RESUMO
Affect recognition in a real-world, less constrained environment is the principal prerequisite of the industrial-level usefulness of this technology. Monitoring the psychological profile using smart, wearable electroencephalogram (EEG) sensors during daily activities without external stimuli, such as memory-induced emotions, is a challenging research gap in emotion recognition. This paper proposed a deep learning framework for improved memory-induced emotion recognition leveraging a combination of 1D-CNN and LSTM as feature extractors integrated with an Extreme Learning Machine (ELM) classifier. The proposed deep learning architecture, combined with the EEG preprocessing, such as the removal of the average baseline signal from each sample and extraction of EEG rhythms (delta, theta, alpha, beta, and gamma), aims to capture repetitive and continuous patterns for memory-induced emotion recognition, underexplored with deep learning techniques. This work has analyzed EEG signals using a wearable, ultra-mobile sports cap while recalling autobiographical emotional memories evoked by affect-denoting words, with self-annotation on the scale of valence and arousal. With extensive experimentation using the same dataset, the proposed framework empirically outperforms existing techniques for the emerging area of memory-induced emotion recognition with an accuracy of 65.6%. The EEG rhythms analysis, such as delta, theta, alpha, beta, and gamma, achieved 65.5%, 52.1%, 65.1%, 64.6%, and 65.0% accuracies for classification with four quadrants of valence and arousal. These results underscore the significant advancement achieved by our proposed method for the real-world environment of memory-induced emotion recognition.
Assuntos
Aprendizado Profundo , Eletroencefalografia , Emoções , Rememoração Mental , Humanos , Eletroencefalografia/métodos , Emoções/fisiologia , Rememoração Mental/fisiologia , Masculino , Feminino , Adulto , Adulto JovemRESUMO
BACKGROUND: The high mortality rate associated with coronary heart disease has led to state-of-the-art non-invasive methods for cardiac diagnosis including computed tomography and magnetic resonance imaging. However, stenosis computation and clinical assessment of non-calcified plaques has been very challenging due to their ambiguous intensity response in CT i.e. a significant overlap with surrounding muscle tissues and blood. Accordingly, this research presents an approach for computation of coronary stenosis by investigating cross-sectional lumen behaviour along the length of 3D coronary segments. METHODS: Non-calcified plaques are characterized by comparatively lower-intensity values with respect to the surrounding. Accordingly, segment-wise orthogonal volume was reconstructed in 3D space using the segmented coronary tree. Subsequently, the cross sectional volumetric data was investigated using proposed CNN-based plaque quantification model and subsequent stenosis grading in clinical context was performed. In the last step, plaque-affected orthogonal volume was further investigated by comparing vessel-wall thickness and lumen area obstruction w.r.t. expert-based annotations to validate the stenosis grading performance of model. RESULTS: The experimental data consists of clinical CT images obtained from the Rotterdam CT repository leading to 600 coronary segments and subsequent 15786 cross-sectional images. According to the results, the proposed method quantified coronary vessel stenosis i.e. severity of the non-calcified plaque with an overall accuracy of 83%. Moreover, for individual grading, the proposed model show promising results with accuracy equal to 86%, 90% and 79% respectively for severe, moderate and mild stenosis. The stenosis grading performance of the proposed model was further validated by performing lumen-area versus wall-thickness analysis as per annotations of manual experts. The statistical results for lumen area analysis precisely correlates with the quantification performance of the model with a mean deviation of 5% only. CONCLUSION: The overall results demonstrates capability of the proposed model to grade the vessel stenosis with reasonable accuracy and precision equivalent to human experts.
