Computer vision quantitation of erythrocyte shape abnormalities provides diagnostic, prognostic, and mechanistic insight.
Computer vision quantitation of erythrocyte shape abnormalities provides diagnostic, prognostic, and mechanistic insight.
复制标题
DOI：
10.1182/bloodadvances.2022008967
复制
发表时间：
2023-08-22
期刊：
影响因子：
7.5
通讯作者：
Makar, Robert S.
中科院分区：
文献类型：
--
作者：
Foy, Brody H.;Stefely, Jonathan A.;Bendapudi, Pavan K.;Hasserjian, Robert P.;Al-Samkari, Hanny;Louissaint Jr, Abner;Fitzpatrick, Megan J.;Hutchison, Bailey;Mow, Christopher;Collins, Julia;Patel, Hasmukh R.;Patel, Chhaya H.;Patel, Nikita;Ho, Samantha N.;Kaufman, Richard M.;Dzik, Walter H.;Higgins, John M.;Makar, Robert S.
使用计算机视觉系统可以准确地对外周血涂片中的RBC畸形进行精确定量。这种定量允许改善多种血液病状态的诊断和预后评估。外周血涂片中红细胞（RBC）形态学检查可以帮助诊断血液病，即使在资源有限的环境中，但这种分析仍然是主观和半定量的，通量低。先前开发自动化工具的尝试受到其再现性差和临床验证有限的阻碍。在这里，我们提出了一种新的开源机器学习方法（表示为RBC-diff）来量化外周涂片图像中的异常RBC并生成RBC形态差异。RBC-diff细胞计数显示单细胞分类（平均AUC，0.93）和跨涂片定量（平均R2，0.76，与专家，interexperts R2，0.75相比）的准确性较高。RBC-diff计数与300 000+图像的临床形态学分级一致，并在不同的临床队列中恢复了预期的病理生理信号。使用RBC-diff计数的标准区分血栓性血小板减少性紫癜和溶血性尿毒症综合征与其他血栓性微血管病，提供比临床形态学分级更高的特异性（72%对41%; P <0.001），同时保持高灵敏度（94%至100%）。在58950例住院患者队列中，RBC-diff裂细胞计数升高与6个月全因死亡率增加相关（裂细胞死亡率为9.5%）。> 1%，而片岩为4.7%; <0.5%; P <0.001）。RBC-diff还能够估计单细胞体积-形态分布，从而深入了解形态对常规血细胞计数测量的影响。这里包含了我们的代码库和专家注释的图像，以促进进一步的发展。这些结果表明，计算机视觉可以快速准确地定量RBC形态，这可能在临床和研究背景下提供价值。
Exact quantitation of RBC dysmorphologies in peripheral blood smears can be accurately performed using a computer vision system. This quantitation allowed for improved diagnostic and prognostic evaluations of multiple hematologic disease states. Examination of red blood cell (RBC) morphology in peripheral blood smears can help diagnose hematologic diseases, even in resource-limited settings, but this analysis remains subjective and semiquantitative with low throughput. Prior attempts to develop automated tools have been hampered by their poor reproducibility and limited clinical validation. Here, we present a novel, open-source machine-learning approach (denoted as RBC-diff) to quantify abnormal RBCs in peripheral smear images and generate an RBC morphology differential. RBC-diff cell counts showed high accuracy for single-cell classification (mean AUC, 0.93) and quantitation across smears (mean R2, 0.76 compared with experts, interexperts R2, 0.75). RBC-diff counts were concordant with the clinical morphology grading for 300 000+ images and recovered the expected pathophysiologic signals in diverse clinical cohorts. Criteria using RBC-diff counts distinguished thrombotic thrombocytopenic purpura and hemolytic uremic syndrome from other thrombotic microangiopathies, providing greater specificity than clinical morphology grading (72% vs 41%; P < .001) while maintaining high sensitivity (94% to 100%). Elevated RBC-diff schistocyte counts were associated with increased 6-month all-cause mortality in a cohort of 58 950 inpatients (9.5% mortality for schist. >1%, vs 4.7% for schist; <0.5%; P < .001) after controlling for comorbidities, demographics, clinical morphology grading, and blood count indices. RBC-diff also enabled the estimation of single-cell volume-morphology distributions, providing insight into the influence of morphology on routine blood count measures. Our codebase and expert-annotated images are included here to spur further advancement. These results illustrate that computer vision can enable rapid and accurate quantitation of RBC morphology, which may provide value in both clinical and research contexts.
登录
查看更多内容
影响因子：
3.9
作者：
Song W;Huang P;Wang J;Shen Y;Zhang J;Lu Z;Li D;Liu D
通讯作者：
Liu D
影响因子：
24.7
作者：
Bendapudi, Pavan K.;Hurwitz, Shelley;Makar, Robert S.
通讯作者：
Makar, Robert S.
影响因子：
3
作者：
Bruegel, Mathias;George, Tracy I.;Russcher, Henk
通讯作者：
Russcher, Henk
影响因子：
3
作者：
Lesesve, J-F;Speyer, E.;Perol, J-P
通讯作者：
Perol, J-P
影响因子：
3
作者：
Noutsos T;Currie BJ;Brown SG;Isbister GK
通讯作者：
Isbister GK