XNet A convolutional neural network CNN implementation for medical XRay image segmentation suitable for small datasets

• Arnau Quera-Bofarull  |  Carolina Cuesta  |  Joseph Aylett-Bullock

• Bandyopadhyay et al.6  |  BIS National E-infrastructure  |  Hamilton HPC Service of Durham University  |  IBEX Innovations Ltd.32  |  Institute for Computational Cosmology  |  Kazeminia et al.4

XNet is a Convolutional Neural Network designed for the segmentation of X-Ray images into bone, soft tissue and open beam regions. Specifically, it performs well on small datasets with the aim to minimise the number of false positives in the soft tissue class. It uses an encoder decoder model which is an enhanced version of Segnet. It is benched marked against Unet and Segnet and tends to perform better than both of them with a weighted average f1 score of 0.92. Its a state of the art model which uses 4 augmentations. (Shearing, flipping, zooming and resizing).

Input variables : Normal Xray images
Output Variables : Annotated Xray Images

Visit Model : github.com