Multi-Task Lung Nodule Detection in Chest Radiographs with a Dual Head Network

TL;DR

This paper introduces a multi-task deep learning algorithm with a dual head network for improved lung nodule detection in chest radiographs, combining global and local predictions to enhance accuracy and robustness.

Contribution

The study proposes a novel dual head network architecture with a specialized augmentation strategy for simultaneous global and local nodule detection in chest X-rays.

Findings

Enhanced nodule detection performance compared to traditional methods

Effective multi-task learning with dual head architecture

Significant improvement using the proposed augmentation strategy

Abstract

Lung nodules can be an alarming precursor to potential lung cancer. Missed nodule detections during chest radiograph analysis remains a common challenge among thoracic radiologists. In this work, we present a multi-task lung nodule detection algorithm for chest radiograph analysis. Unlike past approaches, our algorithm predicts a global-level label indicating nodule presence along with local-level labels predicting nodule locations using a Dual Head Network (DHN). We demonstrate the favorable nodule detection performance that our multi-task formulation yields in comparison to conventional methods. In addition, we introduce a novel Dual Head Augmentation (DHA) strategy tailored for DHN, and we demonstrate its significance in further enhancing global and local nodule predictions.