Learning to recognize Abnormalities in Chest X-Rays with Location-Aware Dense Networks

TL;DR

This paper introduces a location-aware Dense Network approach for chest X-ray abnormality detection, leveraging high-resolution images and spatial info, achieving state-of-the-art results on a large dataset.

Contribution

The novel DNetLoc model explicitly incorporates spatial information and high-resolution data, improving abnormality classification accuracy on large-scale chest X-ray datasets.

Findings

Achieved the best average AUC score on ChestX-Ray14 dataset.

Improved AUC scores when pathology location information is used.

Provided new patient-wise data splits for benchmarking.

Abstract

Chest X-ray is the most common medical imaging exam used to assess multiple pathologies. Automated algorithms and tools have the potential to support the reading workflow, improve efficiency, and reduce reading errors. With the availability of large scale data sets, several methods have been proposed to classify pathologies on chest X-ray images. However, most methods report performance based on random image based splitting, ignoring the high probability of the same patient appearing in both training and test set. In addition, most methods fail to explicitly incorporate the spatial information of abnormalities or utilize the high resolution images. We propose a novel approach based on location aware Dense Networks (DNetLoc), whereby we incorporate both high-resolution image data and spatial information for abnormality classification. We evaluate our method on the largest data set reported in the community, containing a total of 86,876 patients and 297,541 chest X-ray images. We achieve (i) the best average AUC score for published training and test splits on the single benchmarking data set (ChestX-Ray14), and (ii) improved AUC scores when the pathology location information is explicitly used. To foster future research we demonstrate the limitations of the current benchmarking setup and provide new reference patient-wise splits for the used data sets. This could support consistent and meaningful benchmarking of future methods on the largest publicly available data sets.