Recently, with the rapid development of electron microscopy (EM) technology and the increasing demand of neuron circuit reconstruction, the scale of reconstruction data grows significantly. This brings many challenges, one of which is how to effectively manage large-scale data so that researchers can mine valuable information. For this purpose, we developed a data management module equipped with two parts, a storage and retrieval module on the server-side and an image cache module on the client-side. On the server-side, Hadoop and HBase are introduced to resolve massive data storage and retrieval. The pyramid model is adopted to store electron microscope images, which represent multiresolution data of the image. A block storage method is proposed to store volume segmentation results. We design a spatial location-based retrieval method for fast obtaining images and segments by layers rapidly, which achieves a constant time complexity. On the client-side, a three-level image cache module is designed to reduce latency when acquiring data. Through theoretical analysis and practical tests, our tool shows excellent real-time performance when handling large-scale data. Additionally, the server-side can be used as a backend of other similar software or a public database to manage shared datasets, showing strong scalability.
25 August 2025 in IEEE Transactions on Medical Imaging
Accurate segmentation of neurons in 3D fluorescence microscopy images is essential for advancing neuroscience. Prevalent methods split a volume into patches and process each patch separately due to computational resource limitations. However, they fail to capture global neuronal morphology across multiple patches, which results in discontinuous segmentation and poses a challenge for subsequent neuronal reconstruction. In this paper, we propose a dual U-Net architecture termed “Glancing Beyond Patch” Network (GBP-Net) to incorporate contextual information into segmentation. Specifically, GBP-Net encodes contextual and high-resolution information using two U-Nets, respectively, and facilitates their integration through a cross-scale context module (CSCM) and a cross-resolution fusion module (CRFM). CSCM utilizes a cross-attention mechanism to enable interaction of features from diverse fields of view after encoders, and CRFM employs Mamba to adaptively fuse high-resolution features after decoders. These two modules complement each other at different levels, enabling to capture both global neuronal structures and fine-grained details. Additionally, the proposed cross-network loss guides to focus on challenging samples by penalizing misclassified voxels from both networks, which further promotes the exploitation of contextual information. Experimental results on three datasets demonstrate that our method outperforms other advanced segmentation methods while maintaining computational efficiency. It keeps the global structure of neurons and achieves the highest F1 scores.
27 August 2024 in IEEE Transactions on Neural Networks and Learning Systems
The multicut problem, also known as correlation clustering, is a classic combinatorial optimization problem that aims to optimize graph partitioning given only node (dis)similarities on edges. It serves as an elegant generalization for several graph partitioning problems and has found successful applications in various areas such as data mining and computer vision. However, the multicut problem with an exponentially large number of cycle constraints proves to be NP-hard, and existing solvers either suffer from exponential complexity or often give unsatisfactory solutions due to inflexible heuristics driven by hand-designed mechanisms. In this article, we propose a deep graph reinforcement learning method to solve the multicut problem within a combinatorial decision framework involving sequential edge contractions. The customized subgraph neural network adapts to the dynamically edge-contracted graph environment by extracting bilevel connected features from both contracted and original graphs. Our method can learn to infer feasible multicut solutions end-to-end toward optimization of the multicut objective in a data-driven manner. More specifically, by exploring the decision space adaptively, it implicitly gains heuristic knowledge from topological patterns of instances and thereby generates more targeted heuristics overcoming the short-sightedness inherent in the hand-designed ones. During testing, the learned heuristics iteratively contract graphs to construct high-quality solutions within polynomial time. Extensive experiments on synthetic and real-world multicut instances show the superiority of our method over existing combinatorial solvers, while also maintaining a certain level of out-of-distribution generalization ability.