Privacy-preserving blockchain-based federated learning for brain tumor segmentation.

Improved data sharing between healthcare providers can lead to a higher probability of accurate diagnosis, more effective treatments, and enhanced capabilities of healthcare organizations. One critical area of focus is brain tumor segmentation, a complex task due to the heterogeneous appearance, irregular shape, and variable location of tumors. Accurate segmentation is essential for proper diagnosis and effective treatment planning, yet current techniques often fall short due to these complexities. However, the sensitive nature of health data often prohibits its sharing. Moreover, the healthcare industry faces significant issues, including preserving the privacy of the model and instilling trust in the model. This paper proposes a framework to address these privacy and trust issues by introducing a mechanism for training the global model using federated learning and sharing the encrypted learned parameters via a permissioned blockchain. The blockchain-federated learning algorithm we designed aggregates gradients in the permissioned blockchain to decentralize the global model, while the introduced masking approach retains the privacy of the model parameters. Unlike traditional raw data sharing, this approach enables hospitals or medical research centers to contribute to a globally learned model, thereby enhancing the performance of the central model for all participating medical entities. As a result, the global model can learn about several specific diseases and benefit each contributor with new disease diagnosis tasks, leading to improved treatment options. The proposed algorithm ensures the quality of model data when aggregating the local model, using an asynchronous federated learning procedure to evaluate the shared model's quality. The experimental results demonstrate the efficacy of the proposed scheme for the critical and challenging task of brain tumor segmentation. Specifically, our method achieved a 1.99% improvement in Dice similarity coefficient for enhancing tumors and a 19.08% reduction in Hausdorff distance for whole tumors compared to the baseline methods, highlighting the significant advancement in segmentation performance and reliability.

Hierarchical Threshold Pruning Based on Uniform Response Criterion.

Convolutional neural networks (CNNs) have been successfully applied to various fields. However, CNNs' overparameterization requires more memory and training time, making it unsuitable for some resource-constrained devices. To address this issue, filter pruning as one of the most efficient ways was proposed. In this article, we propose a feature-discrimination-based filter importance criterion, uniform response criterion (URC), as a key component of filter pruning. It converts the maximum activation responses into probabilities and then measures the importance of the filter through the distribution of these probabilities over classes. However, applying URC directly to global threshold pruning may cause some problems. The first problem is that some layers will be completely pruned under global pruning settings. The second problem is that global threshold pruning neglects that filters in different layers have different importance. To address these issues, we propose hierarchical threshold pruning (HTP) with URC. It performs a pruning step limited in a relatively redundant layer rather than comparing the filters' importance across all layers, which can avoid some important filters being pruned. The effectiveness of our method benefits from three techniques: 1) measuring filter importance by URC; 2) normalizing filter scores; and 3) conducting prune in relatively redundant layers. Extensive experiments on CIFAR-10/100 and ImageNet show that our method achieves the state-of-the-art performance on multiple benchmarks.

COVID-19 CT ground-glass opacity segmentation based on attention mechanism threshold.

The ground glass opacity (GGO) of the lung is one of the essential features of COVID-19. The GGO in computed tomography (CT) images has various features and low-intensity contrast between the GGO and edge structures. These problems pose significant challenges for segmenting the GGO. To tackle these problems, we propose a new threshold method for accurate segmentation of GGO. Specifically, we offer a framework for adjusting the threshold parameters according to the image contrast. Three functions include Attention mechanism threshold, Contour equalization, and Lung segmentation (ACL). The lung is divided into three areas using the attention mechanism threshold. Further, the segmentation parameters of the attention mechanism thresholds of the three parts are adaptively adjusted according to the image contrast. Only the segmentation regions restricted by the lung segmentation results are retained. Extensive experiments on four COVID datasets show that ACL can segment GGO images at low contrast well. Compared with the state-of-the-art methods, the similarity Dice of the ACL segmentation results is improved by 8.9%, the average symmetry surface distance ASD is reduced by 23%, and the required computational power F L O P s are only 0.09% of those of deep learning models. For GGO segmentation, ACL is more lightweight, and the accuracy is higher. Code will be released at https://github.com/Lqs-github/ACL.

Blockchain and homomorphic encryption based privacy-preserving model aggregation for medical images.

