RESUMO
BACKGROUND: Stroke infarct volume predicts patient disability and has utility for clinical trial outcomes. Accurate infarct volume measurement requires manual segmentation of stroke boundaries in diffusion-weighted magnetic resonance imaging scans which is time-consuming and subject to variability. Automatic infarct segmentation should be robust to rotation and reflection; however, prior work has not encoded this property into deep learning architecture. Here, we use rotation-reflection equivariance and train a deep learning model to segment stroke volumes in a large cohort of well-characterized patients with acute ischemic stroke in different vascular territories. METHODS: In this retrospective study, patients were selected from a stroke registry at Houston Methodist Hospital. Eight hundred seventy-five patients with acute ischemic stroke in any brain area who had magnetic resonance imaging with diffusion-weighted imaging were included for analysis and split 80/20 for training/testing. Infarct volumes were manually segmented by consensus of 3 independent clinical experts and cross-referenced against radiology reports. A rotation-reflection equivariant model was developed based on U-Net and grouped convolutions. Segmentation performance was evaluated using Dice score, precision, and recall. Ninety-day modified Rankin Scale outcome prediction was also evaluated using clinical variables and segmented stroke volumes in different brain regions. RESULTS: Segmentation model Dice scores are 0.88 (95% CI, 0.87-0.89; training) and 0.85 (0.82-0.88; testing). The modified Rankin Scale outcome prediction AUC using stroke volume in 30 refined brain regions based upon modified Rankin Scale-relevance areas adjusted for clinical variables was 0.80 (0.76-0.83) with an accuracy of 0.75 (0.72-0.78). CONCLUSIONS: We trained a deep learning model with encoded rotation-reflection equivariance to segment acute ischemic stroke lesions in diffusion- weighted imaging using a large data set from the Houston Methodist stroke center. The model achieved competitive performance in 175 well-balanced hold-out testing cases that include strokes from different vascular territories. Furthermore, the location specific stroke volume segmentations from the deep learning model combined with clinical factors demonstrated high AUC and accuracy for 90-day modified Rankin Scale in an outcome prediction model.
Assuntos
Isquemia Encefálica , AVC Isquêmico , Acidente Vascular Cerebral , Isquemia Encefálica/diagnóstico por imagem , Imagem de Difusão por Ressonância Magnética/métodos , Humanos , Infarto , AVC Isquêmico/diagnóstico por imagem , Prognóstico , Estudos Retrospectivos , Acidente Vascular Cerebral/diagnóstico por imagem , Acidente Vascular Cerebral/patologia , Volume SistólicoRESUMO
The fields of medicine science and health informatics have made great progress recently and have led to in-depth analytics that is demanded by generation, collection and accumulation of massive data. Meanwhile, we are entering a new period where novel technologies are starting to analyze and explore knowledge from tremendous amount of data, bringing limitless potential for information growth. One fact that cannot be ignored is that the techniques of machine learning and deep learning applications play a more significant role in the success of bioinformatics exploration from biological data point of view, and a linkage is emphasized and established to bridge these two data analytics techniques and bioinformatics in both industry and academia. This survey concentrates on the review of recent researches using data mining and deep learning approaches for analyzing the specific domain knowledge of bioinformatics. The authors give a brief but pithy summarization of numerous data mining algorithms used for preprocessing, classification and clustering as well as various optimized neural network architectures in deep learning methods, and their advantages and disadvantages in the practical applications are also discussed and compared in terms of their industrial usage. It is believed that in this review paper, valuable insights are provided for those who are dedicated to start using data analytics methods in bioinformatics.
