Shaofu Lin

Multimorbidity risk identification in elderly health consultation scenarios is challenging because chronic disease history and medication exposure may interact through complex causal pathways. Existing LLM-, RAG-, and graph-based retrieval methods often rely on semantic relevance or graph connectivity, which may be insufficient for identifying user-specific risk mechanisms. This study proposes RACPR, a generative AI-enhanced risk-aware causal path re-ranking framework for interpretable multimorbidity risk identification. RACPR ranks candidate causal paths by integrating user-entity alignment, causal coherence, risk contribution, and path length control and uses generative AI to transform selected paths into readable, path-grounded explanations. To support controlled algorithmic evaluation, we constructed a normalized benchmark of 1002 elderly multimorbidity consultation cases covering diabetes, hypertension, and chronic kidney disease. The benchmark and supporting knowledge graph were derived from publicly available biomedical and health information resources, including PubMed abstracts, guideline and review sources, DrugBank medication-safety evidence, and MedlinePlus-based terminology. Under a leakage-controlled setting, only age, diagnosed diseases, and medication exposures were used as model-accessible inputs, while abnormal indicators, support paths, and rationales were reserved for evaluation. On the test set, RACPR achieved an Accuracy of 0.659, a Precision of 0.602, a Recall of 0.938, an F1-score of 0.733, and an AUC of 0.777. Ablation analysis showed that removing the risk-aware component reduced AUC from 0.777 to 0.428. Benchmark-level explanation evaluation further showed improvements in path consistency, path hit rate, and health-oriented plausibility. These findings indicate that risk-aware causal path re-ranking improved risk ranking and explanation grounding relative to the evaluated baselines under the controlled benchmark setting.

In recent years, suicide rates among Chinese adolescents have risen markedly. Moreover, the onset of suicidal behaviour has shifted toward increasingly younger age groups. As a result, expressions of suicidal ideation on social media have become an important public health concern. Although the deep integration of natural language processing and large language models offers new opportunities for adolescent mental-health detection, significant technical challenges remain in terms of model generalizability and detection accuracy. To address these challenges, we propose SentinelGPT, a large language model designed to detect suicidal and depressive tendencies in adolescents. First, we construct a Chinese social-media corpus for suicide and depression detection by performing multi-dimensional cleaning, screening, and preprocessing on original data collected from the Micro-blog “Tree Hole”. Second, we adopt a collaborative GraphRAG architecture that combines a knowledge graph with a vector database to enhance the model’s ability to integrate information across multiple dialogue turns and capture semantic relationships in complex contexts. Third, we fine-tune a dialogue-enhanced LLaMA 3.1 model using LoRA and deploy it privately via the Ollama framework to improve parsing of fragmented expressions, metaphorical language, and cross-platform contextual variations commonly found in online text. To further verify model effectiveness, we conduct comparative experiments against four representative baselines DeepSeek V3, Qwen 2.5, Fine-tuned Qwen 2.5 and ELECTRA using the same dataset and evaluation metrics including precision, recall, and F1-score. SentinelGPT achieves the highest overall F1-scores, outperforming all baselines and demonstrating superior robustness and generalization ability. The results demonstrate that SentinelGPT effectively detects suicidal and depressive tendencies in the complex context of Micro-blog comments, maintains semantic consistency across multi-turn conversations, and offers a reliable technical solution for adolescent mental-health monitoring and early warning.

Accurate kidney tumor segmentation from abdominal CT is essential for quantitative assessment and treatment planning. However, indistinct tumor boundaries and substantial inter-patient shape variability render traditional hand-crafted feature-based methods unreliable for precise delineation. Although deep learning has advanced this task, these methods still struggle with multi-scale tumor characteristics, complex morphological variations, and background noise in medical images. To address these challenges, we propose MDA-Net, an end-to-end segmentation method based on enhanced multi-scale feature extraction and attention refinement. Specifically, we introduce a Multi-Scale Feature Extraction (MSFE) module into encoder–decoder skip connections to aggregate dilated features across multiple receptive fields and learn branch-wise weights for adaptive refinement and fusion, thereby enhancing boundary details and semantic cues to reduce tumor-tissue ambiguity. At the bottleneck, a Deformable Pyramid Feature Refinement (DPFR) module combines deformable sampling with pyramid contextual modeling, thereby improving adaptability to variations in tumor shape and scale while preserving feature resolution. Moreover, a Channel and Spatial Attention (CASA) module is embedded in the decoder to suppress background interference and enhance boundary-sensitive structures during upsampling via coordinated channel and spatial reweighting, thereby improving the reconstruction of fine-grained tumor morphology and contours. Experiments on both KiTS19 and KiTS21 show that MDA-Net consistently improves tumor boundary delineation, lesion localization, and mask reconstruction, demonstrating stronger robustness and cross-dataset generalizability than representative baseline methods. Ablation studies further confirm the complementary effects of MSFE, DPFR, and CASA. In addition, Grad-CAM visualizations improve the clinical transparency and interpretability of the model. Overall, this method advances deep learning for medical image analysis and supports precise diagnosis and treatment of renal tumors.

