles, respectively. We have obtained state-of-the-art results with a detection rate of 88%, 91% and 98% for follicles of size 2-4mm, 4-12mm and >12mm.


Assuntos
Aprendizado Profundo , Ovário , Feminino , Humanos , Imageamento Tridimensional , Folículo Ovariano/diagnóstico por imagem , Ovário/diagnóstico por imagem , Ultrassonografia
13.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 4357-4360, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018960

RESUMO

We have developed a series of 3D endoscopic systems where a micro-sized pattern projector is inserted through the instrument channel of the endoscope and shapes are reconstructed by a structured light technique using captured images of the endoscopic camera. One problem of the previous works is that the accuracy of shape reconstruction is low, because the projector cannot be fixed to the endoscope, and thus, the pose of the pattern projector w.r.t. the camera cannot be pre-calibrated. In this paper, we propose a method to auto-calibrate the pose of the projector without using any special devices nor manual process. Since the technique is one-shot, multiple shapes can be reconstructed from an image sequence and a large 3D scene can be recovered by merging them. Experiments are conducted using the real system.


Assuntos
Algoritmos , Imageamento Tridimensional , Endoscópios , Endoscopia , Fotografação
14.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 4848-4853, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33019076

RESUMO

In this study, we present a human body shape statistical model including elderly people, which is constructed using principal component analysis (PCA) on 3D body scan data of approximately 130 people. As a pre-process step, a template human body mesh model is fitted to 3D scan data using a coarse-to-fine surface registration technique based on a conformal deformation method, in order to establish correspondences between the scans of different subjects possibly in different poses. To change body style by a small set of parameters, such as "age", "weight" and "height" or the easily measurable anthropometric parameters like "shoulder width", the linear transformations between these attributes and the first 10 principal component scores are obtained. We design a simple user interface to use this deformation model to generate different body styles easily. As a result, we were able to produce and show body styles capturing the characteristics of elderly people whose shoulders fell and back bent. Finally, as an application, we used our deformation method to generate different body types, performed forward dynamics simulations in an assistive device setting and visualized the differences in contact pressure distributions due to body shape changes.


Assuntos
Corpo Humano , Imageamento Tridimensional , Idoso , Antropometria , Humanos , Modelos Estatísticos , Análise de Componente Principal
15.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 803-807, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018107

RESUMO

Motion rehabilitation is increasingly required owing to an aging population and suffering of stroke, which means human motion analysis must be valued. Based on the concept mentioned above, a deep-learning-based system is proposed to track human motion based on three-dimensional (3D) images in this work; meanwhile, the features of traditional red green blue (RGB) images, known as two-dimensional (2D) images, were used as a comparison. The results indicate that 3D images have an advantage over 2D images due to the information of spatial relationships, which implies that the proposed system can be a potential technology for human motion analysis applications.


Assuntos
Algoritmos , Aprendizado Profundo , Idoso , Humanos , Imageamento Tridimensional , Movimento (Física)
16.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1088-1091, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018175

RESUMO

A unified framework for the analysis of fluorescence data taken by a two-photon imaging system is presented. As in the processing of blood-oxygen-level-dependent signals of functional magnetic resonance imaging, the acquired functional images have to be co-registered with a structural brain atlas before delineating the regions activated by a given stimulus. The voxels whose calcium traces are highly correlated with the predicted responses are demarcated without the need for subjective reasoning. Experimental data acquired while presenting olfactory stimuli are used to demonstrate the efficacy of the proposed schemes. The results indicate that the functional images of a Drosophila individual can be normalized into a standard stereotactic space, and the expected brain regions can be delineated adequately. This framework provides an opportunity to enable the development of a Drosophila functional connectome database.