Assuntos
Biologia Computacional , Mineração de Dados , Algoritmos , Teorema de Bayes , Aprendizado de Máquina , Inquéritos e QuestionáriosRESUMO
Cardiac dysfunction constitutes common cardiovascular health issues in the society, and has been an investigation topic of strong focus by researchers in the medical imaging community. Diagnostic modalities based on echocardiography, magnetic resonance imaging, chest radiography and computed tomography are common techniques that provide cardiovascular structural information to diagnose heart defects. However, functional information of cardiovascular flow, which can in fact be used to support the diagnosis of many cardiovascular diseases with a myriad of hemodynamics performance indicators, remains unexplored to its full potential. Some of these indicators constitute important cardiac functional parameters affecting the cardiovascular abnormalities. With the advancement of computer technology that facilitates high speed computational fluid dynamics, the realization of a support diagnostic platform of hemodynamics quantification and analysis can be achieved. This article reviews the state-of-the-art medical imaging and high fidelity multi-physics computational analyses that together enable reconstruction of cardiovascular structures and hemodynamic flow patterns within them, such as of the left ventricle (LV) and carotid bifurcations. The combined medical imaging and hemodynamic analysis enables us to study the mechanisms of cardiovascular disease-causing dysfunctions, such as how (1) cardiomyopathy causes left ventricular remodeling and loss of contractility leading to heart failure, and (2) modeling of LV construction and simulation of intra-LV hemodynamics can enable us to determine the optimum procedure of surgical ventriculation to restore its contractility and health This combined medical imaging and hemodynamics framework can potentially extend medical knowledge of cardiovascular defects and associated hemodynamic behavior and their surgical restoration, by means of an integrated medical image diagnostics and hemodynamic performance analysis framework.
Assuntos
Sistema Cardiovascular/anatomia & histologia , Sistema Cardiovascular/diagnóstico por imagem , Diagnóstico por Imagem/métodos , Hemodinâmica , Doenças Cardiovasculares/diagnóstico por imagem , Doenças Cardiovasculares/patologia , Doenças Cardiovasculares/fisiopatologia , Sistema Cardiovascular/patologia , Humanos , Processamento de Imagem Assistida por Computador , Modelos CardiovascularesRESUMO
In this paper, a novel training/testing process for building/using a classification model based on human activity recognition (HAR) is proposed. Traditionally, HAR has been accomplished by a classifier that learns the activities of a person by training with skeletal data obtained from a motion sensor, such as Microsoft Kinect. These skeletal data are the spatial coordinates (x, y, z) of different parts of the human body. The numeric information forms time series, temporal records of movement sequences that can be used for training a classifier. In addition to the spatial features that describe current positions in the skeletal data, new features called 'shadow features' are used to improve the supervised learning efficacy of the classifier. Shadow features are inferred from the dynamics of body movements, and thereby modelling the underlying momentum of the performed activities. They provide extra dimensions of information for characterising activities in the classification process, and thereby significantly improve the classification accuracy. Two cases of HAR are tested using a classification model trained with shadow features: one is by using wearable sensor and the other is by a Kinect-based remote sensor. Our experiments can demonstrate the advantages of the new method, which will have an impact on human activity detection research.
Assuntos
Atividades Humanas , Humanos , Movimento (Física) , MovimentoRESUMO
Simulation of blood flow in a stenosed artery using Smoothed Particle Hydrodynamics (SPH) is a new research field, which is a particle-based method and different from the traditional continuum modelling technique such as Computational Fluid Dynamics (CFD). Both techniques harness parallel computing to process hemodynamics of cardiovascular structures. The objective of this study is to develop and test a new robust method for comparison of arterial flow velocity contours by SPH with the well-established CFD technique, and the implementation of SPH in computed tomography (CT) reconstructed arteries. The new method was developed based on three-dimensional (3D) straight and curved arterial models of millimeter range with a 25% stenosis in the middle section. In this study, we employed 1,000 to 13,000 particles to study how the number of particles influences SPH versus CFD deviation for blood-flow velocity distribution. Because further increasing the particle density has a diminishing effect on this deviation, we have determined a critical particle density of 1.45 particles/mm2 based on Reynolds number (Reâ=â200) at the inlet for an arterial flow simulation. Using this critical value of particle density can avoid unnecessarily big computational expenses that have no further effect on simulation accuracy. We have particularly shown that the SPH method has a big potential to be used in the virtual surgery system, such as to simulate the interaction between blood flow and the CT reconstructed vessels, especially those with stenosis or plaque when encountering vasculopathy, and for employing the simulation results output in clinical surgical procedures.
