Rethinking annotation granularity for overcoming deep shortcut learning: A retrospective study on chest radiographs

by   Luyang Luo, et al.

Deep learning has demonstrated radiograph screening performances that are comparable or superior to radiologists. However, recent studies show that deep models for thoracic disease classification usually show degraded performance when applied to external data. Such phenomena can be categorized into shortcut learning, where the deep models learn unintended decision rules that can fit the identically distributed training and test set but fail to generalize to other distributions. A natural way to alleviate this defect is explicitly indicating the lesions and focusing the model on learning the intended features. In this paper, we conduct extensive retrospective experiments to compare a popular thoracic disease classification model, CheXNet, and a thoracic lesion detection model, CheXDet. We first showed that the two models achieved similar image-level classification performance on the internal test set with no significant differences under many scenarios. Meanwhile, we found incorporating external training data even led to performance degradation for CheXNet. Then, we compared the models' internal performance on the lesion localization task and showed that CheXDet achieved significantly better performance than CheXNet even when given 80 visualizing the models' decision-making regions, we revealed that CheXNet learned patterns other than the target lesions, demonstrating its shortcut learning defect. Moreover, CheXDet achieved significantly better external performance than CheXNet on both the image-level classification task and the lesion localization task. Our findings suggest improving annotation granularity for training deep learning systems as a promising way to elevate future deep learning-based diagnosis systems for clinical usage.


page 3

page 5

page 6

page 14


Deep Mining External Imperfect Data for Chest X-ray Disease Screening

Deep learning approaches have demonstrated remarkable progress in automa...

Deep Learning-Based Grading of Ductal Carcinoma In Situ in Breast Histopathology Images

Ductal carcinoma in situ (DCIS) is a non-invasive breast cancer that can...

OXnet: Omni-supervised Thoracic Disease Detection from Chest X-rays

Chest X-ray (CXR) is the most typical medical image worldwide to examine...

An Accurate and Explainable Deep Learning System Improves Interobserver Agreement in the Interpretation of Chest Radiograph

Recent artificial intelligence (AI) algorithms have achieved radiologist...

Deep image mining for diabetic retinopathy screening

Deep learning is quickly becoming the leading methodology for medical im...

COVCOR20 at WNUT-2020 Task 2: An Attempt to Combine Deep Learning and Expert rules

In the scope of WNUT-2020 Task 2, we developed various text classificati...

Deep ensemble learning for segmenting tuberculosis-consistent manifestations in chest radiographs

Automated segmentation of tuberculosis (TB)-consistent lesions in chest ...