RESUMO

This article addresses the semantic segmentation of laparoscopic surgery images, placing special emphasis on the segmentation of structures with a smaller number of observations. As a result of this study, adjustment parameters are proposed for deep neural network architectures, enabling a robust segmentation of all structures in the surgical scene. The U-Net architecture with five encoder-decoders (U-Net5ed), SegNet-VGG19, and DeepLabv3+ employing different backbones are implemented. Three main experiments are conducted, working with Rectified Linear Unit (ReLU), Gaussian Error Linear Unit (GELU), and Swish activation functions. The applied loss functions include Cross Entropy (CE), Focal Loss (FL), Tversky Loss (TL), Dice Loss (DiL), Cross Entropy Dice Loss (CEDL), and Cross Entropy Tversky Loss (CETL). The performance of Stochastic Gradient Descent with momentum (SGDM) and Adaptive Moment Estimation (Adam) optimizers is compared. It is qualitatively and quantitatively confirmed that DeepLabv3+ and U-Net5ed architectures yield the best results. The DeepLabv3+ architecture with the ResNet-50 backbone, Swish activation function, and CETL loss function reports a Mean Accuracy (MAcc) of 0.976 and Mean Intersection over Union (MIoU) of 0.977. The semantic segmentation of structures with a smaller number of observations, such as the hepatic vein, cystic duct, Liver Ligament, and blood, verifies that the obtained results are very competitive and promising compared to the consulted literature. The proposed selected parameters were validated in the YOLOv9 architecture, which showed an improvement in semantic segmentation compared to the results obtained with the original architecture.

RESUMO

In recent decades, many studies have been published on the use of automatic smear microscopy for diagnosing pulmonary tuberculosis (TB). Most of them deal with a preliminary step of the diagnosis, the bacilli detection, whereas sputum smear microscopy for diagnosis of pulmonary TB comprises detecting and reporting the number of bacilli found in at least 100 microscopic fields, according to the 5 grading scales (negative, scanty, 1+, 2+ and 3+) endorsed by the World Health Organization (WHO). Pulmonary TB diagnosis in bright-field smear microscopy, however, depends upon the attention of a trained and motivated technician, while the automated TB diagnosis requires little or no interpretation by a technician. As far as we know, this work proposes the first automatic method for pulmonary TB diagnosis in bright-field smear microscopy, according to the WHO recommendations. The proposed method comprises a semantic segmentation step, using a deep neural network, followed by a filtering step aiming to reduce the number of false positives (false bacilli): color and shape filtering. In semantic segmentation, different configurations of encoders are evaluated, using depth-wise separable convolution layers and channel attention mechanism. The proposed method was evaluated with a large, robust, and annotated image dataset designed for this purpose, consisting of 250 testing sets, 50 sets for each of the 5 TB diagnostic classes. The following performance metrics were obtained for automatic pulmonary TB diagnosis by smear microscopy: mean precision of 0.894, mean recall of 0.896, and mean F1-score of 0.895.

Assuntos

RESUMO

Most of the methods for real-time semantic segmentation do not take into account temporal information when working with video sequences. This is counter-intuitive in real-world scenarios where the main application of such methods is, precisely, being able to process frame sequences as quickly and accurately as possible. In this paper, we address this problem by exploiting the temporal information provided by previous frames of the video stream. Our method leverages a previous input frame as well as the previous output of the network to enhance the prediction accuracy of the current input frame. We develop a module that obtains feature maps rich in change information. Additionally, we incorporate the previous output of the network into all the decoder stages as a way of increasing the attention given to relevant features. Finally, to properly train and evaluate our methods, we introduce CityscapesVid, a dataset specifically designed to benchmark semantic video segmentation networks. Our proposed network, entitled FASSVid improves the mIoU accuracy performance over a standard non-sequential baseline model. Moreover, FASSVid obtains state-of-the-art inference speed and competitive mIoU results compared to other state-of-the-art lightweight networks, with significantly lower number of computations. Specifically, we obtain 71% of mIoU in our CityscapesVid dataset, running at 114.9 FPS on a single NVIDIA GTX 1080Ti and 31 FPS on the NVIDIA Jetson Nano embedded board with images of size 1024×2048 and 512×1024, respectively.

RESUMO

Single image depth estimation works fail to separate foreground elements because they can easily be confounded with the background. To alleviate this problem, we propose the use of a semantic segmentation procedure that adds information to a depth estimator, in this case, a 3D Convolutional Neural Network (CNN)-segmentation is coded as one-hot planes representing categories of objects. We explore 2D and 3D models. Particularly, we propose a hybrid 2D-3D CNN architecture capable of obtaining semantic segmentation and depth estimation at the same time. We tested our procedure on the SYNTHIA-AL dataset and obtained σ3=0.95, which is an improvement of 0.14 points (compared with the state of the art of σ3=0.81) by using manual segmentation, and σ3=0.89 using automatic semantic segmentation, proving that depth estimation is improved when the shape and position of objects in a scene are known.