Assuntos
Velocidade do Fluxo Sanguíneo/fisiologia , Simulação por Computador , Processamento de Imagem Assistida por Computador/métodos , Modelos Cardiovasculares , Tomografia Computadorizada por Raios X/métodos , Arteriopatias Oclusivas/diagnóstico por imagem , Arteriopatias Oclusivas/patologia , Arteriopatias Oclusivas/fisiopatologia , Humanos , HidrodinâmicaRESUMO
This paper presents a framework for modelling biological tissues based on discrete particles. Cell components (e.g. cell membranes, cell cytoskeleton, cell nucleus) and extracellular matrix (e.g. collagen) are represented using collections of particles. Simple particle to particle interaction laws are used to simulate and control complex physical interaction types (e.g. cell-cell adhesion via cadherins, integrin basement membrane attachment, cytoskeletal mechanical properties). Particles may be given the capacity to change their properties and behaviours in response to changes in the cellular microenvironment (e.g., in response to cell-cell signalling or mechanical loadings). Each particle is in effect an 'agent', meaning that the agent can sense local environmental information and respond according to pre-determined or stochastic events. The behaviour of the proposed framework is exemplified through several biological problems of ongoing interest. These examples illustrate how the modelling framework allows enormous flexibility for representing the mechanical behaviour of different tissues, and we argue this is a more intuitive approach than perhaps offered by traditional continuum methods. Because of this flexibility, we believe the discrete modelling framework provides an avenue for biologists and bioengineers to explore the behaviour of tissue systems in a computational laboratory.
Assuntos
Fenômenos Fisiológicos Celulares , Matriz Extracelular/fisiologia , Mecanotransdução Celular/fisiologia , Modelos Biológicos , Frações Subcelulares/fisiologia , Animais , Simulação por Computador , Humanos , Modelos EstatísticosRESUMO
BACKGROUND: Double injection of blood into cisterna magna using a rabbit model results in cerebral vasospasm. An unacceptably high mortality rate tends to limit the application of model. Ultrasound guided puncture can provide real-time imaging guidance for operation. The aim of this paper is to establish a safe and effective rabbit model of cerebral vasospasm after subarachnoid hemorrhage with the assistance of ultrasound medical imaging. METHODS: A total of 160 New Zealand white rabbits were randomly divided into four groups of 40 each: (1) manual control group, (2) manual model group, (3) ultrasound guided control group, and (4) ultrasound guided model group. The subarachnoid hemorrhage was intentionally caused by double injection of blood into their cisterna magna. Then, basilar artery diameters were measured using magnetic resonance angiography before modeling and 5 days after modeling. RESULTS: The depth of needle entering into cisterna magna was determined during the process of ultrasound guided puncture. The mortality rates in manual control group and model group were 15 and 23 %, respectively. No rabbits were sacrificed in those two ultrasound guided groups. We found that the mortality rate in ultrasound guided groups decreased significantly compared to manual groups. Compared with diameters before modeling, the basilar artery diameters after modeling were significantly lower in manual and ultrasound guided model groups. The vasospasm aggravated and the proportion of severe vasospasms was greater in ultrasound guided model group than that of manual group. In manual model group, no vasospasm was found in 8 % of rabbits. CONCLUSIONS: The ultrasound guided double injection of blood into cisterna magna is a safe and effective rabbit model for treatment of cerebral vasospasm.
