文献速递临床血液与体液质量管理与危急值

血液学数字与计算形态学:现有平台、临床证据与未来需求

Digital and computational morphology in hematology: current platforms, clinical evidence, and future requirements.
PubMed 文献 · Clin Chem Lab Med · 2026-09-14

本综述评估数字形态学在外周血、骨髓、疟疾检测和体液分析中的应用,比较自动预分类与专家数字后分类及独立光学显微镜检查。数字系统对常见成熟白细胞表现良好,但对原始细胞等稀有或诊断关键细胞可靠性不足。临床价值取决于验证、参考方法透明报告和专家复核标准。

要点

  • 数字形态学可提升标准化、图像追溯、远程会诊和质量管理。
  • 对原始细胞、异常淋巴细胞等关键细胞,数字系统仍不够可靠。
  • 未来需多中心数据集、术语统一、外部验证和系统互操作性。

核心结论

  • 数字形态学可改善标准化、远程会诊和质量保证,但临床价值取决于适当验证和专家复核。

研究设计与样本

  • 综述评估数字形态学在外周血、骨髓、疟疾检测和体液分析中的当前应用与商业平台。

主要发现

  • 数字系统对常见成熟白细胞群表现良好,但对原始细胞、异常淋巴细胞、浆细胞等稀有或诊断关键细胞可靠性较低。
  • 研究系统正从单细胞分类扩展到全片、标本和患者层面的评估。

临床怎么用

  • 数字形态学可用于标准化、图像追溯、远程咨询、教育、能力验证和质量保证。
  • 需明确专家复核和传统显微镜检查的标准,并认识算法特定失效模式。

局限性

  • 未来需代表性多中心数据集、统一形态学术语、外部验证和软件硬件更新后持续监测。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

INTRODUCTION: Morphologic examination of peripheral blood and bone marrow remains central to the diagnosis and classification of hematologic disorders. Conventional optical microscopy, however, is labor-intensive, dependent on operator expertise, and affected by interobserver variability. Digital morphology has developed from automated image acquisition and cell pre-classification into a broader field that includes whole-slide imaging, remote review, quantitative morphometry, and artificial intelligence-based analysis. CONTENT: This review examines current applications of digital morphology in peripheral blood, bone marrow aspirates, malaria detection, and body-fluid analysis. Commercial platforms are evaluated with particular attention to the distinction between raw automated pre-classification, expert digital post-classification, and comparison with independent optical microscopy. Digital systems generally perform well for common mature leukocyte populations but remain less reliable for rare or diagnostically critical cells, including blasts, abnormal lymphoid cells, plasma cells, and intermediate maturation stages. Research systems increasingly extend analysis from individual-cell classification to whole-slide, specimen-level, and patient-level assessment. SUMMARY: Digital morphology can improve standardization, image traceability, remote consultation, education, proficiency testing, quality assurance, and selected aspects of laboratory workflow. Its clinical value depends on appropriate validation, transparent reporting of reference methods, recognition of algorithm-specific failure modes, and clearly defined criteria for expert review and conventional microscopy. Human expertise remains essential not only for validating results but also for adapting cell taxonomies and interpretive rules to evolving classifications of hematologic diseases. OUTLOOK: Future progress will require representative multicenter datasets, harmonized morphologic terminology, external validation, interoperability with laboratory information systems, and continuous monitoring after software or hardware updates. Integration of morphology with quantitative hematology, flow cytometry, cytogenetics, genomics, and clinical data may support more comprehensive computational diagnosis. Digital platforms may also broaden access to specialist expertise, training, and quality programs in resource-limited institutions and regions, provided that infrastructure, governance, and professional competency are adequately supported.

来源:Clin Chem Lab Med,查看原文。版权归原作者所有。

引用

  1. Zini G. Digital and computational morphology in hematology: current platforms, clinical evidence, and future requirements. Clin Chem Lab Med. 2026. PMID: 42732457. doi:10.1515/cclm-2026-1265

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。