Assuntos
Estenose Coronária , Placa Aterosclerótica , Tomografia Computadorizada por Raios X , Estenose Coronária/diagnóstico por imagem , Humanos , Placa Aterosclerótica/diagnóstico por imagem , Meios de Contraste , MasculinoRESUMO
The Oxford English Dictionary includes 17 definitions for the word "model" as a noun and another 11 as a verb. Therefore, context is necessary to understand the meaning of the word model. For instance, "model railways" refer to replicas of railways and trains at a smaller scale and a "model student" refers to an exemplary individual. In some cases, a specific context, like cancer research, may not be sufficient to provide one specific meaning for model. Even if the context is narrowed, specifically, to research related to the tumour microenvironment, "model" can be understood in a wide variety of ways, from an animal model to a mathematical expression. This paper presents a review of different "models" of the tumour microenvironment, as grouped by different definitions of the word into four categories: model organisms, in vitro models, mathematical models and computational models. Then, the frequencies of different meanings of the word "model" related to the tumour microenvironment are measured from numbers of entries in the MEDLINE database of the United States National Library of Medicine at the National Institutes of Health. The frequencies of the main components of the microenvironment and the organ-related cancers modelled are also assessed quantitatively with specific keywords. Whilst animal models, particularly xenografts and mouse models, are the most commonly used "models", the number of these entries has been slowly decreasing. Mathematical models, as well as prognostic and risk models, follow in frequency, and these have been growing in use.
RESUMO
This paper investigates the impact of the amount of training data and the shape variability on the segmentation provided by the deep learning architecture U-Net. Further, the correctness of ground truth (GT) was also evaluated. The input data consisted of a three-dimensional set of images of HeLa cells observed with an electron microscope with dimensions 8192×8192×517. From there, a smaller region of interest (ROI) of 2000×2000×300 was cropped and manually delineated to obtain the ground truth necessary for a quantitative evaluation. A qualitative evaluation was performed on the 8192×8192 slices due to the lack of ground truth. Pairs of patches of data and labels for the classes nucleus, nuclear envelope, cell and background were generated to train U-Net architectures from scratch. Several training strategies were followed, and the results were compared against a traditional image processing algorithm. The correctness of GT, that is, the inclusion of one or more nuclei within the region of interest was also evaluated. The impact of the extent of training data was evaluated by comparing results from 36,000 pairs of data and label patches extracted from the odd slices in the central region, to 135,000 patches obtained from every other slice in the set. Then, 135,000 patches from several cells from the 8192×8192 slices were generated automatically using the image processing algorithm. Finally, the two sets of 135,000 pairs were combined to train once more with 270,000 pairs. As would be expected, the accuracy and Jaccard similarity index improved as the number of pairs increased for the ROI. This was also observed qualitatively for the 8192×8192 slices. When the 8192×8192 slices were segmented with U-Nets trained with 135,000 pairs, the architecture trained with automatically generated pairs provided better results than the architecture trained with the pairs from the manually segmented ground truths. This suggests that the pairs that were extracted automatically from many cells provided a better representation of the four classes of the various cells in the 8192×8192 slice than those pairs that were manually segmented from a single cell. Finally, the two sets of 135,000 pairs were combined, and the U-Net trained with these provided the best results.
RESUMO
In this work, the performance of five deep learning architectures in classifying COVID-19 in a multi-class set-up is evaluated. The classifiers were built on pretrained ResNet-50, ResNet-50r (with kernel size 5×5 in the first convolutional layer), DenseNet-121, MobileNet-v3 and the state-of-the-art CaiT-24-XXS-224 (CaiT) transformer. The cross entropy and weighted cross entropy were minimised with Adam and AdamW. In total, 20 experiments were conducted with 10 repetitions and obtained the following metrics: accuracy (Acc), balanced accuracy (BA), F1 and F2 from the general Fß macro score, Matthew's Correlation Coefficient (MCC), sensitivity (Sens) and specificity (Spec) followed by bootstrapping. The performance of the classifiers was compared by using the Friedman-Nemenyi test. The results show that less complex architectures such as ResNet-50, ResNet-50r and DenseNet-121 were able to achieve better generalization with rankings of 1.53, 1.71 and 3.05 for the Matthew Correlation Coefficient, respectively, while MobileNet-v3 and CaiT obtained rankings of 3.72 and 5.0, respectively.