Assuntos
Sangue , Cisterna Magna , Cirurgia Assistida por Computador , Vasoespasmo Intracraniano/diagnóstico por imagem , Vasoespasmo Intracraniano/cirurgia , Animais , Modelos Animais de Doenças , Injeções , Imageamento por Ressonância Magnética , Masculino , Punções , Coelhos , Segurança , Hemorragia Subaracnóidea/cirurgia , Tomografia Computadorizada por Raios X , UltrassonografiaRESUMO
BACKGROUND: The objective of this study is to assess standardized histograms of signal intensities of T1 signal and T2 signal on sagittal view without enhancement during (1) acute stage, and (2) convalescence stage of pediatric patients with Enterovirus 71 related brainstem encephalitis (BE), and with respect to (3) healthy normal. METHODS: Our subjects were hospitalized between March 2010 and October 2012, and underwent pre- and post-contrast MRI studies. The research question to be answered is whether the comparison of the MRI image intensity histograms and relevant statistical quantification can add new knowledge to the diagnosis of BE patients. So, both 25 cases in acute stage with prolonged T1 and T2 signal, without enhancement, and 13 cases in convalescence stage were introduced. In additional, a healthy group with 25 cases was recruited for comparison. RESULTS: MRI signal intensity histogram changes of the lesions were compared at the acute and convalescence stages of the disease. Our preliminary results suggest that standardized histograms of signal intensities and their statistical properties are able to provide diagnostic information for the clinical assessment of the disease. Different stages pertaining to the histogram plots comparison showed that overall T1 signal intensity values increase as we traverse from the acute stage to the convalescence stage. And then for the healthy subjects, the T2 signal intensity values changed their magnitudes in a reverse direction. However, exceptions of this can happen in four cases where the primary lesions occurred in the brainstem that developed encephalomalacia resulting in a lower signal in T1WI and higher signal in T2WI. Statistical analysis revealed there was significant difference of T1 signal intensity among the three groups; and also, the T2 signal intensity was lower than other two groups. CONCLUSIONS: Standardized histogram of T1 and T2 intensity provide valuable and useful information for disease diagnosis and evaluation, which can potentially help medical doctors to save the lives of children.
Assuntos
Tronco Encefálico/virologia , Convalescença , Encefalite Viral/diagnóstico , Enterovirus Humano A/fisiologia , Infecções por Enterovirus/diagnóstico , Processamento de Imagem Assistida por Computador , Imageamento por Ressonância Magnética , Doença Aguda , Estudos de Casos e Controles , Criança , Encefalite Viral/complicações , Encefalite Viral/terapia , Infecções por Enterovirus/complicações , Infecções por Enterovirus/terapia , Doença de Mão, Pé e Boca/complicações , Hospitalização , Humanos , Estudos RetrospectivosRESUMO
Automated identification of cardiac vortices is a formidable task due to the complex nature of blood flow within the heart chambers. This study proposes a novel approach that algorithmically characterizes the identification criteria of these cardiac vortices based on Lagrangian Averaged Vorticity Deviation (LAVD). For this purpose, the Recurrent All-Pairs Field Transforms (RAFT) is employed to assess the optical flow over the Phase Contrast Magnetic Resonance Imaging (PC-MRI), and to construct a continuous blood flow velocity field and reduce errors that arise from the integral process of LAVD. Additionally, Generalized Hough Transform (GHT) is applied for automated depiction of the structure of cardiac vortices. The effectiveness of this method is demonstrated and validated by the computation of the acquired cardiac flow data. The results of this comprehensive visual and analytical study show that the evolution of cardiac vortices can be effectively described and displayed, and the RAFT framework for optical flow can synthesize the in-between PC-MRIs with high accuracy. This allows cardiologists to acquire a deeper understanding of intracardiac hemodynamics and its impact on cardiac functional performance.
Assuntos
Algoritmos , Humanos , Velocidade do Fluxo Sanguíneo/fisiologia , Imageamento por Ressonância Magnética/métodos , Modelos Cardiovasculares , Coração/fisiologia , Coração/diagnóstico por imagem , Processamento de Imagem Assistida por Computador/métodosRESUMO
Calcifying marine invertebrates with complex life cycles are particularly at risk to climate changes as they undergo an abrupt ontogenetic shift during larval metamorphosis. Although our understanding of the larval response to climate changes is rapidly advancing, the proteome plasticity involved in a compensatory response to climate change is still unknown. In this study, we investigated the proteomic response of metamorphosing larvae of the tubeworm Hydroides elegans, challenged with two climate change stressors, ocean acidification (OA; pH 7.6) and hypoxia (HYP; 2.8 mg O2 l(-1)), and with both combined. Using a two-dimensional gel electrophoresis (2-DE)-based approach coupled with mass spectrometry, we found that climate change stressors did not affect metamorphosis except under OA, but altered the larval proteome and phosphorylation status. Metabolism and various stress and calcification-related proteins were downregulated in response to OA. In OA and HYP combined, HYP restored the expression of the calcification-related proteins to the control levels. We speculate that mild HYP stress could compensate for the negative effects of OA. This study also discusses the potential functions of selected proteins that might play important roles in larval acclimation and adaption to climate change.
