MURA Dataset: Towards Radiologist-Level Abnormality Detection in Musculoskeletal Radiographs
By Pranav Rajpurkar, Jeremy Irvin, Aarti Bagul, Daisy Yi Ding, Tony Duan, Hershel Mehta, Brandon Yang, Kaylie Zhu, Dillon Laird, Robyn L. Ball and 4 more. 2017
89 citations recorded by the source corpus.
We introduce MURA, a large dataset of musculoskeletal radiographs containing 40,895 images from 14,982 studies, where each study is manually labeled by radiologists as either normal or abnormal. On this dataset, we train a 169-layer densely connected convolutional network to detect and localize abnormalities. Six board-certified radiologists provide additional labels for a test set of 209 studies, on which we compare model and radiologist performance. We find that our model achieves performance comparable to that of radiologists. On finger, hand, and wrist studies, our model's F1 scores are slightly higher than those of radiologists, but not statistically significantly so; on elbow, forearm, humerus, and shoulder studies our model's F1 scores are slightly lower than those of radiologists, but also not statistically significantly so, indicating that the dataset presents a good challenge problem for future research. The dataset is freely available at this https URL
Subfields: Biomedical Engineering, Health Informatics, Radiology, Nuclear Medicine and Imaging
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Biomedical Engineering
In Health Informatics
In Radiology, Nuclear Medicine and Imaging
Citations in the corpus
- Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks (2018, unknown): “access to a large training set plays an important role for deep CNNs to achieve human-level performance (Esteva et al., 2017; Rajpurkar et al., 2017).”
- MURA: Large Dataset for Abnormality Detection in Musculoskeletal Radiographs. (2017, background): “…of diabetic retinopathy (Gulshan et al., 2016), skin cancer (Esteva et al., 2017), lymph node metastases (Bejnordi et al., 2017), heart arrhythmias (Rajpurkar et al., 2017a), brain hemorrhage (Grewal et al., 2017), pneumonia (Rajpurkar et al., 2017b), and hip fractures (Gale et al., 2017).”
- Deep Learning in the Wild (2018, background): “Figure 6 illustrates some human-interpretable representations of the inner workings of a CNN on the recently published MUsculoskeletal RAdiographs (MURA) dataset [45] that we use here as a proxy for the balloon dataset.”