Polar contrast attention and skip cross-channel aggregation for efficient learning in U-Net.

The performance of existing lesion semantic segmentation models has shown a steady improvement with the introduction of mechanisms like attention, skip connections, and deep supervision. However, these advancements often come at the expense of computational requirements, necessitating powerful graphics processing units with substantial video memory. Consequently, certain models may exhibit poor or non-existent performance on more affordable edge devices, such as smartphones and other point-of-care devices. To tackle this challenge, our paper introduces a lesion segmentation model with a low parameter count and minimal operations. This model incorporates polar transformations to simplify images, facilitating faster training and improved performance. We leverage the characteristics of polar images by directing the model's focus to areas most likely to contain segmentation information, achieved through the introduction of a learning-efficient polar-based contrast attention (PCA). This design utilizes Hadamard products to implement a lightweight attention mechanism without significantly increasing model parameters and complexities. Furthermore, we present a novel skip cross-channel aggregation (SC2A) approach for sharing cross-channel corrections, introducing Gaussian depthwise convolution to enhance nonlinearity. Extensive experiments on the ISIC 2018 and Kvasir datasets demonstrate that our model surpasses state-of-the-art models while maintaining only about 25K parameters. Additionally, our proposed model exhibits strong generalization to cross-domain data, as confirmed through experiments on the PH2 dataset and CVC-Polyp dataset. In addition, we evaluate the model's performance in a mobile setting against other lightweight models. Notably, our proposed model outperforms other advanced models in terms of IoU and Dice score, and running time.

Near-Infrared Forearm Vascular Width Calculation Using Radius Estimation of Tangent Circle.

In response to the analysis of the functional status of forearm blood vessels, this paper fully considers the orientation of the vascular skeleton and the geometric characteristics of blood vessels and proposes a blood vessel width calculation algorithm based on the radius estimation of the tangent circle (RETC) in forearm near-infrared images. First, the initial infrared image obtained by the infrared camera is preprocessed by image cropping, contrast stretching, denoising, enhancement, and initial segmentation. Second, the Zhang-Suen refinement algorithm is used to extract the vascular skeleton. Third, the Canny edge detection method is used to perform vascular edge detection. Finally, a RETC algorithm is developed to calculate the vessel width. This paper evaluates the accuracy of the proposed RETC algorithm, and experimental results show that the mean absolute error between the vessel width obtained by our algorithm and the reference vessel width is as low as 0.36, with a variance of only 0.10, which can be significantly reduced compared to traditional calculation measurements.

Temporal Properties of Self-Prioritization.

Using electroencephalogram (EEG), we tested the hypothesis that the association of a neutral stimulus with the self would elicit ultra-fast neural responses from early top-down feedback modulation to late feedforward periods for cognitive processing, resulting in self-prioritization in information processing. In two experiments, participants first learned three associations between personal labels (self, friend, stranger) and geometric shapes (Experiment 1) and three colors (Experiment 2), and then they judged whether the shape/color-label pairings matched. Stimuli in Experiment 2 were shown in a social communicative setting with two avatars facing each other, one aligned with the participant's view (first-person perspective) and the other with a third-person perspective. The color was present on the t-shirt of one avatar. This setup allowed for an examination of how social contexts (i.e., perspective taking) affect neural connectivity mediating self-related processing. Functional connectivity analyses in the alpha band (8-12 Hz) revealed that self-other discrimination was mediated by two distinct phases of neural couplings between frontal and occipital regions, involving an early phase of top-down feedback modulation from frontal to occipital areas followed by a later phase of feedforward signaling from occipital to frontal regions. Moreover, while social communicative settings influenced the later feedforward connectivity phase, they did not alter the early feedback coupling. The results indicate that regardless of stimulus type and social context, the early phase of neural connectivity represents an enhanced state of awareness towards self-related stimuli, whereas the later phase of neural connectivity may be associated with cognitive processing of socially meaningful stimuli.

Label-free Medical Image Quality Evaluation by Semantics-aware Contrastive Learning in IoMT.

With the rapid development of the Internet-of-Medical-Things (IoMT) in recent years, it has emerged as a promising solution to alleviate the workload of medical staff, particularly in the field of Medical Image Quality Assessment (MIQA). By deploying MIQA based on IoMT, it proves to be highly valuable in assisting the diagnosis and treatment of various types of medical images, such as fundus images, ultrasound images, and dermoscopic images. However, traditional MIQA models necessitate a substantial number of labeled medical images to be effective, which poses a challenge in acquiring a sufficient training dataset. To address this issue, we present a label-free MIQA model developed through a zero-shot learning approach. This paper introduces a Semantics-Aware Contrastive Learning (SCL) model that can effectively generalise quality assessment to diverse medical image types. The proposed method integrates features extracted from zero-shot learning, the spatial domain, and the frequency domain. Zero-shot learning is achieved through a tailored Contrastive Language-Image Pre-training (CLIP) model. Natural Scene Statistics (NSS) and patch-based features are extracted in the spatial domain, while frequency features are hierarchically extracted from both local and global levels. All of this information is utilised to derive a final quality score for a medical image. To ensure a comprehensive evaluation, we not only utilise two existing datasets, EyeQ and LiverQ, but also create a dataset specifically for skin image quality assessment. As a result, our SCL method undergoes extensive evaluation using all three medical image quality datasets, demonstrating its superiority over advanced models.