Assuntos
Calcificação Fisiológica , Poliquetos/fisiologia , Proteoma/metabolismo , Aclimatação , Animais , Mudança Climática , Eletroforese em Gel Bidimensional , Concentração de Íons de Hidrogênio , Hipóxia/metabolismo , Larva/crescimento & desenvolvimento , Larva/fisiologia , Metamorfose Biológica , Oxigênio/análise , Oxigênio/metabolismo , Poliquetos/crescimento & desenvolvimento , Proteômica , Água do Mar/químicaRESUMO
OBJECTIVE: The traditional ICM is widely used in applications, such as image edge detection and image segmentation. However, several model parameters must be set, which tend to lead to reduced accuracy and increased cost. As medical images have more complex edges, contours and details, more suitable combinatorial algorithms are needed to handle the pathological diagnosis of multiple cerebral infarcts and acute strokes, resulting in the findings being more applicable, as well as having good clinical value. METHODS: To better solve the medical image fusion and diagnosis problems, this paper introduces the image fusion algorithm based on the combination of NSCT and improved ICM and proposes low-frequency, sub-band fusion rules and high-frequency sub-band fusion rules. The above method is applied to the fusion of CT/MRI images, subsequently, three other fusion algorithms, including NSCT-SF-PCNN, NSCT-SR-PCNN and Adaptive-PCNN are compared, and the simulation results of image fusion are analyzed and validated. RESULTS: According to the experimental findings, the suggested algorithm performs better than other fusion algorithms in terms of five objective evaluation metrics or subjective evaluation. The NSCT transform and the improved ICM were combined, and the outcomes were evaluated against those of other fusion algorithms. The CT/MRI medical images of healthy brain tissue, numerous cerebral infarcts and acute strokes were combined using this technique. CONCLUSION: Medical image fusion using Adaptive-PCNN produces satisfactory results, not only in relation to improved image clarity but also in terms of outstanding edge information, high contrast and brightness.
Assuntos
Acidente Vascular Cerebral , Córtex Visual , Humanos , Interpretação de Imagem Assistida por Computador/métodos , Redes Neurais de Computação , Algoritmos , Acidente Vascular Cerebral/diagnóstico por imagem , Infarto Cerebral/diagnóstico por imagem , Processamento de Imagem Assistida por Computador/métodosRESUMO
BACKGROUND AND OBJECTIVE: Diabetes is a disease that requires early detection and early treatment, and complications are likely to occur in late stages of the disease, threatening the life of patients. Therefore, in order to diagnose diabetic patients as early as possible, it is necessary to establish a model that can accurately predict diabetes. METHODOLOGY: This paper proposes an ensemble learning framework: KFPredict, which combines multi-input models with key features and machine learning algorithms. We first propose a multi-input neural network model (KF_NN) that fuses key features and uses a decision tree-based selection recursive feature elimination algorithm and correlation coefficient method to screen out the key feature inputs and secondary feature inputs in the model. We then ensemble KF_NN with three machine learning algorithms (i.e., Support Vector Machine, Random Forest and K-Nearest Neighbors) for soft voting to form our predictive classifier for diabetes prediction. RESULTS: Our framework demonstrates good prediction results on the test set with a sensitivity of 0.85, a specificity of 0.98, and an accuracy of 93.5%. Compared with the single prediction method KFPredict, the accuracy is up to 18.18% higher. Concurrently, we also compared KFPredict with the existing prediction methods. It still has good prediction performance, and the accuracy rate is improved by up to 14.93%. CONCLUSION: This paper constructs a diabetes prediction framework that combines multi-input models with key features and machine learning algorithms. Taking tthe PIMA diabetes dataset as the test data, the experiment shows that the framework presents good prediction results.
