Classification of gestational diabetes mellitus placental histopathological images based on MONAI and improved GAMatrix-YOLOv8.
4 affiliations
- 1Jinan University, Guangzhou, China.
- 2Obstetrics and Gynaecology, Huizhou First Maternal and Child Health Care Hospital, Huizhou, Guangdong, China.
- 3Department of Soft Engineering, Huizhou University, Huizhou, Guangdong, China.
- 4Department of Soft Engineering, Huizhou University, Huizhou, Guangdong, China. [email protected].
Abstract
Background
Gestational diabetes mellitus (GDM) poses a significant threat to maternal and fetal health, and placental pathology is crucial for assessing disease severity. Traditional diagnosis relies on clinicians' visual assessment, which is often inefficient and subjective.Methods
To address these issues, we developed a MONAI-based image augmentation strategy and integrated the GAMatrix attention module within the YOLOv8 framework. Our study used a dataset of 5,065 placental pathology images (80% for training and 20% for validation) covering five pathology categories, supplemented by an external validation set of 500 images. A MONAI-based image augmentation pipeline was used to increase training-image diversity.Results
Four model configurations were compared: YOLOv8 trained on the original dataset, YOLOv8 trained with MONAI-based augmentation, GAMatrix-YOLOv8 trained on the original dataset, and GAMatrix-YOLOv8 trained with MONAI-based augmentation. In the original comparative analysis, YOLOv8 trained on the original dataset achieved a macro-F1 score of 95.56% (95% CI, 94.86%-96.26%) in internal validation. The GAMatrix-YOLOv8 model trained with MONAI-based augmentation showed the strongest overall performance among the evaluated configurations, while external image-level evaluation indicated persistent class-specific errors, particularly for thrombotic vascular lesion and chorangiosis.Conclusion
These findings suggest that attention-enhanced YOLOv8 combined with MONAI-based augmentation may support placental histopathological image classification, but the results should be interpreted as image-level findings and require patient-level, multicenter, and prospective validation.
Related
- Quercetin ameliorates gestational diabetes mellitus-induced placental dysfunction by modulating the AGE-RAGE-associated oxidative and inflammatory signaling network
- A machine learning-based predictive model for gestational diabetes mellitus based on first trimester clinical indicators
- Pregnancy Complicated by Citrin Deficiency and Gestational Diabetes Mellitus: Perioperative Metabolic Management and Postoperative Analgesic Challenges Following Caesarean Delivery
- Preconceptional Risk Factors Associated With Gestational Diabetes Mellitus and Pregnancy Outcomes: A Case-Control Study
- Performance and Safety of a Continuous Glucose Monitoring System in Pregnant Women With Gestational Diabetes Mellitus: A Multicenter, Open-Label Trial
BibTeX
@article{jiang2026classification,
title = {{Classification of gestational diabetes mellitus placental histopathological images based on MONAI and improved GAMatrix-YOLOv8.}},
author = {Zhifa Jiang and Weirui Wu and Jingwen Liu and Xiekun Chen and Ruoping Lin and Xiangyun Ye and Donghui Huang and Zhen Zhang},
journal = {Biomedical engineering online},
year = {2026},
month = jul,
doi = {10.1186/s12938-026-01602-w},
url = {https://europepmc.org/article/MED/42919373},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Classification of gestational diabetes mellitus placental histopathological images based on MONAI and improved GAMatrix-YOLOv8.
AU - Zhifa Jiang
AU - Weirui Wu
AU - Jingwen Liu
AU - Xiekun Chen
AU - Ruoping Lin
AU - Xiangyun Ye
AU - Donghui Huang
AU - Zhen Zhang
PY - 2026
DA - 2026/07/01
JO - Biomedical engineering online
DO - 10.1186/s12938-026-01602-w
UR - https://europepmc.org/article/MED/42919373
AB - Background Gestational diabetes mellitus (GDM) poses a significant threat to maternal and fetal health, and placental pathology is crucial for assessing disease severity. Traditional diagnosis relies on clinicians' visual assessment, which is often inefficient and subjective.Methods To address these issues, we developed a MONAI-based image augmentation strategy and integrated the GAMatrix attention module within the YOLOv8 framework. Our study used a dataset of 5,065 placental pathology images (80% for training and 20% for validation) covering five pathology categories, supplemented by an external validation set of 500 images. A MONAI-based image augmentation pipeline was used to increase training-image diversity.Results Four model configurations were compared: YOLOv8 trained on the original dataset, YOLOv8 trained with MONAI-based augmentation, GAMatrix-YOLOv8 trained on the original dataset, and GAMatrix-YOLOv8 trained with MONAI-based augmentation. In the original comparative analysis, YOLOv8 trained on the original dataset achieved a macro-F1 score of 95.56% (95% CI, 94.86%-96.26%) in internal validation. The GAMatrix-YOLOv8 model trained with MONAI-based augmentation showed the strongest overall performance among the evaluated configurations, while external image-level evaluation indicated persistent class-specific errors, particularly for thrombotic vascular lesion and chorangiosis.Conclusion These findings suggest that attention-enhanced YOLOv8 combined with MONAI-based augmentation may support placental histopathological image classification, but the results should be interpreted as image-level findings and require patient-level, multicenter, and prospective validation.
SN - 1475-925X
ER -