EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Dynamic attention ordering in self-adaptive convolutional block attention module for chest X-ray image classification

Fatin Nabilah Shaari iD, Aimi Salihah Abdul Nasir iD, Wan Aireene Wan Ahmed

Biomedical Physics & Engineering Express · 17.11.2026

DOI: 10.1088/2057-1976/aec2f7 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biomedical Physics & Engineering Express, online erschienen am 17.11.2026. Verfasst von Fatin Nabilah Shaari und 2 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Electrical Engineering & Technology, Universiti Malaysia Perlis, Pauh Putra Campus, Arau, 02600, Malaysia. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Fundamental Research Grant Scheme.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Convolutional Block Attention Module (CBAM) is widely used attention mechanism to boost the feature representation ability of the Convolutional Neural Network (CNN) by performing channel attention and spatial attention alternately. However, CBAM is hard to use for various distributions of the features, because of its fixed channel-to-spatial attention ordering. Therefore, this study proposes a Self-Adaptive Convolutional Block Attention Module (SA-CBAM) which introduces a lightweight auxiliary decision network for dynamically ordering the attention. The auxiliary network uses the global characteristics of the attention feature map to determine whether the spatial attention or the channel attention should be applied first, which is more flexible than the conventional CBAM for feature refinement. The effectiveness of the proposed module is assessed by systematically investigating six variants of the SA-CBAM where the standard, lightweight, early-stage, middle-stage and late-stage are investigated using a baseline CNN. The top-performing variant was then compared to representative attentional approaches and shown to be valid on several benchmark datasets that included chest X-ray (CXR) classification, remote sensing, natural image, microscopic image, and histopathological image classification tasks. Finally, the optimal SA-CBAM was combined with a hybrid CNN-LSTM and evaluated for sequential feature learning. Through experimental results, it was verified that the proposed SA-CBAM outperformed the conventional CBAM on all the test sets with similar complexity. The CNN+SA-CBAM model had accuracy of 90.88% and MCC of 86.52% on the primary CXR dataset, while the hybrid CNN+SA-CBAM-LSTM model further improved the accuracy to 99.88%. The results on different datasets show the effectiveness and computational efficiency of the proposed dynamic attention-ordering mechanism and make SA-CBAM robust and generalizable attention module for deep image classification.

Schlagwörter

Stichwörter der Autor:innen

CNN-LSTMattention mechanismchest X-rayexplainable AIself-adaptive CBAM

Autor:innen und Institutionen

  • Faculty of Electrical Engineering & Technology, Universiti Malaysia Perlis, Pauh Putra Campus, Arau, 02600, Malaysia.
  • Universiti Sains Malaysia - Kampus Kesihatan, Kubang Kerian, Kubang Kerian, Kelantan, 16150, Malaysia.

ORCID-Profile: Fatin Nabilah Shaari, Aimi Salihah Abdul Nasir

Förderung

  • Fundamental Research Grant Scheme

Zitieren

APA

Shaari, F. N., Abdul Nasir, A. S., & Wan Ahmed, W. A. (2026). Dynamic attention ordering in self-adaptive convolutional block attention module for chest X-ray image classification. Biomedical Physics & Engineering Express. https://doi.org/10.1088/2057-1976/aec2f7

Vancouver

Shaari FN, Abdul Nasir AS, Wan Ahmed WA. Dynamic attention ordering in self-adaptive convolutional block attention module for chest X-ray image classification. Biomedical Physics & Engineering Express. 2026. doi:10.1088/2057-1976/aec2f7

BibTeX
@article{Shaari2026,
  author = {Shaari, Fatin Nabilah and Abdul Nasir, Aimi Salihah and Wan Ahmed, Wan Aireene},
  title = {Dynamic attention ordering in self-adaptive convolutional block attention module for chest X-ray image classification},
  journal = {Biomedical Physics & Engineering Express},
  year = {2026},
  doi = {10.1088/2057-1976/aec2f7}
}
Kostenlos starten →