Assuntos
Diabetes Mellitus , Humanos , Algoritmos , Redes Neurais de Computação , Aprendizado de Máquina , Algoritmo Florestas Aleatórias , Máquina de Vetores de SuporteRESUMO
INTRODUCTION: In patients with acute ischemic stroke (AIS), the National Institutes of Health Stroke Scale (NIHSS) is essential to establishing a patient's initial stroke severity. While previous research has validated NIHSS scoring reliability between neurologists and other clinicians, it has not specifically evaluated NIHSS scoring reliability between emergency room (ER) and neurology physicians within the same clinical scenario and timeframe in a large cohort of patients. This study specifically addresses the key question: does an ER physician's NIHSS score agree with the neurologist's NIHSS score in the same patient at the same time in a real-world context? METHODS: Data was retrospectively collected from 1,946 patients being evaluated for AIS at Houston Methodist Hospital from 05/2016 - 04/2018. Triage NIHSS scores assessed by both the ER and neurology providers within one hour of each other under the same clinical context were evaluated for comparison. Ultimately, 129 patients were included in the analysis. All providers in this study were NIHSS rater-certified. RESULTS: The distribution of the NIHSS score differences (ER score - neurology score) had a mean of -0.46 and a standard deviation of 2.11. The score difference between provider teams ranged ±5 points. The intraclass correlation coefficient (ICC) for the NIHSS scores between the ER and neurology teams was 0.95 (95% CI, 0.93 - 0.97) with an F-test of 42.41 and a p-value of 4.43E-69. Overall reliability was excellent between the ER and neurology teams. CONCLUSION: We evaluated triage NIHSS scores performed by ER and neurology providers under matching time and treatment conditions and found excellent interrater reliability. The excellent score agreement has important implications for treatment decision-making during patient handoff and further in stroke modeling, prediction, and clinical trial registries where missing NIHSS scores may be equivalently substituted from either provider team.
RESUMO
The efficacy of acupuncture and moxibustion in the treatment of depression has been fully recognized internationally. However, its central mechanism is still not developed into a unified standard, and it is generally believed that the central mechanism is regulation of the cortical striatum thalamic neural pathway of the limbic system. In recent years, some scholars have applied functional magnetic resonance imaging (fMRI) to study the central mechanism and the associated brain effects of acupuncture and moxibustion treatment for depression. This study reviews the acupuncture and moxibustion treatment of depression from two aspects: (1) fMRI study of the brain function related to the acupuncture treatment of depression: different acupuncture and moxibustion methods are summarized, the fMRI technique is elaborately explained, and the results of fMRI study of the effects of acupuncture are analyzed in detail, and (2) fMRI associated "brain functional network" effects of acupuncture and moxibustion on depression, including the effects on the hippocampus, the amygdala, the cingulate gyrus, the frontal lobe, the temporal lobe, and other brain regions. The study of the effects of acupuncture on brain imaging is not adequately developed and still needs further improvement and development. The brain function networks associated with the acupuncture treatment of depression have not yet been adequately developed to provide a scientific and standardized mechanism of the effects of acupuncture. For this purpose, this study analyzes in-depth the clinical studies on the treatment of anxiety and depression by acupuncture and moxibustion, by depicting how the employment of fMRI technology provides significant imaging changes in the brain regions. Therefore, the study also provides a reference for future clinical research on the treatment of anxiety and depression.
RESUMO
BACKGROUND: Medical image classification is crucial for accurate and efficient diagnosis, and deep learning frameworks have shown significant potential in this area. When a general learning deep model is directly deployed to a new dataset with heterogeneous features, the effect of domain shifts is usually ignored, which degrades the performance of deep learning models and leads to inaccurate predictions. PURPOSE: This study aims to propose a framework that utilized the cross-modality domain adaptation and accurately diagnose and classify MRI scans and domain knowledge into stable and vulnerable plaque categories by a modified Vision Transformer (ViT) model for the classification of MRI scans and transformer model for domain knowledge classification. METHODS: This study proposes a Hybrid Vision Inspired Transformer (HViT) framework that employs a convolutional layer for image pre-processing and normalization and a 3D convolutional layer to enable ViT to classify 3D images. Our proposed HViT framework introduces a slim design with a multi-branch network and channel attention, improving patch embedding extraction and information learning. Auxiliary losses target shallow features, linking them with deeper ones, enhancing information gain, and model generalization. Furthermore, replacing the MLP Head with RNN enables better backpropagation for improved performance. Moreover, we utilized a modified transformer model with LSTM positional encoding and Golve word vector to classify domain knowledge. By using ensemble learning techniques, specifically stacking ensemble learning with hard and soft prediction, we combine the predictive power of both models to address the cross-modality domain adaptation problem and improve overall performance. RESULTS: The proposed framework achieved an accuracy of 94.32% for carotid artery plaque classification into stable and vulnerable plaque by addressing the cross-modality domain adaptation problem and improving overall performance. CONCLUSION: The model was further evaluated using an independent dataset acquired from different hardware protocols. The results demonstrate that the proposed deep learning model significantly improves the generalization ability across different MRI scans acquired from different hardware protocols without requiring additional calibration data.