RESUMO
This paper investigates the classification of radiographic images with eleven convolutional neural network (CNN) architectures (GoogleNet, VGG-19, AlexNet, SqueezeNet, ResNet-18, Inception-v3, ResNet-50, VGG-16, ResNet-101, DenseNet-201 and Inception-ResNet-v2). The CNNs were used to classify a series of wrist radiographs from the Stanford Musculoskeletal Radiographs (MURA) dataset into two classes-normal and abnormal. The architectures were compared for different hyper-parameters against accuracy and Cohen's kappa coefficient. The best two results were then explored with data augmentation. Without the use of augmentation, the best results were provided by Inception-ResNet-v2 (Mean accuracy = 0.723, Mean kappa = 0.506). These were significantly improved with augmentation to Inception-ResNet-v2 (Mean accuracy = 0.857, Mean kappa = 0.703). Finally, Class Activation Mapping was applied to interpret activation of the network against the location of an anomaly in the radiographs.
Assuntos
Redes Neurais de Computação , RadiografiaRESUMO
The segmentation of power lines (PLs) from aerial images is a crucial task for the safe navigation of unmanned aerial vehicles (UAVs) operating at low altitudes. Despite the advances in deep learning-based approaches for PL segmentation, these models are still vulnerable to the class imbalance present in the data. The PLs occupy only a minimal portion (1-5%) of the aerial images as compared to the background region (95-99%). Generally, this class imbalance problem is addressed via the use of PL-specific detectors in conjunction with the popular class balanced cross entropy (BBCE) loss function. However, these PL-specific detectors do not work outside their application areas and a BBCE loss requires hyperparameter tuning for class-wise weights, which is not trivial. Moreover, the BBCE loss results in low dice scores and precision values and thus, fails to achieve an optimal trade-off between dice scores, model accuracy, and precision-recall values. In this work, we propose a generalized focal loss function based on the Matthews correlation coefficient (MCC) or the Phi coefficient to address the class imbalance problem in PL segmentation while utilizing a generic deep segmentation architecture. We evaluate our loss function by improving the vanilla U-Net model with an additional convolutional auxiliary classifier head (ACU-Net) for better learning and faster model convergence. The evaluation of two PL datasets, namely the Mendeley Power Line Dataset and the Power Line Dataset of Urban Scenes (PLDU), where PLs occupy around 1% and 2% of the aerial images area, respectively, reveal that our proposed loss function outperforms the popular BBCE loss by 16% in PL dice scores on both the datasets, 19% in precision and false detection rate (FDR) values for the Mendeley PL dataset and 15% in precision and FDR values for the PLDU with a minor degradation in the accuracy and recall values. Moreover, our proposed ACU-Net outperforms the baseline vanilla U-Net for the characteristic evaluation parameters in the range of 1-10% for both the PL datasets. Thus, our proposed loss function with ACU-Net achieves an optimal trade-off for the characteristic evaluation parameters without any bells and whistles. Our code is available at Github.
RESUMO
In this work, an unsupervised volumetric semantic instance segmentation of the plasma membrane of HeLa cells as observed with serial block face scanning electron microscopy is described. The resin background of the images was segmented at different slices of a 3D stack of 518 slices with 8192 × 8192 pixels each. The background was used to create a distance map, which helped identify and rank the cells by their size at each slice. The centroids of the cells detected at different slices were linked to identify them as a single cell that spanned a number of slices. A subset of these cells, i.e., the largest ones and those not close to the edges were selected for further processing. The selected cells were then automatically cropped to smaller regions of interest of 2000 × 2000 × 300 voxels that were treated as cell instances. Then, for each of these volumes, the nucleus was segmented, and the cell was separated from any neighbouring cells through a series of traditional image processing steps that followed the plasma membrane. The segmentation process was repeated for all the regions of interest previously selected. For one cell for which the ground truth was available, the algorithm provided excellent results in Accuracy (AC) and the Jaccard similarity Index (JI): nucleus: JI =0.9665, AC =0.9975, cell including nucleus JI =0.8711, AC =0.9655, cell excluding nucleus JI =0.8094, AC =0.9629. A limitation of the algorithm for the plasma membrane segmentation was the presence of background. In samples with tightly packed cells, this may not be available. When tested for these conditions, the segmentation of the nuclear envelope was still possible. All the code and data were released openly through GitHub, Zenodo and EMPIAR.
