LandmarkNet: a 2D digital radiograph landmark estimator for registration

Cong Liu,Zhen Wang,Longhua Ma

doi:10.1186/s12911-020-01164-4

Abstract

BackgroundRadiation therapy requires precision to target and escalate the doses to affected regions while reducing the adjacent normal tissue exposed to high radiotherapy doses. Image guidance has become the start of the art in the treating process. Registering the digital radiographs megavoltage x ray (MV-DRs) and the kilovoltage digital reconstructed radiographs (KV-DRRs) is difficult because of the poor quality of MV-DRs. We simplify the problem by registering between landmarks instead of entire image information, thence we propose a model to estimate the landmark accurately.MethodsAfter doctors’ analysis, it is proved that it is effective to register through several physiological features such as spinous process, tracheal bifurcation, Louis angle. We propose the LandmarkNet, a novel keypoint estimation architecture, can automatically detect keypoints in blurred medical images. The method applies the idea of Feature Pyramid Network (FPN) twice to merge the cross-scale and cross-layer features for feature extraction and landmark estimation successively. Intermediate supervision is used at the end of the first FPN to ensure that the underlying parameters are updated normally. The network finally produces heatmap to display the approximate location of landmarks and we obtain accurate position estimation after non-maximum suppression (NMS) processing.ResultsOur method could obtain accurate landmark estimation in the dataset provided by several cancer hospitals and labeled by ourselves. The standard percentage of correct keypoints (PCK) within 8 pixels of estimation for the spinous process, tracheal bifurcation and Louis angle is 81.24%, 98.95% and 85.61% respectively. For the above three landmarks, the mean deviation between the predicted location of each landmark and corresponding ground truth is 2.38, 0.98 and 2.64 pixels respectively.ConclusionLandmark estimation based on LandmarkNet has high accuracy for different kinds of landmarks. Our model estimates the location of tracheal bifurcation especially accurately because of its obvious features. For the spinous process, our model performs well in quantity estimation as well as in position estimation. The wide application of our method assists doctors in image-guided radiotherapy (IGRT) and provides the possibility of precise treatment in the true sense.

Highlights

Radiation therapy requires precision to target and escalate the doses to affected regions while reducing the adjacent normal tissue exposed to high radiotherapy doses
In the process of image-guided radiotherapy, it is usually registered by MVDRs (Digital Radiography, generated by mega-level X-rays through the human body on the Electronic Portal Imaging Device) and kilovoltage digital reconstructed radiographs (KV-DRRs) (Digitally Reconstructed Radiography, re-projected from computed tomography of kilovolt X-ray), so that the treatment position is aligned with the planned position for precise radiotherapy
The landmarks in the image are extracted, and the registration of the two images is achieved by landmark alignment, which reduces the deviation caused by the different postures of the patient at different time periods, avoids the interference of the non-interest area on the attention area

Summary

Introduction

Radiation therapy requires precision to target and escalate the doses to affected regions while reducing the adjacent normal tissue exposed to high radiotherapy doses. In the process of image-guided radiotherapy, it is usually registered by MVDRs (Digital Radiography, generated by mega-level X-rays through the human body on the Electronic Portal Imaging Device) and KV-DRRs (Digitally Reconstructed Radiography, re-projected from computed tomography of kilovolt X-ray), so that the treatment position is aligned with the planned position for precise radiotherapy. The registration of the two images is very technically demanding for the physician because of the poor quality of the MV-DRs. In this paper, the landmarks in the image (spinous process, tracheal bifurcation, etc.) are extracted, and the registration of the two images is achieved by landmark alignment, which reduces the deviation caused by the different postures of the patient at different time periods, avoids the interference of the non-interest area on the attention area. Estimation of key points has many applications in pulmonary nodules, The work by Shi et al [3] input segmented lung image to conventional neural networks for extract the feature of pulmonary nodules and adopted position-sensitive score maps to represent the location information of lung nodules

Methods

Results

Discussion

Conclusion