血型
ABO血型系统
重复性
计算机科学
人工智能
计算机视觉
模式识别(心理学)
色谱法
生物
化学
遗传学
免疫学
作者
Hong Zhang,Ruining Liu,Qingmei Li,Xiaolin Hu,Lixiang Wu,Ye Zhou,Guangchao Qing,Rui Yuan,Junjie Huang,Wei Gu,Yanyao Ye,Chao Qi,Meng Han,Xiaohui Chen,Xun Zhu,Yun Deng,Liangliang Zhang,Hengyi Chen,Haoran Zhang,Weiyin Gao,Yao Liu,Yang Luo
出处
期刊:ACS Nano
[American Chemical Society]
日期:2021-04-19
卷期号:15 (4): 7649-7658
被引量:11
标识
DOI:10.1021/acsnano.1c01215
摘要
Accurate and rapid blood typing plays a vital role in a variety of biomedical and forensic scenarios, but recognizing weak agglutination remains challenging. Herein, we demonstrated a flipping identification with a prompt error-discrimination (FLIPPED) platform for automatic blood group readouts. Bromocresol green dye was exploited as a characteristic chromatography indicator for the differentiation of plasma from whole blood by presenting a teal color against a brown color. After integrating these color changes into a quick-response (QR) code, prompt typing of ABO and Rhesus groups was automatically achieved and data could be uploaded wirelessly within 30 s using a commercially available smartphone to facilitate blood cross-matching. We further designed a color correction model and algorithm to remove potential errors from scanning angles and ambient light intensities, by which weak agglutination could be accurately recognized. With comparable accuracy and repeatability to classical column assay in grouping 450 blood samples, the proposed approach further demonstrates to be a versatile sample-to-result platform for clinical diagnostics, food safety, and environmental monitoring.
科研通智能强力驱动
Strongly Powered by AbleSci AI