RESUMO
Histological evaluation plays a major role in cancer diagnosis and treatment. The appearance of H&E-stained images can vary significantly as a consequence of differences in several factors, such as reagents, staining conditions, preparation procedure and image acquisition system. Such potential sources of noise can all have negative effects on computer-assisted classification. To minimize such artefacts and their potentially negative effects several color pre-processing methods have been proposed in the literature-for instance, color augmentation, color constancy, color deconvolution and color transfer. Still, little work has been done to investigate the efficacy of these methods on a quantitative basis. In this paper, we evaluated the effects of color constancy, deconvolution and transfer on automated classification of H&E-stained images representing different types of cancers-specifically breast, prostate, colorectal cancer and malignant lymphoma. Our results indicate that in most cases color pre-processing does not improve the classification accuracy, especially when coupled with color-based image descriptors. Some pre-processing methods, however, can be beneficial when used with some texture-based methods like Gabor filters and Local Binary Patterns.
RESUMO
Accurate measurements of cell morphology and behaviour are fundamentally important for understanding how disease, molecules and drugs affect cell function in vivo Here, by using muscle stem cell (muSC) responses to injury in zebrafish as our biological paradigm, we established a 'ground truth' for muSC behaviour. This revealed that segmentation and tracking algorithms from commonly used programs are error-prone, leading us to develop a fast semi-automated image analysis pipeline that allows user-defined parameters for segmentation and correction of cell tracking. Cell Tracking Profiler (CTP) is a package that runs two existing programs, HK Means and Phagosight within the Icy image analysis suite, to enable user-managed cell tracking from 3D time-lapse datasets to provide measures of cell shape and movement. We demonstrate how CTP can be used to reveal changes to cell behaviour of muSCs in response to manipulation of the cell cytoskeleton by small-molecule inhibitors. CTP and the associated tools we have developed for analysis of outputs thus provide a powerful framework for analysing complex cell behaviour in vivo from 4D datasets that are not amenable to straightforward analysis.
Assuntos
Rastreamento de Células , Peixe-Zebra , Algoritmos , Animais , Processamento de Imagem Assistida por Computador , Imageamento Tridimensional , MovimentoRESUMO
The quantitative study of cell morphology is of great importance as the structure and condition of cells and their structures can be related to conditions of health or disease. The first step towards that, is the accurate segmentation of cell structures. In this work, we compare five approaches, one traditional and four deep-learning, for the semantic segmentation of the nuclear envelope of cervical cancer cells commonly known as HeLa cells. Images of a HeLa cancer cell were semantically segmented with one traditional image-processing algorithm and four three deep learning architectures: VGG16, ResNet18, Inception-ResNet-v2, and U-Net. Three hundred slices, each 2000 × 2000 pixels, of a HeLa Cell were acquired with Serial Block Face Scanning Electron Microscopy. The first three deep learning architectures were pre-trained with ImageNet and then fine-tuned with transfer learning. The U-Net architecture was trained from scratch with 36, 000 training images and labels of size 128 × 128. The image-processing algorithm followed a pipeline of several traditional steps like edge detection, dilation and morphological operators. The algorithms were compared by measuring pixel-based segmentation accuracy and Jaccard index against a labelled ground truth. The results indicated a superior performance of the traditional algorithm (Accuracy = 99%, Jaccard = 93%) over the deep learning architectures: VGG16 (93%, 90%), ResNet18 (94%, 88%), Inception-ResNet-v2 (94%, 89%), and U-Net (92%, 56%).