Assuntos
Conectoma , Drosophila , Animais , Encéfalo/diagnóstico por imagem , Imageamento Tridimensional , Imagem por Ressonância Magnética
17.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1207-1210, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018204

RESUMO

Accurate segmentation of pulmonary vein (PV) and left atrium (LA) is essential for the preoperative evaluation and planning of total anomalous pulmonary venous connection (TAPVC), which is a rare but mortal congenital heart disease of children. However, manual segmentation is time-consuming and insipid. To free radiologists from the repetitive work, we propose an automatic deep learning method to segment PV and LA from Low-Dose CT images. In the method, attention mechanism is incorporated into the widely used V-Net and a novel grouped attention module is applied to enforce the segmentation performance of the V-Net. We evaluate our method on 68 3D Low-Dose CT images scanned from patients with TAPVC. The experiment result shows that our method outperforms the popular 3D-UNet and V-Net, with mean dice similarity coefficient (DSC) of 0.795 and 0.834 for the PV and LA respectively.Clinical relevance-We proposed a CNNs-based method for the automatic segmentation of PV and LA with good accuracy, which can be used for the preoperative evaluation and planning of TAPVC. Our method can improve the efficiency and reduce the workloads of radiologists (400 milliseconds vs. 2-3 hours per-case).


Assuntos
Veias Pulmonares , Síndrome de Cimitarra , Atenção , Criança , Átrios do Coração/diagnóstico por imagem , Humanos , Imageamento Tridimensional , Veias Pulmonares/diagnóstico por imagem
18.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1294-1298, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018225

RESUMO

Analyzing and interpreting cone-beam computed tomography (CBCT) images is a complicated and often time-consuming process. In this study, we present two different architectures of multi-channel deep learning (DL) models: "Ensemble" and "Synchronized multi-channel", to automatically identify and classify skeletal malocclusions from 3D CBCT craniofacial images. These multi-channel models combine three individual single-channel base models using a voting scheme and a two-step learning process, respectively, to simultaneously extract and learn a visual representation from three different directional views of 2D images generated from a single 3D CBCT image. We also employ a visualization method called "Class-selective Relevance Mapping" (CRM) to explain the learned behavior of our DL models by localizing and highlighting a discriminative area within an input image. Our multi-channel models achieve significantly better performance overall (accuracy exceeding 93%), compared to single-channel DL models that only take one specific directional view of 2D projected image as an input. In addition, CRM visually demonstrates that a DL model based on the sagittal-left view of 2D images outperforms those based on other directional 2D images.Clinical Relevance- the proposed method aims at assisting orthodontist to determine the best treatment path for the patient be it orthodontic or surgical treatment or a combination of both.


Assuntos
Aprendizado Profundo , Má Oclusão , Tomografia Computadorizada de Feixe Cônico , Humanos , Imageamento Tridimensional
19.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1299-1302, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018226

RESUMO

We proposed a target-based cone beam computed tomography (CBCT) imaging framework in order to optimize a free three dimensional (3D) source-detector trajectory by incorporating prior 3D image data. We aim to enable CBCT systems to provide topical information about a region of interest (ROI) using a short-scan trajectory with a reduced number of projections. The best projection views are selected by maximizing an objective function fed by the image quality by means of applying different x-ray positions on the digital phantom data. Finally, an optimized trajectory is selected which is applied to a C-arm device able to perform general source-detector positioning. An Alderson-Rando head phantom is used in order to investigate the performance of the proposed framework. Our experiments showed that the optimized trajectory could achieve a comparable image quality in the ROI with respect to the reference C-arm CBCT while using approximately one-quarter of projections. An angular range of 156° was used for the optimized trajectory.


Assuntos
Tomografia Computadorizada de Feixe Cônico Espiral , Tomografia Computadorizada de Feixe Cônico , Imageamento Tridimensional , Imagens de Fantasmas , Cintilografia
20.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1347-1350, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018238

RESUMO

2D/3D registration of preoperative computed tomography angiography with intra-operative X-ray angiography improves image guidance in percutaneous coronary intervention. However, previous registration methods are inaccurate and time-consuming due to simple deformation and iterative optimization, respectively. In this paper, we propose a novel method for non-rigid registration of coronary arteries based on a point set registration network, which predicts the complex deformation field directly without iterative optimization. In order to maintain the structure of coronary arteries, we advance the classical point set registration network with a loss function containing global and local topological constraints. The method was evaluated on ten clinical data, and it achieved a median chamfer distance of 73.60 pixels with a run time of less than 1s on CPU. Experimental results demonstrate that the proposed method is highly accurate and efficient.


Assuntos
Vasos Coronários , Imageamento Tridimensional , Algoritmos , Angiografia por Tomografia Computadorizada , Angiografia Coronária , Vasos Coronários/diagnóstico por imagem
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA