Skin cancer is one of the most prevalent forms of cancer worldwide, and early and accurate diagnosis critically impacts patient outcomes. Given the sensitive nature of medical data and its fragmented distribution across institutions (data silos), privacy-preserving collaborative learning is essential to enable knowledge-sharing without compromising patient confidentiality. While federated learning (FL) offers a promising solution, existing methods struggle with heterogeneous and missing modalities across institutions, which reduce the diagnostic accuracy. To address these challenges, we propose an effective and flexible Personalized Multimodal Federated Learning framework (PMM-FL), which enables efficient cross-client knowledge transfer while maintaining personalized performance under heterogeneous and incomplete modality conditions. Our study contains three key contributions: (1) A hierarchical aggregation strategy that decouples multi-module aggregation from local deployment via global modular-separated aggregation and local client fine-tuning. Unlike conventional FL (which synchronizes all parameters in each round), our method adopts a frequency-adaptive synchronization mechanism, updating parameters based on their stability and functional roles. (2) A multimodal fusion approach based on multitask learning, integrating learnable modality imputation and attention-based feature fusion to handle missing modalities. (3) A custom dataset combining multi-year International Skin Imaging Collaboration(ISIC) challenge data (2018–2024) to ensure comprehensive coverage of diverse skin cancer types. We evaluate PMM-FL through diverse experiment settings, demonstrating its effectiveness in heterogeneous and incomplete modality federated learning settings, achieving 92.32% diagnostic accuracy with only a 2% drop in accuracy under 30% modality missingness, with a 32.9% communication overhead decline compared with baseline FL methods.

With the development of information technology, industry 4.0, mobile applications, etc., the volume of data has shown explosive growth, NoSQL databases are widely used to support data storage and access services. To resolve the write amplification problem for the common storage architecture LSM-Tree, this paper proposes a novel merge scheme, namely LSMA-Tree, by combining the reading technique of leveling scheme and the writing technique of tiering scheme. For each layer of LSM-Tree, this paper maintains the order of the whole key by files like leveling scheme. Furthermore, in the merging stage of the upper and lower files, LSMA-Tree adopts the "append" rather than "reorganization" method to avoid the read and write operations of the lower files, thereby reducing the number of disk IOs. This paper also designs a new sstable file structure to adapt to the proposed LSMA scheme, and implements and verifies it on LevelDB. Experimental results show that the proposed LSMA scheme can effectively reduce write amplification and improve write performance by more than 50%, while the impact on read performance is less than 5%.