RadioX-Labs/USI-FLFib

If you are not subscribed to Abdominal Radiology - Springer Link Journal, you can view the full version of our paper by clicking on this link

RXL-USI-FLFib is an original dataset collected at a tertiary care referral center, as part of a prospective study on hepatocellular carcinoma screening (NCT05716620) after informed written consent.

Image acquisition protocol:

• All US scans were performed using a GE LOGIQ S8 scanner with a C1-6-D curvilinear transducer (frequency range: 1–6 MHz).
• To standardize the acquisitions, US examinations were performed after at least 6h of fasting by a radiologist with 1–10 years of experience in abdominal imaging.
• Images were acquired from the right intercostal space, ensuring inclusion of the liver capsule and diaphragm, and from the subxiphoid window to view the left lobe.
• Acquisition depth was set between 15 and 18 cm with the focus placed at the posterior aspect of the liver, and gain was adjusted per patient to optimize image quality.

---

Dataset Characteristics

Characteristic	Value
Total Image	1,131
Total Patients	111
Modality	B-mode Ultrasound Images

Metadata

The complete dataset includes a metadata file with label information for each image:

filename,fl_label,fib_label
P157/1.jpg,1,1
P157/2.jpg,1,1
P157/3.jpg,1,1
P157/4.jpg,1,1
...

Metadata fields:

• filename: Individual Images are formmated as {Patient Folder}/{Image Number}
• fl_label: Fatty Liver Label
• fib_label: Fibrosis Label

The full metadata file is available at train_DS1.csv.

Dataset Access

To request access to the RXL-USI-FLFib dataset, please complete the following form:

Request Dataset Access

Access Process

1. Complete the access request form with your institutional details
2. Specify your intended use case for the dataset
3. Agree to the dataset license terms (CC BY-NC-SA 4.0)
4. You will receive download instructions via email

---

Citation

If you use RXL-RADSet in your research, please cite:

Bose, K., Mudgil, P., Gupta, P. et al. Deep learning for non-invasive detection of steatosis and fibrosis in MASLD: a multicenter study with a new fibroscan-labelled ultrasound dataset. Abdom Radiol (2025). https://doi.org/10.1007/s00261-025-05309-9

---

License

This dataset is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0).

You are free to:

• Share: Copy and redistribute the material in any medium or format
• Adapt: Remix, transform, and build upon the material

Under the following terms:

• Attribution: You must give appropriate credit
• NonCommercial: You may not use the material for commercial purposes
• ShareAlike: If you remix, transform, or build upon the material, you must distribute your contributions under the same license

---

Contact

For questions about the dataset, access requests, or collaboration opportunities, please email us by clicking here

---