Enhancing Conjunctival Image-Based Anemia Classification Using Coordinate Attention on MobileNetV3-Small

Abstract

This study evaluates the integration of attention mechanisms (Efficient Channel Attention [ECA] and Coordinate Attention [CoordAtt]) into MobileNetV3-Small for preliminary conjunctival anemia screening. Using a single public dataset of 710 pediatric images from Ghana (CP-AnemiC dataset), the model was evaluated retrospectively via stratified 10-fold cross-validation (80%) and tested on a hold-out test set (20%). CoordAtt achieved the highest performance (Accuracy 78%, Sensitivity 84%, Specificity 70%, AUC 0.85), outperforming the Baseline (Accuracy 75%, Sensitivity 80%, Specificity 67%, AUC 0.80) and ECA (Accuracy 76%, Sensitivity 83%, Specificity 66%, AUC 0.81). A 95% Bootstrap BCa confidence interval test confirmed that the CoordAtt improvement over the Baseline is statistically significant for Accuracy, F1-Macro, Sensitivity, and AUC. Grad-CAM visualizations further showed more focused conjunctival localization. This improvement came with only a minimal increase in theoretical computational footprint (1.06M parameters, 0.0584 GFLOPs), keeping a lightweight profile similar to the Baseline (1.03M parameters, 0.0581 GFLOPs). These results show that CoordAtt improves classification performance without sacrificing computational efficiency, suggesting potential suitability for resource-constrained devices prior to clinical validation.

Enhancing prenatal diagnostic workflow: a deep learning-based system for multi-class fetal ultrasound classification across all gestational stages
Chen-Shen Shih, Chia-Hao Danny Yang, Tung-Yao Chang +1
Scientific Reports28 Oct 2026Peer-reviewedCC BY-NC-ND 4.0Obstetrics, gynaecology & paediatricsCrossref
Feature-aware attention convolutional neural network for efficient pulmonary fibrosis classification from chest CT images with FPGA deployment
K. Mahapackialakshmi, G. Jaffino
Frontiers in medicine16 Oct 2026Peer-reviewedCC BYFull textGeneral medicineEurope PMC
TiLCA-Net: nuclei segmentation and classification in pathological images via task-specialized attention and efficient decoding
Liu Xiang, Xuejia Zheng, Huihui Tao +7
Quantitative imaging in medicine and surgery18 Sep 2026Peer-reviewedCC BY-NC-NDFull textMedical imaging & radiologyEurope PMC
CSwinFK: a bidirectional cross-attention CNN-Swin Transformer fusion model for fungal keratitis classification in in vivo confocal microscopy images
Miaoting Hu, Liang Liang, Lifei Chen +11
Scientific Reports13 Oct 2026Peer-reviewedCC BY-NC-ND 4.0Full textGeneral medicineCrossref
Deep Learning for Multi‐Disease OCT Classification in Clinical Retina Practice Using Attention Mechanisms and Contrastive Learning
Khosro Rezaee, Safoura Farsi, Hossein Ghayoumi Zadeh
Healthcare Technology Letters25 Oct 2026Peer-reviewedCC BY 4.0Full textMedical imaging & radiologyPMC

Cite

BibTeX RIS

BibTeX

@article{mahendra2026enhancing,
  title = {{Enhancing Conjunctival Image-Based Anemia Classification Using Coordinate Attention on MobileNetV3-Small}},
  author = {Arnanda Asshaffa Mahendra and Aris Marjuni and Affandy and Muhammad Naufal},
  journal = {Journal of Information Systems and Informatics},
  year = {2026},
  month = oct,
  doi = {10.63158/journalisi.v8i5.1796},
  url = {https://doi.org/10.63158/journalisi.v8i5.1796},
  note = {CC BY},
}

RIS

TY  - JOUR
TI  - Enhancing Conjunctival Image-Based Anemia Classification Using Coordinate Attention on MobileNetV3-Small
AU  - Arnanda Asshaffa Mahendra
AU  - Aris Marjuni
AU  - Affandy
AU  - Muhammad Naufal
PY  - 2026
DA  - 2026/10/24
JO  - Journal of Information Systems and Informatics
DO  - 10.63158/journalisi.v8i5.1796
UR  - https://doi.org/10.63158/journalisi.v8i5.1796
AB  - This study evaluates the integration of attention mechanisms (Efficient Channel Attention [ECA] and Coordinate Attention [CoordAtt]) into MobileNetV3-Small for preliminary conjunctival anemia screening. Using a single public dataset of 710 pediatric images from Ghana (CP-AnemiC dataset), the model was evaluated retrospectively via stratified 10-fold cross-validation (80%) and tested on a hold-out test set (20%). CoordAtt achieved the highest performance (Accuracy 78%, Sensitivity 84%, Specificity 70%, AUC 0.85), outperforming the Baseline (Accuracy 75%, Sensitivity 80%, Specificity 67%, AUC 0.80) and ECA (Accuracy 76%, Sensitivity 83%, Specificity 66%, AUC 0.81). A 95% Bootstrap BCa confidence interval test confirmed that the CoordAtt improvement over the Baseline is statistically significant for Accuracy, F1-Macro, Sensitivity, and AUC. Grad-CAM visualizations further showed more focused conjunctival localization. This improvement came with only a minimal increase in theoretical computational footprint (1.06M parameters, 0.0584 GFLOPs), keeping a lightweight profile similar to the Baseline (1.03M parameters, 0.0581 GFLOPs). These results show that CoordAtt improves classification performance without sacrificing computational efficiency, suggesting potential suitability for resource-constrained devices prior to clinical validation.
SN  - 2656-5935
SN  - 2656-4882
ER  -

Abstract

Related

Cite