https://atlas.rsna.org/schemas/2025-11/dataset.json

RSNA Abdominal Traumatic Injury CT (RATIC) Dataset

https://mira.rsna.org/dataset/5

1.0

informatics@rsna.org

The RSNA Abdominal Traumatic Injury CT (RATIC) dataset is the largest publicly available adult abdominal traumatic injury CT dataset, with contributions from 23 institutions across 14 countries and six continents. It contains CT studies with annotations related to traumatic injuries of the liver, spleen, kidneys, bowel, mesentery, and active extravasation. The dataset was used for the RSNA 2023 Abdominal Trauma Detection competition.

Early accurate diagnosis and grading of traumatic injuries is critical. Automated assessment of traumatic abdominal injuries is an excellent use case for AI algorithms to prioritize studies and augment radiologist accuracy and efficiency. There is a need for large multi-institutional publicly available annotated abdominal trauma datasets.

Participating sites were asked to enrich the dataset with representative injuries given the relatively low prevalence of traumatic abdominal injuries at CT. Broad inclusion criteria for CT scans were chosen to facilitate a larger and more diverse dataset for training robust ML models.

Reference standard labels for solid organ injury grades were established using majority voting among three annotators and divided into low-grade (AAST I–III) and high-grade (IV and V) injury groups. Image-level labels for bowel and mesenteric injuries and active extravasation were based on annotator consensus. An explicit effort was made to reduce potential biases by considering sex, age, injuries, and contributing site when assigning scans to training, public test, and private test datasets.

Absence of delayed phase imaging is a limitation. Other injuries such as hematomas, fractures, and lower thoracic injuries are present but were not explicitly annotated due to challenge timeline and focus on critical findings.

Diagnostic errors in trauma interpretation are common, and there is high interrater variability in the AAST grading system. Large variation in protocols used at different hospitals, including single portal venous phase, multiphasic imaging, and split bolus approaches, complicates the task. Dramatic differences in z-axis coverage in included CT scans and low prevalence of traumatic abdominal injuries were also challenges.