Medical healthcare centers are envisioned as a promising paradigm to handle the massive volume of data for COVID-19 patients using artificial intelligence (AI). Traditionally, AI techniques require centralized data collection and training models within a single organization. This practice can be considered a weakness as it leads to several privacy and security concerns related to raw data communication. To overcome this weakness and secure raw data communication, we propose a blockchain-based federated learning framework that provides a solution for collaborative data training. The proposed framework enables the coordination of multiple hospitals to train and share encrypted federated models while preserving data privacy. Blockchain ledger technology provides decentralization of federated learning models without relying on a central server. Moreover, the proposed homomorphic encryption scheme encrypts and decrypts the gradients of the model to preserve privacy. More precisely, the proposed framework: (i) train the local model by a novel capsule network for segmentation and classification of COVID-19 images, (ii) furthermore, we use the homomorphic encryption scheme to secure the local model that encrypts and decrypts the gradients, (iii) finally, the model is shared over a decentralized platform through the proposed blockchain-based federated learning algorithm. The integration of blockchain and federated learning leads to a new paradigm for medical image data sharing over the decentralized network. To validate our proposed model, we conducted comprehensive experiments and the results demonstrate the superior performance of the proposed scheme.

Fuzzy Graph Subspace Convolutional Network.

Graph convolutional networks (GCNs) are a popular approach to learn the feature embedding of graph-structured data, which has shown to be highly effective as well as efficient in performing node classification in an inductive way. However, with massive nongraph-organized data existing in application scenarios nowadays, it is critical to exploit the relationships behind the given groups of data, which makes better use of GCN and broadens the application field. In this article, we propose the fuzzy graph subspace convolutional network (FGSCN) to provide a brand-new paradigm for feature embedding and node classification with graph convolution (GC) when given an arbitrary collection of data. The FGSCN performs GC on the fuzzy subspace ( F -space), which simultaneously learns from the underlying subspace information in the low-dimensional space as well as its neighborliness information in the high-dimensional space. In particular, we construct the fuzzy homogenous graph GF on the F -space by fusing the homogenous graph of neighborliness GN and homogenous graph of subspace GS (defined by the affinity matrix of the low-rank representation). Here, it is proven that the GC on F -space will propagate both the local and global information through fuzzy set theory. We evaluated FGSCN on 15 unique datasets with different tasks (e.g., feature embedding, visual recognition, etc.). The experimental results showed that the proposed FGSCN has significant superiority compared with current state-of-the-art methods.

Multifunctional Fiber-Enabled Intelligent Health Agents.

The application of wearable devices is promoting the development toward digitization and intelligence in the field of health. However, the current smart devices centered on human health have disadvantages such as weak perception, high interference degree, and unfriendly interaction. Here, an intelligent health agent based on multifunctional fibers, with the characteristics of autonomy, activeness, intelligence, and perceptibility enabling health services, is proposed. According to the requirements for healthcare in the medical field and daily life, four major aspects driven by intelligent agents, including health monitoring, therapy, protection, and minimally invasive surgery, are summarized from the perspectives of materials science, medicine, and computer science. The function of intelligent health agents is realized through multifunctional fibers as sensing units and artificial intelligence technology as a cognitive engine. The structure, characteristics, and performance of fibers and analysis systems and algorithms are reviewed, while discussing future challenges and opportunities in healthcare and medicine. Finally, based on the above four aspects, future scenarios related to health protection of a person's life are presented. Intelligent health agents will have the potential to accelerate the realization of precision medicine and active health.

Regularization on Augmented Data to Diversify Sparse Representation for Robust Image Classification.

Image classification is a fundamental component in modern computer vision systems, where sparse representation-based classification has drawn a lot of attention due to its robustness. However, on the optimization of sparse learning systems, regularization and data augmentation are both powerful, but currently isolated. We believe that regularization and data augmentation can cooperate to generate a breakthrough in robust image classification. In this article, we propose a novel framework, regularization on augmented data (READ), which creates diversification in the data using the generic augmentation techniques to implement robust sparse representation-based image classification. When the training data are augmented, READ applies a distinct regularizer, l1 or l2 , in particular, on the augmented training data apart from the original data, so that regularization and data augmentation are utilized and enhanced synchronously. We introduce an elaborate theoretical analysis on how to optimize the sparse representation by both l1 -norm and l2 -norm with the generic data augmentation and demonstrate its performance in extensive experiments. The results obtained on several facial and object datasets show that READ outperforms many state-of-the-art methods when using deep features.

