top of page
segmentation.png

Segmentation

Topology-informed segmentation

There are many cases in which the topology of a structure to be segmented is known a priori. For example, when segmenting the left ventricle of the heart from short-axis cine MR images, it is known that the correct segmentation will be a topological closed loop in two dimensions. However, traditional deep learning based segmentation models make no use of this knowledge. In this work, we explicitly incorporate topological priors into a deep learning segmentation model for the first time, based on the concept of persistent homology. These priors require no additional ground truth segmentations and so the model is suitable for use in a semi-supervised setting. We found that by including the topological priors the number of topological errors was reduced whilst maintaining high overlap with the ground truth. The initial work demonstrated this principle for single-class segmentation but we subsequently extended it to the multi-class case.

Publications

N. Byrne, J. R. Clough, I. Valverde, G. Montana, A. P. King, "A Persistent Homology-based Topological Loss for CNN-based Multi-class Segmentation of CMR", IEEE Transactions on Medical Imaging, 2022. (paper)
 

N. Byrne, J. R. Clough, G. Montana, A. P. King, "A Persistent Homology-based Topological Loss Function for Multi-class CNN Segmentation of Cardiac MRI", Proceedings MICCAI STACOM, 2020. (paper)


J. