跳转至

LUCAID：用于肺癌精准病理学的智能体多模态人工智能

文章背景与核心概要

肺癌组织诊断是精准肿瘤学中的复杂环节，治疗决策高度依赖于组织形态学、免疫组化及分子特征的综合分析。然而，传统的病理评估仍以视觉观察和半定量分析为主，存在显著的观察者间差异。此外，现有的AI工具大多仅针对单一任务，缺乏通用性，且极少经过前瞻性临床验证。

为了解决这些痛点，研究团队开发了LUCAID系统。这是一个专为肺癌精准病理学设计的智能体多模态AI系统。该系统通过一个整合型智能体，将诊断推理与九个功能模块相结合，涵盖了从质量控制到结构化报告生成的全流程。在临床验证中，LUCAID展现出了卓越的性能，其与专家组共识标准的一致性高达93.0%，显著优于经验丰富的胸科病理学家。

执行摘要

Lung cancer tissue diagnostics is a complex domain where therapy decisions depend heavily on integrating histomorphological, immunohistochemical, and molecular features. Traditional pathological assessments remain largely visual, semi-quantitative, and prone to interobserver variability. While prior artificial intelligence (AI) tools have targeted isolated tasks, they have rarely achieved generalizable, expert-level performance or undergone prospective clinical validation.

To overcome these hurdles, researchers developed LUCAID, an agentic multimodal AI system engineered for precision lung cancer pathology. LUCAID features an integrative agent that combines diagnostic reasoning with nine distinct modules, streamlining the entire routine workflow from quality control to structured reporting.

作者列表

Marie-Lisa Eich, Kai Standvoss, Timo Milbich, Alexander Möllers, Miriam Hägele, Philipp Anders, Lars Tharun, Hanna Kontradiuk, Sebastian Kons, Nader Aldoj, Recepcan Adigüzel, Adam Narai, Lukas Hönig, Jonathan Striebel, Binru Yang, Mihnea P. Dragomir, Marvin Sextro, Philipp Keyl, Philipp Jurmeister, Rosemarie Krupar, Evelyn Ramberger, James Wells, Julika Ribbat-Idel, Andreas Kunft, Hussam Shuaib, Christian Grohé, Reinhard Büttner, David Horst, Klaus-Robert Müller, Lukas Ruff, Maximilian Alber, Frederick Klauschen, and Simon Schallenberg

摘要

肺癌组织诊断过程复杂，因为精准肿瘤学的治疗决策依赖于组织形态学、免疫组化和分子特征的整合。然而，病理评估目前仍主要依赖视觉和半定量分析，且存在观察者间差异；现有的AI工具仅覆盖部分任务，难以达到通用的专家级性能，且缺乏前瞻性临床验证。

为了应对这些挑战，我们开发并临床验证了LUCAID，这是一个用于肺癌精准病理学的智能体AI系统。一个整合型智能体将诊断推理与九个模块相结合，涵盖了常规工作流程的全过程，包括质量控制、肿瘤检测与分割、组织学亚型分类、肿瘤微环境分析、肿瘤细胞含量定量、预测性生物标志物评分（PD-L1, MET, TROP-2）以及自动生成结构化报告。LUCAID允许用户交互式查询模块输出，并生成将结果情境化的报告。

针对大规模专家标注的基准，分析模块的F1分数达到了0.82–0.95。在前瞻性临床验证中，LUCAID在临床可操作决策方面与专家组裁定的参考标准达到了93.0%的一致性，而五位经验丰富的胸科病理学家的一致性范围为68.3–81.1%。

Lung cancer tissue diagnostics is complex, as therapy decisions in precision oncology rely on the integration of histomorphological, immunohistochemical, and molecular features. Yet pathological assessment remains largely visual and semi-quantitative and shows interobserver variability, while existing artificial intelligence (AI) tools cover only selected tasks, rarely reach generalizable expert-level performance, and lack prospective clinical validation.

To address these challenges, we developed and clinically validated LUCAID, an agentic AI system for precision lung cancer pathology. An integrative agent couples diagnostic reasoning with nine modules that cover the full routine workflow, from quality control, tumor detection and segmentation, histological subtyping, tumor microenvironment profiling, tumor cellularity quantification, and predictive biomarker scoring (PD-L1, MET, TROP-2) to automated structured report generation. LUCAID enables users to interactively query the module outputs and generate reports that contextualize the results.

Against large-scale expert ground-truth annotations, the analysis modules achieved F1 scores of 0.82–0.95. In prospective clinical validation, LUCAID reached 93.0% concordance with an expert-panel adjudicated reference standard across clinically actionable decisions, compared with 68.3–81.1% for five experienced thoracic pathologists.

关键功能与工作流模块

LUCAID的整合型智能体协调了九个专业模块，覆盖了端到端的病理工作流程：

  1. 质量控制： 确保诊断级图像的完整性。
  2. 肿瘤检测与分割： 准确识别并勾勒癌变组织区域。
  3. 组织学亚型分类： 区分特定的肺癌亚型。
  4. 肿瘤微环境分析： 分析周围的细胞环境。
  5. 肿瘤细胞含量定量： 测量对分子检测至关重要的肿瘤纯度。
  6. 预测性生物标志物评分： 自动评估关键生物标志物，包括：
  7. PD-L1
  8. MET
  9. TROP-2
  10. 自动结构化报告生成： 将分析结果汇编成临床级报告，并提供背景信息，支持用户交互式查询。

LUCAID's integrative agent orchestrates nine specialized modules covering the end-to-end pathology workflow:

  1. Quality Control: Ensures diagnostic-grade image integrity.
  2. Tumor Detection and Segmentation: Accurately localizes and boundaries cancerous tissue regions.
  3. Histological Subtyping: Distinguishes specific lung cancer subtypes.
  4. Tumor Microenvironment Profiling: Analyzes surrounding cellular contexts.
  5. Tumor Cellularity Quantification: Measures tumor purity crucial for molecular testing.
  6. Predictive Biomarker Scoring: Automatically evaluates key biomarkers, including:
  7. PD-L1
  8. MET
  9. TROP-2
  10. Automated Structured Report Generation: Compiles findings into clinical-grade reports that contextualize analysis outputs, allowing interactive user queries.

性能与验证结果

  • 分析性能： 针对大规模专家标注的基准测试，各分析模块表现优异，F1分数在0.82至0.95之间。
  • 临床验证： 在前瞻性临床评估中，LUCAID在临床可操作决策方面与专家组裁定的参考标准达到了93.0%的一致性。相比之下，五位经验丰富的胸科病理学家的一致性在68.3%至81.1%之间。
  • Analytical Performance: Tested against large-scale expert ground-truth annotations, the individual analysis modules attained high accuracy with F1 scores ranging from 0.82 to 0.95.
  • Clinical Validation: In prospective clinical evaluations, LUCAID achieved a 93.0% concordance rate with an expert-panel adjudicated reference standard for clinically actionable decisions. This compared favorably against five experienced thoracic pathologists, who recorded concordance rates between 68.3% and 81.1%.

全文及其他资源

(注：有关本文的许可详情及权利，请参阅 知识共享署名-非商业性使用-禁止演绎 4.0 国际许可协议。)

(Note: License details and rights to this article can be reviewed via the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International license.)