Compound Scaling Encoder-Decoder (CoSED) Network for Diabetic Retinopathy Related Bio-marker Detection.

Biomedical image segmentation plays an important role in Diabetic Retinopathy (DR)-related biomarker detection. DR is an ocular disease that affects the retina in people with diabetes and could lead to visual impairment if management measures are not taken in a timely manner. In DR screening programs, the presence and severity of DR are identified and classified based on various microvascular lesions detected by qualified ophthalmic screeners. Such a detection process is time-consuming and error-prone, given the small size of the microvascular lesions and the volume of images, especially with the increasing prevalence of diabetes. Automated image processing using deep learning methods is recognized as a promising approach to support diabetic retinopathy screening. In this paper, we propose a novel compound scaling encoder-decoder network architecture to improve the accuracy and running efficiency of microvascular lesion segmentation. In the encoder phase, we develop a lightweight encoder to speed up the training process, where the encoder network is scaled up in depth, width, and resolution dimensions. In the decoder phase, an attention mechanism is introduced to yield higher accuracy. Specifically, we employ Concurrent Spatial and Channel Squeeze and Channel Excitation (scSE) blocks to fully utilise both spatial and channel-wise information. Additionally, a compound loss function is incorporated with transfer learning to handle the problem of imbalanced data and further improve performance. To assess performance, our method is evaluated on two large-scale lesion segmentation datasets: DDR and FGADR datasets. Experimental results demonstrate the superiority of our method compared to other competent methods. Our codes are available at https://github.com/DeweiYi/CoSED-Net.

Neural Connectivity Underlying Reward and Emotion-Related Processing: Evidence From a Large-Scale Network Analysis.

Neuroimaging techniques have advanced our knowledge about neurobiological mechanisms of reward and emotion processing. It remains unclear whether reward and emotion-related processing share the same neural connection topology and how intrinsic brain functional connectivity organization changes to support emotion- and reward-related prioritized effects in decision-making. The present study addressed these challenges using a large-scale neural network analysis approach. We applied this approach to two independent functional magnetic resonance imaging datasets, where participants performed a reward value or emotion associative matching task with tight control over experimental conditions. The results revealed that interaction between the Default Mode Network, Frontoparietal, Dorsal Attention, and Salience networks engaged distinct topological structures to support the effects of reward, positive and negative emotion processing. Detailed insights into the properties of these connections are important for understanding in detail how the brain responds in the presence of emotion and reward related stimuli. We discuss the linking of reward- and emotion-related processing to emotional regulation, an important aspect of regulation of human behavior in relation to mental health.

Fabrication of patterned calcium carbonate materials through template-assisted microbially induced calcium carbonate precipitation.

Patterned calcium carbonate materials with controlled morphologies have broad applications in both environmental and engineering fields. However, how to fabricate such materials through environmental-friendly methods under ambient conditions is still challenging. Here, we report a green approach for fabricating patterned calcium carbonate materials. This eco-friendly approach is based on template-assisted microbially induced calcium carbonate precipitation. As a proof of concept, by varying the templates and optimizing fabrication parameters, different patterned calcium carbonate materials were obtained. The optimized parameters include C Ca2+ = 80 mM, T i = 15 °C, and templates made of small-sized CaCO3 particles with a concentration of 1.5 mg mL-1, under which better and more sharp patterns were obtained. Materials with periodic patterns were also fabricated through a periodic template, showing good scalability of this approach. The results of this study could mean great potential in applications where spatially controlled calcium carbonate depositions with user-designed patterns are needed.

Dimension Reduction Aided Hyperspectral Image Classification with a Small-sized Training Dataset: Experimental Comparisons.

Hyperspectral images (HSI) provide rich information which may not be captured by other sensing technologies and therefore gradually find a wide range of applications. However, they also generate a large amount of irrelevant or redundant data for a specific task. This causes a number of issues including significantly increased computation time, complexity and scale of prediction models mapping the data to semantics (e.g., classification), and the need of a large amount of labelled data for training. Particularly, it is generally difficult and expensive for experts to acquire sufficient training samples in many applications. This paper addresses these issues by exploring a number of classical dimension reduction algorithms in machine learning communities for HSI classification. To reduce the size of training dataset, feature selection (e.g., mutual information, minimal redundancy maximal relevance) and feature extraction (e.g., Principal Component Analysis (PCA), Kernel PCA) are adopted to augment a baseline classification method, Support Vector Machine (SVM). The proposed algorithms are evaluated using a real HSI dataset. It is shown that PCA yields the most promising performance in reducing the number of features or spectral bands. It is observed that while significantly reducing the computational complexity, the proposed method can achieve better classification results over the classic SVM on a small training dataset, which makes it suitable for real-time applications or when only limited training data are available. Furthermore, it can also achieve performances similar to the classic SVM on large datasets but with much less computing time.