RadLite：用于CPU部署的放射学AI的多任务LoRA微调小型语言模型
RadLite: Multi-Task LoRA Fine-Tuning of Small Language Models for CPU-Deployable Radiology AI
AI总结：
研究通过LoRA微调小型语言模型，实现放射学多任务性能，展示在消费级CPU上部署的可行性，提出RadLite模型。
摘要：
大型语言模型（LLMs）在放射学领域显示出潜力，但其部署受限于计算需求，无法在资源受限的临床环境中使用。我们研究是否可以通过LoRA微调使3-40亿参数的小型语言模型（SLMs）在放射学任务中表现优异，从而在消费级CPU上部署。我们训练了Qwen2.5-3B-Instruct和Qwen3-4B模型，使用162,000个样本覆盖9个放射学任务，包括10种系统的RADS分类、印象生成、时间比较、放射学NLI、NER、异常检测、N/M分期和放射学问答，这些数据来自12个公开数据集。两种模型在每个任务上最多评估500个保留测试样本，使用标准化指标。我们的关键发现是：（1）LoRA微调显著提高了零样本基线的性能（RADS准确率提高53%，NLI提高60%，N-staging提高89%）；（2）两种模型表现出互补的优势——Qwen2.5在结构生成任务中表现优异，而Qwen3在提取任务中占主导地位；（3）结合两种模型的任务外Oracle集成在所有任务中表现最佳；（4）使用微调模型的少样本提示会损害性能，证明LoRA适应比上下文学习在专用领域更有效；（5）模型可以量化为GGUF格式（约1.8-2.4GB）以在消费级硬件上部署，速度为每秒4-8个token。我们的工作证明，经过高效微调的小型模型——我们统称为RadLite——可以作为实用的多任务放射学AI助手，完全在消费级硬件上部署，无需GPU要求。
英文摘要：
Large language models (LLMs) show promise in radiology but their deployment is limited by computational requirements that preclude use in resource-constrained clinical environments. We investigate whether small language models (SLMs) of 3-4 billion parameters can achieve strong multi-task radiology performance through LoRA fine-tuning, enabling deployment on consumer-grade CPUs. We train Qwen2.5-3B-Instruct and Qwen3-4B on 162K samples spanning 9 radiology tasks - RADS classification across 10 systems, impression generation, temporal comparison, radiology NLI, NER, abnormality detection, N/M staging, and radiology Q&A - compiled from 12 public datasets. Both models are evaluated on up to 500 held-out test samples per task with standardized metrics. Our key findings are: (1) LoRA fine-tuning dramatically improves performance over zero-shot baselines (RADS accuracy +53%, NLI +60%, N-staging +89%); (2) the two models exhibit complementary strengths - Qwen2.5 excels at structured generation tasks while Qwen3 dominates extractive tasks; (3) a task-outed oracle ensemble combining both models achieves the best performance across all tasks; (4) few-shot prompting with fine-tuned models hurts performance, demonstrating that LoRA adaptation is more effective than in-context learning for specialized domains; and (5) models can be quantized to GGUF format (~1.8-2.4GB) for CPU deployment at 4-8 tokens/second on consumer hardware. Our work demonstrates that small, efficiently fine-tuned models - which we collectively call RadLite - can serve as practical multi-task radiology AI assistants deployable entirely on consumer hardware without GPU requirements.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