Assuntos
Células HeLa/citologia , Processamento de Imagem Assistida por Computador/métodos , Algoritmos , Aprendizado Profundo , Humanos , Microscopia de Força AtômicaRESUMO
Fractures of the wrist are common in Emergency Departments, where some patients are treated with a procedure called Manipulation under Anaesthesia. In some cases, this procedure is unsuccessful and patients need to revisit the hospital where they undergo surgery to treat the fracture. This work describes a geometric semi-automatic image analysis algorithm to analyse and compare the x-rays of healthy controls and patients with dorsally displaced wrist fractures (Colles' fractures) who were treated with Manipulation under Anaesthesia. A series of 161 posterior-anterior radiographs from healthy controls and patients with Colles' fractures were acquired and analysed. The patients' group was further subdivided according to the outcome of the procedure (successful/unsuccessful) and pre- or post-intervention creating five groups in total (healthy, pre-successful, pre-unsuccessful, post-successful, post-unsuccessful). The semi-automatic analysis consisted of manual location of three landmarks (finger, lunate and radial styloid) and automatic processing to generate 32 geometric and texture measurements, which may be related to conditions such as osteoporosis and swelling of the wrist. Statistical differences were found between patients and controls, as well as between pre- and post-intervention, but not between the procedures. The most distinct measurements were those of texture. Although the study includes a relatively low number of cases and measurements, the statistical differences are encouraging.
Assuntos
Fratura de Colles/diagnóstico por imagem , Fratura de Colles/terapia , Interpretação de Imagem Radiográfica Assistida por Computador/métodos , Adulto , Idoso , Idoso de 80 Anos ou mais , Estudos de Casos e Controles , Feminino , Humanos , Masculino , Manipulação Ortopédica , Pessoa de Meia-Idade , Radiografia , Resultado do TratamentoRESUMO
Centrosome separation in late G2/ early prophase requires precise spatial coordination that is determined by a balance of forces promoting and antagonizing separation. The major effector of centrosome separation is the kinesin Eg5. However, the identity and regulation of Eg5-antagonizing forces is less well characterized. By manipulating candidate components, we find that centrosome separation is reversible and that separated centrosomes congress toward a central position underneath the flat nucleus. This positioning mechanism requires microtubule polymerization, as well as actin polymerization. We identify perinuclear actin structures that form in late G2/early prophase and interact with microtubules emanating from the centrosomes. Disrupting these structures by breaking the interactions of the linker of nucleoskeleton and cytoskeleton (LINC) complex with perinuclear actin filaments abrogates this centrosome positioning mechanism and causes an increase in subsequent chromosome segregation errors. Our results demonstrate how geometrical cues from the cell nucleus coordinate the orientation of the emanating spindle poles before nuclear envelope breakdown.
Assuntos
Actinas/metabolismo , Centrossomo/metabolismo , Segregação de Cromossomos/genética , Prófase/genética , HumanosRESUMO
This paper describes a methodology that extracts key morphological features from histological breast cancer images in order to automatically assess Tumour Cellularity (TC) in Neo-Adjuvant treatment (NAT) patients. The response to NAT gives information on therapy efficacy and it is measured by the residual cancer burden index, which is composed of two metrics: TC and the assessment of lymph nodes. The data consist of whole slide images (WSIs) of breast tissue stained with Hematoxylin and Eosin (H&E) released in the 2019 SPIE Breast Challenge. The methodology proposed is based on traditional computer vision methods (K-means, watershed segmentation, Otsu's binarisation, and morphological operations), implementing colour separation, segmentation, and feature extraction. Correlation between morphological features and the residual TC after a NAT treatment was examined. Linear regression and statistical methods were used and twenty-two key morphological parameters from the nuclei, epithelial region, and the full image were extracted. Subsequently, an automated TC assessment that was based on Machine Learning (ML) algorithms was implemented and trained with only selected key parameters. The methodology was validated with the score assigned by two pathologists through the intra-class correlation coefficient (ICC). The selection of key morphological parameters improved the results reported over other ML methodologies and it was very close to deep learning methodologies. These results are encouraging, as a traditionally-trained ML algorithm can be useful when limited training data are available preventing the use of deep learning approaches.
