EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

AI ‐Enabled Automated Schistocyte Classification in Peripheral Blood for TMA Auxiliary Diagnosis

Lei Shang iD, Chao Fang, Dongshuo Li, Xuesong Wang, Jiani Yu, Yunchu Cui, Jingliang Chen, Xuekai Liu

International Journal of Laboratory Hematology · 2026-12 · 48 (6) · 1486-1493

DOI: 10.1111/ijlh.70209 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Laboratory Hematology, online erschienen am 05.08.2026. Verfasst von Lei Shang und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Clinical Laboratory, Aerospace Center Hospital, Beijing, China. Hauptthemen laut MEDLINE: Thrombotic Microangiopathies, Artificial Intelligence, Erythrocytes, Abnormal. Die Arbeit nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Peking University, Aerospace Center Hospital.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Schistocytes are critical morphological markers for thrombotic microangiopathy (TMA) diagnosis. Manual identification is hampered by inconsistent standards and high interobserver variability, compromising accuracy. Although AI has advanced in hematology, AI‐enabled schistocyte classification remains under‐studied, creating a clinical gap.

Methods

We constructed a two‐stage AI system for segmentation and classification. Three segmentation models (U‐Net, ENet, and R2U‐Net) were assessed on 25 067 RBCs from 183 patients. ResNet‐50 and Xception were trained on 28 586 RBCs, including 13 125 ICSH‐classified schistocytes, using grayscale and RGB images. Clinical validation was performed on 156 784 RBCs from 219 patients. Performance was evaluated using recall, specificity, precision, and F 1 score.

Results

R2U‐Net achieved the best segmentation (recall = 0.868, F 1 = 0.881, mIoU = 0.807). The Xception‐RGB model performed best in classification (weighted F 1 = 0.957), with high precision and recall for schistocyte subtypes. Clinical validation showed excellent reliability with all weighted metrics at 0.998.

Conclusions

This two‐stage framework enables accurate schistocyte analysis with better feature representation using RGB images and Xception. It improves diagnostic accuracy and reproducibility, supporting TMA auxiliary diagnosis and clinical decision‐making.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Thrombotic MicroangiopathiesArtificial IntelligenceErythrocytes, AbnormalHumansFemaleImage Processing, Computer-Assisted

Stichwörter der Autor:innen

ResNetXceptionartificial intelligencedeep learningimage classificationimage segmentationperipheral blood smearschistocytesthrombotic microangiopathy (TMA)

Autor:innen und Institutionen

  • Department of Clinical Laboratory, Aerospace Center Hospital, Beijing, China.
  • Faculty of Engineering and Design, University of Auckland, Auckland, New Zealand.
  • School of Science, Beijing Jiaotong University, Beijing, China.
  • School of Public Health, Xiamen University, Xiamen, China.

ORCID-Profile: Lei Shang

Förderung

  • Peking University
  • Aerospace Center Hospital

Zitieren

APA

Shang, L., Fang, C., Li, D., Wang, X., Yu, J., Cui, Y., Chen, J., & Liu, X. (2026). AI ‐Enabled Automated Schistocyte Classification in Peripheral Blood for TMA Auxiliary Diagnosis. International Journal of Laboratory Hematology, 48(6), 1486-1493. https://doi.org/10.1111/ijlh.70209

Vancouver

Shang L, Fang C, Li D, Wang X, Yu J, Cui Y, et al. AI ‐Enabled Automated Schistocyte Classification in Peripheral Blood for TMA Auxiliary Diagnosis. International Journal of Laboratory Hematology. 2026;48(6):1486-1493. doi:10.1111/ijlh.70209

BibTeX
@article{Shang2026,
  author = {Shang, Lei and Fang, Chao and Li, Dongshuo and Wang, Xuesong and Yu, Jiani and Cui, Yunchu and Chen, Jingliang and Liu, Xuekai},
  title = {AI ‐Enabled Automated Schistocyte Classification in Peripheral Blood for TMA Auxiliary Diagnosis},
  journal = {International Journal of Laboratory Hematology},
  year = {2026},
  volume = {48},
  number = {6},
  pages = {1486--1493},
  doi = {10.1111/ijlh.70209}
}
Kostenlos starten →