Deep Feature Learning from a Hospital-Scale Chest X-ray Dataset with Application to TB Detection on a Small-Scale Dataset

Abstract

The use of ImageNet pre-trained networks is becoming widespread in the medical imaging community. It enables training on small datasets, commonly available in medical imaging tasks. The recent emergence of a large Chest X-ray dataset opened the possibility for learning features that are specific to the X-ray analysis task. In this work, we demonstrate that the features learned allow for better classification results for the problem of Tuberculosis detection and enable generalization to an unseen dataset.To accomplish the task of feature learning, we train a DenseNet-121 CNN on 112K images from the ChestXray14 dataset which includes labels of 14 common thoracic pathologies. In addition to the pathology labels, we incorporate metadata which is available in the dataset: Patient Positioning, Gender and Patient Age. We term this architecture MetaChexNet. As a by-product of the feature learning, we demonstrate state of the art performance on the task of patient Age & Gender estimation using CNN's. Finally, we show the features learned using ChestXray14 allow for better transfer learning on smallscale datasets for Tuberculosis.

Deep learning applications in pulmonary medical imaging: recent updates and insights on COVID-19

“…For stability, many approaches compared the performance of their models using different datasets whether for training, feature learning, validation, or all together. Examples of approaches in training, validation and all together are, respectively: Pan et al [62], Ho et al [133] and Gozes and Greenspan [166]. Some also experimented on different dataset sizes such as Shen et al [192].…”

“…On the same side, Gozes and Greenspan [166] targeted TB willing to study impact of feature learning (pre-training deep model) specifically on CXR using DenseNet-121 CNN. The application incorporated meta data and trained the model on ChestX-ray14 dataset which includes 14 thoracic pathologies.…”

“…Few contributions to chest were based on auto-encoders such as those of Xie et al [202] and Nemoto et al [163], while many were based on U-nets such as segmentation approaches of Ahn et al [88] and Furutani et al [103]. Likewise, other CNN architectures were used as FCNN by Dou et al [143], ResNet by Kim et al [171], and DenseNet by Gozes and Greenspan [166]. Moreover, combinations of architectures took place as by Gozes and Greenspan [164] where ResNet and DenseNet were integrated, and comparisons of multiple architectures were explicit as by Gozes and Greenspan [133], Ayan and Unver [131], Pan et al [62], and Wang et al [184].…”

Deep learning applied to automatic disease detection using chest X‐rays

“…Several groups added metadata to DL models, to further improve performance. Gozes 92 used a pretrained DenseNet‐121 CNN, retrained to extract both image features and metadata prediction which included binary classification of the patient’s position (AP vs. PA) and gender (M vs. F), and the patient’s age. Their model, termed MetaChexNet, was again retrained on the (small) Shenzen dataset in order to detect pulmonary TB, and it achieved an AUROC of 0.965.…”

Designing Chest X-ray Datasets for Improving Lung Nodules Detection Through Convolutional Neural Networks

“…ChestX-ray14 was used to train a specialized 121-layers CNN called CheXNet [2]. This model was mainly targeted at detecting pneumonia, but also performed well in the other diseases.…”