2D/3D Megavoltage Image Registration Using Convolutional Neural Networks
This addresses image registration challenges in medical imaging for megavoltage applications, but it appears incremental as it applies an existing neural network method to a specific domain problem.
The paper tackled the problem of 2D/3D megavoltage image registration, which suffers from low image quality and poor capture range with traditional intensity-based methods, by proposing a Convolutional Neural Network approach that showed promising results on a dataset of 50 brain images.
We presented a 2D/3D MV image registration method based on a Convolutional Neural Network. Most of the traditional image registration method intensity-based, which use optimization algorithms to maximize the similarity between to images. Although these methods can achieve good results for kilovoltage images, the same does not occur for megavoltage images due to the lower image quality. Also, these methods most of the times do not present a good capture range. To deal with this problem, we propose the use of Convolutional Neural Network. The experiments were performed using a dataset of 50 brain images. The results showed to be promising compared to traditional image registration methods.