RESUMO
In this paper, a novel method for interaction detection is presented to compare the contact dynamics of macrophages in the Drosophila embryo. The study is carried out by a framework called macrosight, which analyses the movement and interaction of migrating macrophages. The framework incorporates a segmentation and tracking algorithm into analysing the motion characteristics of cells after contact. In this particular study, the interactions between cells is characterised in the case of control embryos and Shot mutants, a candidate protein that is hypothesised to regulate contact dynamics between migrating cells. Statistical significance between control and mutant cells was found when comparing the direction of motion after contact in specific conditions. Such discoveries provide insights for future developments in combining biological experiments with computational analysis.
RESUMO
The initial host response to fungal pathogen invasion is critical to infection establishment and outcome. However, the diversity of leukocyte-pathogen interactions is only recently being appreciated. We describe a new form of interleukocyte conidial exchange called "shuttling." In Talaromyces marneffei and Aspergillus fumigatus zebrafish in vivo infections, live imaging demonstrated conidia initially phagocytosed by neutrophils were transferred to macrophages. Shuttling is unidirectional, not a chance event, and involves alterations of phagocyte mobility, intercellular tethering, and phagosome transfer. Shuttling kinetics were fungal-species-specific, implicating a fungal determinant. ß-glucan serves as a fungal-derived signal sufficient for shuttling. Murine phagocytes also shuttled in vitro. The impact of shuttling for microbiological outcomes of in vivo infections is difficult to specifically assess experimentally, but for these two pathogens, shuttling augments initial conidial redistribution away from fungicidal neutrophils into the favorable macrophage intracellular niche. Shuttling is a frequent host-pathogen interaction contributing to fungal infection establishment patterns.
Assuntos
Aspergilose/imunologia , Interações Hospedeiro-Patógeno , Macrófagos/fisiologia , Neutrófilos/fisiologia , beta-Glucanas/imunologia , Animais , Aspergillus fumigatus , Camundongos , Fagocitose , Fagossomos , Esporos Fúngicos , Talaromyces , Peixe-ZebraRESUMO
BACKGROUND: For virtually every patient with colorectal cancer (CRC), hematoxylin-eosin (HE)-stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can extract prognosticators directly from these widely available images. METHODS AND FINDINGS: We hand-delineated single-tissue regions in 86 CRC tissue slides, yielding more than 100,000 HE image patches, and used these to train a CNN by transfer learning, reaching a nine-class accuracy of >94% in an independent data set of 7,180 images from 25 CRC patients. With this tool, we performed automated tissue decomposition of representative multitissue HE images from 862 HE slides in 500 stage I-IV CRC patients in the The Cancer Genome Atlas (TCGA) cohort, a large international multicenter collection of CRC tissue. Based on the output neuron activations in the CNN, we calculated a "deep stroma score," which was an independent prognostic factor for overall survival (OS) in a multivariable Cox proportional hazard model (hazard ratio [HR] with 95% confidence interval [CI]: 1.99 [1.27-3.12], p = 0.0028), while in the same cohort, manual quantification of stromal areas and a gene expression signature of cancer-associated fibroblasts (CAFs) were only prognostic in specific tumor stages. We validated these findings in an independent cohort of 409 stage I-IV CRC patients from the "Darmkrebs: Chancen der Verhütung durch Screening" (DACHS) study who were recruited between 2003 and 2007 in multiple institutions in Germany. Again, the score was an independent prognostic factor for OS (HR 1.63 [1.14-2.33], p = 0.008), CRC-specific OS (HR 2.29 [1.5-3.48], p = 0.0004), and relapse-free survival (RFS; HR 1.92 [1.34-2.76], p = 0.0004). A prospective validation is required before this biomarker can be implemented in clinical workflows. CONCLUSIONS: In our retrospective study, we show that a CNN can assess the human tumor microenvironment and predict prognosis directly from histopathological images.