Assuntos

RESUMO

Automatic flood detection may be an important component for triggering damage control systems and minimizing the risk of social or economic impacts caused by flooding. Riverside images from regular cameras are a widely available resource that can be used for tackling this problem. Nevertheless, state-of-the-art neural networks, the most suitable approach for this type of computer vision task, are usually resource-consuming, which poses a challenge for deploying these models within low-capability Internet of Things (IoT) devices with unstable internet connections. In this work, we propose a deep neural network (DNN) architecture pruning algorithm capable of finding a pruned version of a given DNN within a user-specified memory footprint. Our results demonstrate that our proposed algorithm can find a pruned DNN model with the specified memory footprint with little to no degradation of its segmentation performance. Finally, we show that our algorithm can be used in a memory-constraint wireless sensor network (WSN) employed to detect flooding events of urban rivers, and the resulting pruned models have competitive results compared with the original models.

Assuntos

RESUMO

COVID-19 frequently provokes pneumonia, which can be diagnosed using imaging exams. Chest X-ray (CXR) is often useful because it is cheap, fast, widespread, and uses less radiation. Here, we demonstrate the impact of lung segmentation in COVID-19 identification using CXR images and evaluate which contents of the image influenced the most. Semantic segmentation was performed using a U-Net CNN architecture, and the classification using three CNN architectures (VGG, ResNet, and Inception). Explainable Artificial Intelligence techniques were employed to estimate the impact of segmentation. A three-classes database was composed: lung opacity (pneumonia), COVID-19, and normal. We assessed the impact of creating a CXR image database from different sources, and the COVID-19 generalization from one source to another. The segmentation achieved a Jaccard distance of 0.034 and a Dice coefficient of 0.982. The classification using segmented images achieved an F1-Score of 0.88 for the multi-class setup, and 0.83 for COVID-19 identification. In the cross-dataset scenario, we obtained an F1-Score of 0.74 and an area under the ROC curve of 0.9 for COVID-19 identification using segmented images. Experiments support the conclusion that even after segmentation, there is a strong bias introduced by underlying factors from different sources.

Assuntos

RESUMO

This study proposes and evaluates five deep fully convolutional networks (FCNs) for the semantic segmentation of a single tree species: SegNet, U-Net, FC-DenseNet, and two DeepLabv3+ variants. The performance of the FCN designs is evaluated experimentally in terms of classification accuracy and computational load. We also verify the benefits of fully connected conditional random fields (CRFs) as a post-processing step to improve the segmentation maps. The analysis is conducted on a set of images captured by an RGB camera aboard a UAV flying over an urban area. The dataset also contains a mask that indicates the occurrence of an endangered species called Dipteryx alata Vogel, also known as cumbaru, taken as the species to be identified. The experimental analysis shows the effectiveness of each design and reports average overall accuracy ranging from 88.9% to 96.7%, an F1-score between 87.0% and 96.1%, and IoU from 77.1% to 92.5%. We also realize that CRF consistently improves the performance, but at a high computational cost.

RESUMO

Existing computational approaches have not yet resulted in effective and efficient computer-aided tools that are used in pathologists' daily practice. Focusing on a computer-based qualification for breast cancer diagnosis, the present study proposes two deep learning architectures to efficiently and effectively detect and classify mitosis in a histopathological tissue sample. The first method consists of two parts, entailing a preprocessing of the digital histological image and a free-handcrafted-feature Convolutional Neural Network (CNN) used for binary classification. Results show that the methodology proposed can achieve 95% accuracy in testing, with an F1-score of 94.35%. This result is higher than the results using classical image processing techniques and also higher than the approaches combining CCNs with handcrafted features. The second approach is an end-to-end methodology using semantic segmentation. Results showed that this algorithm can achieve an accuracy higher than 95% in testing and an average Dice index of 0.6, higher than the existing results using CNNs (0.9 F1-score). Additionally, due to the semantic properties of the deep learning approach, an end-to-end deep learning framework is viable to perform both tasks: detection and classification of mitosis. The results show the potential of deep learning in the analysis of Whole Slide Images (WSI) and its integration to computer-aided systems. The extension of this work to whole slide images is also addressed in the last sections; as well as, some computational key points that are useful when constructing a computer-aided-system inspired by the proposed technology.