Consensus Sparsity: Multi-context Sparse Image Representation via L∞-induced Matrix Variate.

The sparsity is an attractive property that has been widely and intensively utilized in various image processing fields (e.g., robust image representation, image compression, image analysis, etc.). Its actual success owes to the exhaustive mining of the intrinsic (or homogenous) information from the whole data carrying redundant information. From the perspective of image representation, the sparsity can successfully find an underlying homogenous subspace from a collection of training data to represent a given test sample. The famous sparse representation (SR) and its variants embed the sparsity by representing the test sample using a linear combination of training samples with L0-norm regularization and L1-norm regularization. However, although these state-of-the-art methods achieve powerful and robust performances, the sparsity is not fully exploited on the image representation in the following three aspects: 1) the within-sample sparsity, 2) the between-sample sparsity, and 3) the image structural sparsity. In this paper, to make the above-mentioned multi-context sparsity properties agree and simultaneously learned in one model, we propose the concept of consensus sparsity (Con-sparsity) and correspondingly build a multi-context sparse image representation (MCSIR) framework to realize this. We theoretically prove that the consensus sparsity can be achieved by the L∞-induced matrix variate based on the Bayesian inference. Extensive experiments and comparisons with the state-of-the-art methods (including deep learning) are performed to demonstrate the promising performance and property of the proposed consensus sparsity.

Hierarchical-morphology metafabric for scalable passive daytime radiative cooling.

Incorporating passive radiative cooling structures into personal thermal management technologies could effectively defend humans against intensifying global climate change. We show that large-scale woven metafabrics can provide high emissivity (94.5%) in the atmospheric window and high reflectivity (92.4%) in the solar spectrum because of the hierarchical-morphology design of the randomly dispersed scatterers throughout the metafabric. Through scalable industrial textile manufacturing routes, our metafabrics exhibit desirable mechanical strength, waterproofness, and breathability for commercial clothing while maintaining efficient radiative cooling ability. Practical application tests demonstrated that a human body covered by our metafabric could be cooled ~4.8°C lower than one covered by commercial cotton fabric. The cost-effectiveness and high performance of our metafabrics present substantial advantages for intelligent garments, smart textiles, and passive radiative cooling applications.

Flexible and Robust Biomaterial Microstructured Colored Textiles for Personal Thermoregulation.

Integrating personal thermoregulation technologies into wearable textiles has enabled extensive and profound technological breakthroughs in energy savings, thermal comfort, wearable electronics, intelligent fabrics, and so forth. Nevertheless, previous studies have suffered from long-standing issues such as limited working temperature, poor comfort, and weak reliability of the textiles. Here, we demonstrate a skin-friendly personal insulation textile and a thermoregulation textile that can perform both passive heating and cooling using the same piece of textile with zero energy input. The insulation textile material is composed of biomaterial microstructured fibers that exhibit good thermal insulation, low thermal emissivity, and good dyeability. By filling these microstructure fibers with biocompatible phase-change materials and coating them with polydimethylsiloxane, the insulation textile becomes a thermoregulation textile that shows good water hydrophobicity, high mechanical robustness, and high working stability. The proposed thermoregulation textile exhibits slow heating/cooling rates with improved thermal comfort, offering feasible and adaptive options for personal cooling/heating scenarios and enabling scalable manufacturing for practical applications.

A shell dataset, for shell features extraction and recognition.

Shells are very common objects in the world, often used for decorations, collections, academic research, etc. With tens of thousands of species, shells are not easy to identify manually. Until now, no one has proposed the recognition of shells using machine learning techniques. We initially present a shell dataset, containing 7894 shell species with 29622 samples, where totally 59244 shell images for shell features extraction and recognition are used. Three features of shells, namely colour, shape and texture were generated from 134 shell species with 10 samples, which were then validated by two different classifiers: k-nearest neighbours (k-NN) and random forest. Since the development of conchology is mature, we believe this dataset can represent a valuable resource for automatic shell recognition. The extracted features of shells are also useful in developing and optimizing new machine learning techniques. Furthermore, we hope more researchers can present new methods to extract shell features and develop new classifiers based on this dataset, in order to improve the recognition performance of shell species.