Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2312.16151 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917695775047680 |
|---|---|
| author | Zheng, Qiaoyu Zhao, Weike Wu, Chaoyi Zhang, Xiaoman Dai, Lisong Guan, Hengyu Li, Yuehua Zhang, Ya Wang, Yanfeng Xie, Weidi |
| author_facet | Zheng, Qiaoyu Zhao, Weike Wu, Chaoyi Zhang, Xiaoman Dai, Lisong Guan, Hengyu Li, Yuehua Zhang, Ya Wang, Yanfeng Xie, Weidi |
| contents | Developing a generalist radiology diagnosis system can greatly enhance clinical diagnostics. In this paper, we introduce RadDiag, a foundational model supporting 2D and 3D inputs across various modalities and anatomies, using a transformer-based fusion module for comprehensive disease diagnosis. Due to patient privacy concerns and the lack of large-scale radiology diagnosis datasets, we utilize high-quality, clinician-reviewed radiological images available online with diagnosis labels. Our dataset, RP3D-DiagDS, contains 40,936 cases with 195,010 scans covering 5,568 disorders (930 unique ICD-10-CM codes). Experimentally, our RadDiag achieves 95.14% AUC on internal evaluation with the knowledge-enhancement strategy. Additionally, RadDiag can be zero-shot applied or fine-tuned to external diagnosis datasets sourced from various hospitals, demonstrating state-of-the-art results. In conclusion, we show that publicly shared medical data on the Internet is a tremendous and valuable resource that can potentially support building a generalist AI for healthcare. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_16151 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Large-scale Long-tailed Disease Diagnosis on Radiology Images Zheng, Qiaoyu Zhao, Weike Wu, Chaoyi Zhang, Xiaoman Dai, Lisong Guan, Hengyu Li, Yuehua Zhang, Ya Wang, Yanfeng Xie, Weidi Computer Vision and Pattern Recognition Developing a generalist radiology diagnosis system can greatly enhance clinical diagnostics. In this paper, we introduce RadDiag, a foundational model supporting 2D and 3D inputs across various modalities and anatomies, using a transformer-based fusion module for comprehensive disease diagnosis. Due to patient privacy concerns and the lack of large-scale radiology diagnosis datasets, we utilize high-quality, clinician-reviewed radiological images available online with diagnosis labels. Our dataset, RP3D-DiagDS, contains 40,936 cases with 195,010 scans covering 5,568 disorders (930 unique ICD-10-CM codes). Experimentally, our RadDiag achieves 95.14% AUC on internal evaluation with the knowledge-enhancement strategy. Additionally, RadDiag can be zero-shot applied or fine-tuned to external diagnosis datasets sourced from various hospitals, demonstrating state-of-the-art results. In conclusion, we show that publicly shared medical data on the Internet is a tremendous and valuable resource that can potentially support building a generalist AI for healthcare. |
| title | Large-scale Long-tailed Disease Diagnosis on Radiology Images |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2312.16151 |