Assuntos
Estenose das Carótidas , Humanos , Estenose das Carótidas/diagnóstico por imagem , Imageamento por Ressonância Magnética , Calibragem , Processamento de Imagem Assistida por ComputadorRESUMO
BACKGROUND: Brain stroke is a leading cause of disability and death worldwide, and early diagnosis and treatment are critical to improving patient outcomes. Current stroke diagnosis methods are subjective and prone to errors, as radiologists rely on manual selection of the most important CT slice. This highlights the need for more accurate and reliable automated brain stroke diagnosis and localization methods to improve patient outcomes. PURPOSE: In this study, we aimed to enhance the vision transformer architecture for the multi-slice classification of CT scans of each patient into three categories, including Normal, Infarction, Hemorrhage, and patient-wise stroke localization, based on end-to-end vision transformer architecture. This framework can provide an automated, objective, and consistent approach to stroke diagnosis and localization, enabling personalized treatment plans based on the location and extent of the stroke. METHODS: We modified the Vision Transformer (ViT) in combination with neural network layers for the multi-slice classification of brain CT scans of each patient into normal, infarction, and hemorrhage classes. For stroke localization, we used the ViT architecture and convolutional neural network layers to detect stroke and localize it by bounding boxes for infarction and hemorrhage regions in a patient-wise manner based on multi slices. RESULTS: Our proposed framework achieved an overall accuracy of 87.51% in classifying brain CT scan slices and showed high precision in localizing the stroke patient-wise. Our results demonstrate the potential of our method for accurate and reliable stroke diagnosis and localization. CONCLUSION: Our study enhanced ViT architecture for automated stroke diagnosis and localization using brain CT scans, which could have significant implications for stroke management and treatment. The use of deep learning algorithms can provide a more objective and consistent approach to stroke diagnosis and potentially enable personalized treatment plans based on the location and extent of the stroke. Further studies are needed to validate our method on larger and more diverse datasets and to explore its clinical utility in real-world settings.
Assuntos
Encéfalo , Acidente Vascular Cerebral , Humanos , Encéfalo/diagnóstico por imagem , Acidente Vascular Cerebral/diagnóstico por imagem , Tomografia Computadorizada por Raios X , Hemorragia , InfartoRESUMO
BACKGROUND AND OBJECTIVE: Traditional disease diagnosis is usually performed by experienced physicians, but misdiagnosis or missed diagnosis still exists. Exploring the relationship between changes in the corpus callosum and multiple brain infarcts requires extracting corpus callosum features from brain image data, which requires addressing three key issues. (1) automation, (2) completeness, and (3) accuracy. Residual learning can facilitate network training, Bi-Directional Convolutional LSTM (BDC-LSTM) can exploit interlayer spatial dependencies, and HDC can expand the receptive domain without losing resolution. METHODS: In this paper, we propose a segmentation method by combining BDC-LSTM and U-Net to segment the corpus callosum from multiple angles of brain images based on computed tomography (CT) and magnetic resonance imaging (MRI) in which two types of sequence, namely T2-weighted imaging as well as the Fluid Attenuated Inversion Recovery (Flair), were utilized. The two-dimensional slice sequences are segmented in the cross-sectional plane, and the segmentation results are combined to obtain the final results. Encoding, BDC- LSTM, and decoding include convolutional neural networks. The coding part uses asymmetric convolutional layers of different sizes and dilated convolutions to get multi-slice information and extend the convolutional layers' perceptual field. RESULTS: This paper uses BDC-LSTM between the encoding and decoding parts of the algorithm. On the image segmentation of the brain in multiple cerebral infarcts dataset, accuracy rates of 0.876, 0.881, 0.887, and 0.912 were attained for the intersection of union (IOU), dice similarity coefficient (DS), sensitivity (SE), and predictive positivity value (PPV). The experimental findings demonstrate that the algorithm outperforms its rivals in accuracy. CONCLUSION: This paper obtained segmentation results for three images using three models, ConvLSTM, Pyramid-LSTM, and BDC-LSTM, and compared them to verify that BDC-LSTM is the best method to perform the segmentation task for faster and more accurate detection of 3D medical images. We improve the convolutional neural network segmentation method to obtain medical images with high segmentation accuracy by solving the over-segmentation problem.
