Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CBCTLiTS: A Synthetic, Paired CBCT/CT Dataset For Segmentation And Style Transfer

Maximilian E. Tschuchnig, Philipp Steininger, Michael Gadermayr

TL;DR

CBCTLiTS introduces a synthetic, paired CBCT/CT dataset derived from LiTS to enable liver and liver tumor segmentation and style-transfer research in intraoperative imaging. It provides five CBCT quality levels, ground truth segmentations, and a 131/70 train/test split, plus baseline experiments in uni- and multimodal segmentation, multitask learning, and CT-style transfer. Key findings show strong gains from multimodal learning with good alignment, moderate improvements from multitask approaches, and promising liver segmentation improvements from CBCT-to-CT style transfer, while tumor segmentation remains challenging under most conditions. The dataset offers a reproducible resource for evaluating segmentation and image-translation methods under artifact- and quality-variant CBCT conditions, with code and data accessible on Kaggle, advancing research in intraoperative CBCT analysis.

Abstract

Medical imaging is vital in computer assisted intervention. Particularly cone beam computed tomography (CBCT) with defacto real time and mobility capabilities plays an important role. However, CBCT images often suffer from artifacts, which pose challenges for accurate interpretation, motivating research in advanced algorithms for more effective use in clinical practice. In this work we present CBCTLiTS, a synthetically generated, labelled CBCT dataset for segmentation with paired and aligned, high quality computed tomography data. The CBCT data is provided in 5 different levels of quality, reaching from a large number of projections with high visual quality and mild artifacts to a small number of projections with severe artifacts. This allows thorough investigations with the quality as a degree of freedom. We also provide baselines for several possible research scenarios like uni- and multimodal segmentation, multitask learning and style transfer followed by segmentation of relatively simple, liver to complex liver tumor segmentation. CBCTLiTS is accesssible via https://www.kaggle.com/datasets/maximiliantschuchnig/cbct-liver-and-liver-tumor-segmentation-train-data.

CBCTLiTS: A Synthetic, Paired CBCT/CT Dataset For Segmentation And Style Transfer

TL;DR

CBCTLiTS introduces a synthetic, paired CBCT/CT dataset derived from LiTS to enable liver and liver tumor segmentation and style-transfer research in intraoperative imaging. It provides five CBCT quality levels, ground truth segmentations, and a 131/70 train/test split, plus baseline experiments in uni- and multimodal segmentation, multitask learning, and CT-style transfer. Key findings show strong gains from multimodal learning with good alignment, moderate improvements from multitask approaches, and promising liver segmentation improvements from CBCT-to-CT style transfer, while tumor segmentation remains challenging under most conditions. The dataset offers a reproducible resource for evaluating segmentation and image-translation methods under artifact- and quality-variant CBCT conditions, with code and data accessible on Kaggle, advancing research in intraoperative CBCT analysis.

Abstract

Medical imaging is vital in computer assisted intervention. Particularly cone beam computed tomography (CBCT) with defacto real time and mobility capabilities plays an important role. However, CBCT images often suffer from artifacts, which pose challenges for accurate interpretation, motivating research in advanced algorithms for more effective use in clinical practice. In this work we present CBCTLiTS, a synthetically generated, labelled CBCT dataset for segmentation with paired and aligned, high quality computed tomography data. The CBCT data is provided in 5 different levels of quality, reaching from a large number of projections with high visual quality and mild artifacts to a small number of projections with severe artifacts. This allows thorough investigations with the quality as a degree of freedom. We also provide baselines for several possible research scenarios like uni- and multimodal segmentation, multitask learning and style transfer followed by segmentation of relatively simple, liver to complex liver tumor segmentation. CBCTLiTS is accesssible via https://www.kaggle.com/datasets/maximiliantschuchnig/cbct-liver-and-liver-tumor-segmentation-train-data.
Paper Structure (10 sections, 2 figures, 1 table)

This paper contains 10 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Contributions
  3. Dataset
  4. Experimental Setup
  5. Combining segmentation and image reconstruction using multitask learning
  6. Multimodal learning, combining preoperative CT and intraopertative CBCT
  7. Converting CBCT to CT using style transfer
  8. Results
  9. Discussion
  10. Conclusion

Figures (2)

  • Figure 1: Process of CBCTLiTS generation: after centering the original CT volumes around the liver (using the liver segmentations), DRRs were simulated from these centered CT. Then, CBCT were simulated and aligned with the original CT and masks.
  • Figure 2: Different volumes of CBCTLiTS data (vertically). Horizontally, the presented data consists of the aligned mask, the aligned CT and the $5$ different qualities of synthesised CBCT.