RESUMEN

Transformer-based methods effectively capture global dependencies in images, demonstrating outstanding performance in multiple visual tasks. However, existing Transformers cannot effectively denoise large noisy images captured under low-light conditions owing to (1) the global self-attention mechanism causing high computational complexity in the spatial dimension owing to a quadratic increase in computation with the number of tokens; (2) the channel-wise self-attention computation unable to optimise the spatial correlations in images. We propose a local-global interaction Transformer (LGIT) that employs an adaptive strategy to select relevant patches for global interaction, achieving low computational complexity in global self-attention computation. A top-N patch cross-attention model (TPCA) is designed based on superpixel segmentation guidance. TPCA selects top-N patches most similar to the target image patch and applies cross attention to aggregate information from them into the target patch, effectively enhancing the utilisation of the image's nonlocal self-similarity. A mixed-scale dual-gated feedforward network (MDGFF) is introduced for the effective extraction of multiscale local correlations. TPCA and MDGFF were combined to construct a hierarchical encoder-decoder network, LGIT, to compute self-attention within and across patches at different scales. Extensive experiments using real-world image-denoising datasets demonstrated that LGIT outperformed state-of-the-art (SOTA) convolutional neural network (CNN) and Transformer-based methods in qualitative and quantitative results.

RESUMEN

Compared to computed tomography (CT), magnetic resonance imaging (MRI) is more sensitive to acute ischemic stroke lesion. However, MRI is time-consuming, expensive, and susceptible to interference from metal implants. Generating MRI images from CT images can address the limitations of MRI. The key problem in the process is obtaining lesion information from CT. In this study, we propose a cross-modal image generation algorithm from CT to MRI for acute ischemic stroke by combining radiomics with generative adversarial networks. First, the lesion candidate region was obtained using radiomics, the radiomic features of the region were extracted, and the feature with the largest information gain was selected and visualized as a feature map. Then, the concatenation of the extracted feature map and the CT image was input in the generator. We added a residual module after the downsampling of the generator, following the general shape of U-Net, which can deepen the network without causing degradation problems. In addition, we introduced the lesion feature similarity loss function to focus the model on the similarity of the lesion. Through the subjective judgment of two experienced radiologists and using evaluation metrics, the results showed that the generated MRI images were very similar to the real MRI images. Moreover, the locations of the lesions were correct, and the shapes of lesions were similar to those of the real lesions, which can help doctors with timely diagnosis and treatment.

Asunto(s)

RESUMEN

OBJECTIVE: This work aimed to develop a radiomics nomogram to predict 3-year overall survival of esophageal cancer patients after chemoradiotherapy. METHODS: A total of 109 esophageal cancer patients, diagnosed from November 2012 to February 2015, were enrolled in this retrospective study. They were randomly divided into training set (77 cases) and verification set (32 cases). Image standardization was performed prior to feature extraction. And then, about 1670 radiomics features were extracted from the pretreatment diagnostic computed tomography image. A radiomics signature was constructed with the lasso algorithm; then, a radiomics score was calculated to reflect survival probability using the radiomics signature for each patient. A radiomics nomogram was developed by incorporating the radiomics score and clinical factors. A clinical model was constructed using clinical factors only. The performance of the nomogram was assessed with respect to its calibration and discrimination. Kaplan-Meier survival analysis was performed. RESULTS: Sixteen radiomics features were selected to build the radiomics signature. The radiomics nomogram showed better calibration and classification capacity than the clinical model with AUC 0.96 vs. 0.72 for the training cohort, and 0.87 vs. 0.67 for the validation cohort. The model showed good discrimination with a Harrell's Concordance Index of 0.76 in the training cohort and 0.81 in the validation cohort. Decision curve analysis demonstrated the clinical usefulness of the radiomics nomogram. A significant difference (p value < 0.05; log-rank test) was observed between the survival curves of the nomogram-predicted survival and non-survival groups. CONCLUSIONS: The present study proposed a radiomics-based nomogram involving the radiomics signature and clinical factors. It can be potentially applied in the individual preoperative prediction of 3-year survival in esophageal cancer patients.

