Medical Visual Question Answering (MedVQA) is a crucial intersection of artificial intelligence and healthcare. It enables systems to interpret medical images—such as X-rays, MRIs, and pathology slides—and respond to clinical queries. Early approaches primarily relied on discriminative models, which select answers from predefined candidates. However, these methods struggle to effectively address open-ended, domain-specific, or complex queries. Recent advancements have shifted the focus toward generative models, leveraging autoregressive decoders, large language models (LLMs), and multimodal large language models (MLLMs) to generate more nuanced and free-form answers. This review comprehensively examines the paradigm shift from discriminative to generative systems, examining generative MedVQA works on their model architectures and training process, summarizing evaluation benchmarks and metrics, highlighting key advances and techniques that propels the development of generative MedVQA, such as concept alignment, instruction tuning, and parameter-efficient fine-tuning (PEFT), alongside strategies for data augmentation and automated dataset creation. Finally, we propose future directions to enhance clinical reasoning and intepretability, build robust evaluation benchmarks and metrics, and employ scalable training strategies and deployment solutions. By analyzing the strengths and limitations of existing generative MedVQA approaches, we aim to provide valuable insights for researchers and practitioners working in this domain.

Recent advancements in large language models (LLMs) have significantly enhanced text generation across various sectors; however, their medical application faces critical challenges regarding both accuracy and real-time responsiveness. To address these dual challenges, we propose a novel two-step retrieval and ranking retrieval-augmented generation (RAG) framework that synergistically combines embedding search with Elasticsearch technology. Built upon a dynamically updated medical knowledge base incorporating expert-reviewed documents from leading healthcare institutions, our hybrid architecture employs ColBERTv2 for context-aware result ranking while maintaining computational efficiency. Experimental results show a 10% improvement in accuracy for complex medical queries compared to standalone LLM and single-search RAG variants, while acknowledging that latency challenges remain in emergency situations requiring sub-second responses in an experimental setting, which can be achieved in real-time using more powerful hardware in real-world deployments. This work establishes a new paradigm for reliable medical AI assistants that successfully balances accuracy and practical deployment considerations.

Background: Since the emergence of large language models (LLMs), a large number of applications have emerged in vertical fields, such as medical, legal, and subject education. By fine-tuning the pre-trained base model, professional knowledge can be well parameterized into model capabilities, enabling it to have better performance in specific fields. However, we observed that although the fine-tuned model has improved domain-specific knowledge, the performance of medical LLMs (Med-LLMs) in long-context understanding has declined significantly due to the large amount of knowledge-intensive fine-tuning, especially compared with the general language model with similar parameters. This study aims to investigate the problem of the decline in performance of Med-LLMs in long-context understanding. Methods: We designed a series of experiments to conduct open-book professional knowledge tests related to the medical field on models using different fine-tuning methods to evaluate their long-context understanding capabilities in the medical field. These experiments included benchmarks of general language models, benchmarks of medical language models, tests that adjusted the ratio and amount of general data and professional data during fine-tuning, and experimental data to determine the best data composition to optimize professional models and achieve a balance between long-context performance and specific domain knowledge. Results: Our experimental framework evaluated 5 general-purpose LLMs and 6 medical-adapted models through an open-book knowledge assessment protocol. The results revealed a striking performance hierarchy: even the lowest-performing general model (37.52% accuracy) outperformed non-retrained medical baselines (34.65% peak accuracy). However, medical models employing our optimized fine-tuning strategies demonstrated marked accuracy gains, with maximum improvements reaching 13.5 percentage points. Notably, retrained medical specialists like IvyGPT (40.48%) and WiNGPT2 (38.94%) surpassed several general models of larger parameter scales, establishing new performance benchmarks in medical context processing. Our experiments on fine-tuning data volume revealed a critical saturation threshold near 100,000 domain-specific samples. When approaching this boundary, models exhibited instability in contextual understanding and even performance regression, while further fine-tuning beyond this point failed to induce measurable improvements in long-context comprehension capabilities. This suggests an inherent limitation in scaling domain-specific knowledge integration through continued data exposure alone. Conclusions: The composition and quantity of data for model fine-tuning actually affect the model’s ability to understand context in downstream tasks. The balance between the model’s expertise and context understanding depends on the rationality of the fine-tuning data.

Potassium sodium niobate (KNN)-based ceramics are promising alternatives to lead-containing piezoelectric materials. However, the vast design space, characterized by multiple dopant choices and variable content, presents a considerable challenge in the chemical modification of KNN compositions to improve their piezoelectric performance. In recent years, the rapid advance of machine learning (ML) techniques has facilitated expedited materials design and discovery with deeply sought insights into the materials. In this study, we constructed an interpretable and uncertainty-informed ML framework to optimize the piezoelectric coefficient d₃₃ of a KNN-based lead-free system. We identified and analyzed the influential features for the d₃₃ prediction and conducted three experimental iterations based on the uncertainty-informed predictions obtained from the Monte Carlo dropout (MCDropout). Promising KNN compositions exhibiting large d₃₃ values over 300 pC N⁻¹ were located and synthesized. Furthermore, the MCDropout markedly reduced the computational cost by 33% compared to the commonly used bootstrap method for uncertainty assessment. This study exhibits an ML framework with enhanced interpretability and search efficiency for optimizing the crucial piezoelectric properties of piezoceramics. The application scope of the utilized methods can be extended to various materials with tailored properties.