Daily Arxiv

This page organizes papers related to artificial intelligence published around the world.
This page is summarized using Google Gemini and is operated on a non-profit basis.
The copyright of the paper belongs to the author and the relevant institution. When sharing, simply cite the source.

A Chain-of-thought Reasoning Breast Ultrasound Dataset Covering All Histopathology Categories

Created by
  • Haebom

Author

Haojun Yu, Youcheng Li, Zihan Niu, Nan Zhang, Xuantong Gong, Huan Li, Zhiying Zou, Haifeng Qi, Zhenxiao Cao, Zijie Lan, Xingjian Yuan, Jiating He, Haokai Zhang, Shengtao Zhang, Zicheng Wang, Dong Wang, Ziwei Zhao, Congying Chen, Yong Wang, Wangyan Qin, Qingli Zhu, Liwei Wang

Outline

BUS-CoT is a novel dataset for Chain of Ties (CoT) inference analysis using breast ultrasound (BUS) images. It contains 11,439 images of 10,019 lesions, representing 4,838 patients and covering all 99 histopathological types. Experienced experts annotated and validated observations, features, diagnoses, and pathological labels to build the inference process that drives CoT inference. The goal is to develop a robust AI system that encompasses all histopathological types of lesions, even in rare cases prone to error in clinical practice.

Takeaways, Limitations

•
Takeaways:
◦
Providing a large, diverse breast ultrasound dataset to establish a benchmark for AI development.
◦
Suggesting the possibility of developing AI for rare disease diagnosis, including all histopathological types.
◦
Contributes to improving the explanatory power of AI models by providing annotated data for Chain of Ties (CoT) inference analysis.
◦
Develop more robust AI models that include various types of lesions that can occur in real clinical settings.
•
Limitations:
◦
The size of the dataset may be relatively small compared to other large-scale medical imaging datasets.
◦
The subjectivity of the annotation and verification process by experts may influence the results.
◦
Further validation of the generalization performance of the dataset is needed.
👍
Made with Slashpage