TY - GEN
T1 - Advancing OCT-Based Retinal Disease Classification with XLSTM
T2 - 22nd IEEE International Symposium on Biomedical Imaging, ISBI 2025
AU - Sükei, Emese
AU - Oghbaie, Marzieh
AU - Schmidt-Erfurth, Ursula
AU - Klambauer, Günter
AU - Bogunović, Hrvoje
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025/5/12
Y1 - 2025/5/12
N2 - This paper presents a method for retinal disease classification using optical coherence tomography (OCT) scans, specifically addressing the challenge of variable B-scan density across dataset volumes. Deep learning methods often struggle with fluctuations in the number of B-scans within OCT volumes, which can limit their effectiveness. To address this, we propose an extended long short-term memory (xLSTM) model capable of processing variable-length B-scan sequences, capturing spatial and contextual information across entire volumes regardless of slice count. Leveraging xLSTM's advanced memory structure, our method enhances classification accuracy while maintaining computational efficiency. We evaluated the xLSTM-based approach on public and private datasets, demonstrating its robustness against inter-study variability and domain shifts. Experimental results show that this approach performs on par with or surpasses state-of-the-art baselines and offers a scalable solution for accurate 3D volume classification.
AB - This paper presents a method for retinal disease classification using optical coherence tomography (OCT) scans, specifically addressing the challenge of variable B-scan density across dataset volumes. Deep learning methods often struggle with fluctuations in the number of B-scans within OCT volumes, which can limit their effectiveness. To address this, we propose an extended long short-term memory (xLSTM) model capable of processing variable-length B-scan sequences, capturing spatial and contextual information across entire volumes regardless of slice count. Leveraging xLSTM's advanced memory structure, our method enhances classification accuracy while maintaining computational efficiency. We evaluated the xLSTM-based approach on public and private datasets, demonstrating its robustness against inter-study variability and domain shifts. Experimental results show that this approach performs on par with or surpasses state-of-the-art baselines and offers a scalable solution for accurate 3D volume classification.
KW - 3D volume classification
KW - Optical coherence tomography
KW - retinal disease classification
KW - xLSTM
UR - https://www.scopus.com/pages/publications/105005830741
U2 - 10.1109/ISBI60581.2025.10981206
DO - 10.1109/ISBI60581.2025.10981206
M3 - Conference proceedings
AN - SCOPUS:105005830741
SN - 979-8-3315-2053-3
T3 - Proceedings - International Symposium on Biomedical Imaging
SP - 1
EP - 5
BT - 2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI)
PB - IEEE Computer Society 2012
CY - Houston, TX, USA
Y2 - 14 April 2025 through 17 April 2025
ER -