With the advancement of UAV (Unmanned Aerial Vehicle) technology, flying ad-hoc networks (FANETs), composed of multiple coordinating UAVs, demonstrate tremendous application potential in disaster relief, environmental monitoring and intelligent logistics. However, inherent resource constraints and unpredictable operating environments make UAV failures a frequent and critical challenge. Particularly in mission-critical applications, simultaneous or consecutive failures of multiple UAVs can severely disrupt network topology, leading to catastrophic consequences such as network fragmentation and service interruptions. Furthermore, traditional topology reconstruction algorithms suffer from high computational overhead and significant communication delays. Primarily designed for single-node failure recovery, they are ill-equipped to address the challenge of concurrent multi-node failures. To address these challenges, this paper proposes a topology reconstruction algorithm tailored for multi-node failure scenarios in FANETs. The core objective of this algorithm is to minimize communication overhead and secondary damage to the network during the reconstruction process while ensuring basic reconstruction results, thereby improving the system’s energy efficiency and robustness. The proposed framework integrates three key phases: First, overlapping communication coverage areas among neighbors of failed nodes are leveraged to define first and second regions, enabling rapid identification of connection restoration candidate positions and avoiding computationally intensive global calculations. Second, a comprehensive importance evaluation mechanism is constructed based on the topological and functional attributes of node, categorizing nodes into different importance types. For failed nodes of varying importance, differentiated search ranges and retry strategies are employed to ensure the most suitable nodes are selected for reconstruction tasks. Third, the inflexibility of repulsion ranges in traditional artificial potential field (APF) method is addressed by introducing dynamic repulsion influence zones and a composite repulsion model. The improved APF algorithm enhances safety in high-speed scenarios and reduces the probability of UAVs becoming trapped in local minima. Finally, extensive simulations validate that the proposed algorithm accurately identifies critical network nodes and promptly implements effective reconstruction measures to minimize network damage.

Elderly patients often experience multimorbidity and long-term polypharmacy, making medication safety a critical challenge in disease management. In China, the concurrent use of Western medicines and proprietary Chinese medicines (PCMs) further complicates this issue, as potential drug interactions are often implicit, increasing risks for physiologically vulnerable older adults. Although large language model-based medical question-answering systems have been widely adopted, they remain prone to unsafe outputs in medication-related contexts. Existing retrieval-augmented generation (RAG) frameworks typically rely on static retrieval strategies, limiting their ability to appropriately allocate retrieval and verification efforts across different question types. This paper proposes FusionGraphRAG, an adaptive RAG framework for geriatric disease management. The framework employs query classification-based routing to distinguish questions by complexity and medication relevance; integrates dual-granularity knowledge alignment to connect fine-grained medical entities with higher-level contextual knowledge across diseases, medications, and lifestyle guidance; and incorporates explicit contradiction detection for high-risk medication scenarios. Experiments on the GeriatricHealthQA dataset (derived from the Huatuo corpus) indicate that FusionGraphRAG achieves a Safety Recall of 71.7%. Comparative analysis demonstrates that the framework improves retrieval accuracy and risk interception capabilities compared to existing graph-enhanced baselines, particularly in identifying implicit pharmacological conflicts. The results indicate that the framework supports more reliable geriatric medical question answering while providing enhanced safety verification for medication-related reasoning.

Wireless Body Area Networks (WBANs) enable real-time data collection for medical monitoring, sports tracking, and environmental sensing, driven by Internet of Things advancements. Their layered architecture supports efficient sensing, aggregation, and analysis, but energy constraints from transmission (over 60% of consumption), idle listening, and dynamic conditions like body motion hinder adoption. Challenges include minimizing energy waste while ensuring data reliability, Quality of Service (QoS), and adaptation to channel variations, alongside algorithm complexity and privacy concerns. This paper reviews recent power control mechanisms in WBANs, encompassing feedback control, dynamic and convex optimization, graph theory-based path optimization, game theory, reinforcement learning, deep reinforcement learning, hybrid frameworks, and emerging architectures such as federated learning and cell-free massive MIMO, adopting a systematic review approach with a focus on healthcare and IoT application scenarios. Achieving energy savings ranging from 6% (simple feedback control) to 50% (hybrid frameworks with emerging architectures), depending on method complexity and application scenario, with prolonged network lifetime and improved reliability while preserving QoS requirements in healthcare and IoT applications.