EUVIMEDEUVIMED entdecken
Crossref

Evaluating Kolmogorov-Arnold Network Implementations for Medical Ultrasound Image Classification: A Trade-off Analysis of Computational Cost, Interpretability, and Diagnostic Safety

Saim Ervural iD

Ultrasonic Imaging · 19.08.2026

DOI: 10.1177/01617346261476280 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Ultrasonic Imaging, online erschienen am 19.08.2026. Verfasst von Saim Ervural. Beteiligte Einrichtung: Department of Electrical and Electronics Engineering, KTO Karatay University, Konya, Turkey. Die Arbeit nennt 16 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To compare pykan and Lightweight KAN with MLP and Linear classifier heads for focal liver lesion classification on B-mode ultrasound, with emphasis on diagnostic sensitivity, interpretability and computational cost. A frozen ResNet18 feature extractor was combined with 4 classifier heads and evaluated on 735 annotated liver ultrasound images (Benign, Malignant, Normal) using 5-fold stratified cross-validation. Weighted cross-entropy was applied to address class imbalance. Performance was assessed using Accuracy, F1-Macro, F1-Weighted, AUROC, class-wise recall, malignant-to-normal misclassification count, inference latency, and pykan-specific spline activeness analysis. Overall metrics were similar across models, and Wilcoxon Signed-Rank testing showed no statistically significant pairwise differences ( p > .05). However, clinically relevant class-level differences emerged. pykan achieved the highest Benign recall (65.5%) and the lowest malignant-to-normal error rate (5/435, 1.1%). Lightweight KAN achieved the highest Malignant recall (82.1%) and markedly faster inference (0.69 ms/image) than pykan (4183.67 ms/image). Aggregate metrics alone do not capture clinically meaningful trade-offs among KAN implementations. Lightweight KAN is attractive for rapid screening, whereas pykan shows a directional trend toward fewer malignant-to-normal errors and a more interpretable profile, though this trend did not reach statistical significance and requires confirmation in larger studies.

Schlagwörter

Stichwörter der Autor:innen

Kolmogorov-Arnold networksexplainable AIfocal liver lesionsmedical image classificationultrasound imaging

Autor:innen und Institutionen

  • Department of Electrical and Electronics Engineering, KTO Karatay University, Konya, Turkey.
  • Smart Technologies Design, Development and Prototyping Centre (AKITEK), Konya, Turkey.

ORCID-Profile: Saim Ervural

Zitieren

APA

Ervural, S. (2026). Evaluating Kolmogorov-Arnold Network Implementations for Medical Ultrasound Image Classification: A Trade-off Analysis of Computational Cost, Interpretability, and Diagnostic Safety. Ultrasonic Imaging. https://doi.org/10.1177/01617346261476280

Vancouver

Ervural S. Evaluating Kolmogorov-Arnold Network Implementations for Medical Ultrasound Image Classification: A Trade-off Analysis of Computational Cost, Interpretability, and Diagnostic Safety. Ultrasonic Imaging. 2026. doi:10.1177/01617346261476280

BibTeX
@article{Ervural2026,
  author = {Ervural, Saim},
  title = {Evaluating Kolmogorov-Arnold Network Implementations for Medical Ultrasound Image Classification: A Trade-off Analysis of Computational Cost, Interpretability, and Diagnostic Safety},
  journal = {Ultrasonic Imaging},
  year = {2026},
  doi = {10.1177/01617346261476280}
}
Kostenlos starten →