Enhancing HF-DL Model Validation for Liver Fibrosis Staging Through Sample Optimisation and Technical Integration

We read with great interest the article by Zhang et al. The study demonstrates that the deep learning model based on high-frequency ultrasound images significantly outperforms the low-frequency ultrasound model, FIB-4, APRI, and shear wave elastography in classifying liver fibrosis at each stage in patients with chronic hepatitis B. However, we would like to highlight several areas for further discussion.