Assuntos
Corpo Caloso , Processamento de Imagem Assistida por Computador , Processamento de Imagem Assistida por Computador/métodos , Corpo Caloso/diagnóstico por imagem , Estudos Transversais , Imageamento por Ressonância Magnética/métodos , Tomografia Computadorizada por Raios XRESUMO
Stroke is one of the leading causes of death and disability in the world. Despite intensive research on automatic stroke lesion segmentation from non-invasive imaging modalities including diffusion-weighted imaging (DWI), challenges remain such as a lack of sufficient labeled data for training deep learning models and failure in detecting small lesions. In this paper, we propose BBox-Guided Segmentor, a method that significantly improves the accuracy of stroke lesion segmentation by leveraging expert knowledge. Specifically, our model uses a very coarse bounding box label provided by the expert and then performs accurate segmentation automatically. The small overhead of having the expert provide a rough bounding box leads to large performance improvement in segmentation, which is paramount to accurate stroke diagnosis. To train our model, we employ a weakly-supervised approach that uses a large number of weakly-labeled images with only bounding boxes and a small number of fully labeled images. The scarce fully labeled images are used to train a generator segmentation network, while adversarial training is used to leverage the large number of weakly-labeled images to provide additional learning signals. We evaluate our method extensively using a unique clinical dataset of 99 fully labeled cases (i.e., with full segmentation map labels) and 831 weakly labeled cases (i.e., with only bounding box labels), and the results demonstrate the superior performance of our approach over state-of-the-art stroke lesion segmentation models. We also achieve competitive performance as a SOTA fully supervised method using less than one-tenth of the complete labels. Our proposed approach has the potential to improve stroke diagnosis and treatment planning, which may lead to better patient outcomes.
Assuntos
Imagem de Difusão por Ressonância Magnética , Acidente Vascular Cerebral , Humanos , Acidente Vascular Cerebral/diagnóstico por imagem , Processamento de Imagem Assistida por ComputadorRESUMO
BACKGROUND: Although patients harboring confluent white matter hyperintensities (WMH) are at high risk of cognitive decline, this risk varies among individuals. We investigated the predictors for cognitive decline in stroke patients with confluent WMH. METHODS: We followed up 100 stroke patients with confluent WMH who were participants of the VITAmins TO Prevent Stroke study for 2 years. We investigated the association between clinical features, apolipoprotein E status, imaging measures (infarcts, microbleeds, volumes of WMH, cortical gray matter [cGM], lateral ventricles, and hippocampi), and B vitamins with changes in cognitive measures (clinical dementia rating scale, Mini-Mental State Examination, Mattis dementia rating scale--initiation/perseveration subscale). We performed Pittsburgh compound B imaging among dementia converters. RESULTS: Multivariate regression analysis showed that increase in clinical dementia rating scale grade was associated with cGM atrophy, absence of hyperlipidemia, and lower diastolic blood pressure at baseline. cGM atrophy and absence of hyperlipidemia were also associated with deterioration in Mini-Mental State Examination and Mattis dementia rating scale--initiation/perseveration subscale scores. Pittsburgh compound B retention typical of Alzheimer's disease was found only in 10% of dementia converters. Incident stroke and B vitamins were not associated with cognitive decline. CONCLUSIONS: Among stroke patients with confluent WMH, cGM atrophy and absence of hyperlipidemia are important predictors for cognitive decline. Significant cognitive decline can occur in the absence of incident stroke or Alzheimer's pathology.