Asunto(s)

RESUMEN

OBJECTIVES: To evaluate the utility of radiomics features in differentiating central lung cancers and atelectasis on contrast-enhanced computed tomography (CT) images. This study is retrospective. MATERIALS AND METHODS: In this study, 36 patients with central pulmonary cancer and atelectasis between July 2013 and June 2018 were identified. A total of 1,653 2D and 2,327 3D radiomics features were extracted from segmented lung cancers and atelectasis on contrast-enhanced CT. The refined features were investigated for usefulness in classifying lung cancer and atelectasis according to the information gain, and 10 models were trained based on these features. The classification model is trained and tested at the region level and pixel level, respectively. RESULTS: Among all the extracted features, 334 2D features and 1,507 3D features had an information gain (IG) greater than 0.1. The highest accuracy (AC) of the region classifiers was 0.9375. The best Dice score, Hausdorff distance, and voxel AC were 0.2076, 45.28, and 0.8675, respectively. CONCLUSIONS: Radiomics features derived from contrast-enhanced CT images can differentiate lung cancers and atelectasis at the regional and voxel levels.

Asunto(s)

RESUMEN

This study aims at analyzing the separability of acute cerebral infarction lesions which were invisible in CT. 38 patients, who were diagnosed with acute cerebral infarction and performed both CT and MRI, and 18 patients, who had no positive finding in either CT or MRI, were enrolled. Comparative studies were performed on lesion and symmetrical regions, normal brain and symmetrical regions, lesion, and normal brain regions. MRI was reconstructed and affine transformed to obtain accurate lesion position of CT. Radiomic features and information gain were introduced to capture efficient features. Finally, 10 classifiers were established with selected features to evaluate the effectiveness of analysis. 1301 radiomic features were extracted from candidate regions after registration. For lesion and their symmetrical regions, there were 280 features with information gain greater than 0.1 and 2 features with information gain greater than 0.3. The average classification accuracy was 0.6467, and the best classification accuracy was 0.7748. For normal brain and their symmetrical regions, there were 176 features with information gain greater than 0.1, 1 feature with information gain greater than 0.2. The average classification accuracy was 0.5414, and the best classification accuracy was 0.6782. For normal brain and lesions, there were 501 features with information gain greater than 0.1 and 1 feature with information gain greater than 0.5. The average classification accuracy was 0.7480, and the best classification accuracy was 0.8694. In conclusion, the study captured significant features correlated with acute cerebral infarction and confirmed the separability of acute lesions in CT, which established foundation for further artificial intelligence-assisted CT diagnosis.

Asunto(s)

RESUMEN

Ultrasonography is one of the main imaging methods for diagnosing thyroid nodules. Automatic differentiation between benign and malignant nodules in ultrasound images can greatly assist inexperienced clinicians in their diagnosis. The key of problem is the effective utilization of the features of ultrasound images. In this study, we propose a method that is based on the combination of conventional ultrasound and ultrasound elasticity images based on a convolutional neural network and introduces richer feature information for the classification of benign and malignant thyroid nodules. First, the conventional network model performs pretraining on ImageNet and transfers the feature parameters to the ultrasound image domain by transfer learning so that depth features may be extracted and small samples may be processed. Then, we combine the depth features of conventional ultrasound and ultrasound elasticity images to form a hybrid feature space. Finally, the classification is completed on the hybrid feature space, and an end-to-end CNN model is implemented. The experimental results demonstrate that the accuracy of the proposed method is 0.9470, which is better than that of other single data-source methods under the same conditions.

Asunto(s)

RESUMEN

In this paper, we report our practical experience in designing and implementing a platform with Hadoop/MapReduce framework for supporting health Big Data Analytics. Three billion of emulated health raw data was constructed and cross-referenced with data profiles and metadata based on existing health data at the Island Health Authority, BC, Canada. The patient data was stored over a Hadoop Distributed File System to simulate a presentation of an entire health authority's information system. Then, a High Performance Computing platform called WestGrid was used to benchmark the performance of the platform via several data query tests. The work is important as very few implementation studies existed that tested a BDA platform applied to patient data of a health authority system.

Asunto(s)