As a crucial aspect of data mining, high average utility sequential pattern mining (SPM) aims to discover low frequency and high average utility patterns (subsequences) in sequence data. Most existing high average utility SPM methods overlook the repetitive occurrences of patterns in each sequence, resulting in some important patterns being ignored. To address this issue, we focus on the problem of mining high average utility nonoverlapping patterns (HUPs) from sequential database, and propose an HUP-Miner algorithm. To reduce the need for repeated scanning of the original database, we use a position dictionary to record the occurrence information of each item. To reduce the number of candidate patterns generated, we adopt a pattern join strategy and explore four pruning strategies. To efficiently calculate the average utility of a pattern, we propose an SPC algorithm that utilizes the occurrence positions of sub-patterns. When compared with 12 competitive algorithms, the experimental results on 14 databases show that HUP-Miner gives superior results. Furthermore, we use information gain as the utility for each item, and find that the HUPs discovered in this way can generate better performance via a clustering analysis. All of the algorithms and databases used here are available from https://github.com/wuc567/Pattern-Mining/tree/master/HUP-Miner .

Orthodontic path planning is crucial for precise treatment, yet AI -generated trajectories remain inferior to expert designs due to difficulties in perceiving large dental point clouds and modeling multi-step tooth displacements. In this paper, we introduce Efficient ToothAR, a cascade autoregressive framework that decomposes orthodontic planning into tooth perception and transition prediction stages for improved efficiency and accuracy. In the perception stage, we adapt the efficient sequence model Mamba with a tooth-by-tooth scan and hierarchical architecture to encode dental point clouds. In the transition stage, a bidirectional transformer predicts sequential tooth movements autore-gressively from both initial and target configurations, effectively reducing accumulated transition errors. Trained on 5,000 clinical alignment trajectories, Efficient ToothAR achieves state-of-the-art performance across multiple metrics (e.g., displacement error, collision rate) while maintaining high computational efficiency.

Nucleus segmentation plays a vital role in medical image analysis. However, existing segmentation methods frequently encounter hurdles, such as the loss of crucial image details during downsampling and issues like noise and spatial displacement. In this study, we propose NucleiFormer, where Haar wavelet transforms are employed in the encoder to replace conventional downsampling techniques. Additionally, we utilize an adaptive feature calibration module to align and calibrate features of different scales, reduce feature redundancy, suppress noise, and enhance the spatial awareness of the model. Extensive experiments validate our model’s superior segmentation performance in accurately delineating nucleus boundaries and minimizing errors.

Nuclei segmentation is crucial in medical image analysis, but accurately segmenting complex nuclei shapes under weak supervision with only point annotations remains a significant challenge. To address this, we propose an innovative weakly supervised segmentation method, which includes a novel Hybrid Decoder with two skip connections, a Dual-Gated Attention module, and Laplacian operators. These components collaborate to enhance fine-grained feature representation and refine boundary detection. Additionally, we employ a Graph Convolutional Network (GCN) to model spatial dependencies between pixels via a graph structure, integrating the GCN-generated labels into the loss function to further boost segmentation accuracy and optimize nuclei boundaries and structures. Experimental results demonstrate that our method outperforms existing approaches, particularly in challenging tasks with complex nuclei.

Nuclei segmentation is crucial in medical image analysis. Weakly supervised methods based on point annotations alleviate the labor-intensive process of pixel-level labeling. However, existing approaches struggle with closely adjacent nuclei, leading to imprecise supervision signals, prediction noise, and reduced accuracy. To address these issues, we propose WDRE-NET, a novel end-to-end weakly supervised framework. Our framework incorporates a Wavelet-based Differential Convolutional Module (WDCM) to enhance the model’s ability to capture nuclei boundaries through multi-scale, multi-directional feature extraction, improving the differentiation of adjacent nuclei. Additionally, a Region-Expansion mechanism (RE) iteratively extends point annotations to cover more high-confidence areas, generating reliable supervision signals. This reduces prediction noise and missed detections, improving segmentation performance. Extensive experiments on three public datasets show that WDRE-NET outperforms state-of-the-art methods while effectively addressing the challenge of adjacent nuclei.

ABSTRACT Cell nucleus instance segmentation plays a pivotal role in medical research and clinical diagnosis by providing insights into cell morphology, disease diagnosis, and treatment evaluation. Despite significant efforts from researchers in this field, there remains a lack of a comprehensive and systematic review that consolidates the latest advancements and challenges in this area. In this survey, we offer a thorough overview of existing approaches to nucleus instance segmentation, exploring both traditional and deep learning‐based methods. Traditional methods include watershed, thresholding, active contour model, and clustering algorithms, while deep learning methods include one‐stage methods and two‐stage methods. For these methods, we examine their principles, procedural steps, strengths, and limitations, offering guidance on selecting appropriate techniques for different types of data. Furthermore, we comprehensively investigate the formidable challenges encountered in the field, including ethical implications, robustness under varying imaging conditions, computational constraints, and the scarcity of annotated data. Finally, we outline promising future directions for research, such as privacy‐preserving and fair AI systems, domain generalization and adaptation, efficient and lightweight model design, learning from limited annotations, as well as advancing multimodal segmentation models.

Sequential pattern mining (SPM) is an important branch of knowledge discovery that aims to mine frequent sub-sequences (patterns) in a sequential database. Various SPM methods have been investigated, and most of them are classical SPM methods, since these methods only consider whether or not a given pattern occurs within a sequence. Classical SPM can only find the common features of sequences, but it ignores the number of occurrences of the pattern in each sequence, i.e., the degree of interest of specific users. To solve this problem, this paper addresses the issue of repetitive nonoverlapping sequential pattern (RNP) mining and proposes the RNP-Miner algorithm. To reduce the number of candidate patterns, RNP-Miner adopts an itemset pattern join strategy. To improve the efficiency of support calculation, RNP-Miner utilizes the candidate support calculation algorithm based on the position dictionary. To validate the performance of RNP-Miner, 10 competitive algorithms and 20 sequence databases were selected. The experimental results verify that RNP-Miner outperforms the other algorithms, and using RNPs can achieve a better clustering performance than raw data and classical frequent patterns. All the algorithms were developed using the PyCharm environment and can be downloaded fromhttps://github.com/wuc567/Pattern-Mining/tree/master/RNP-Miner.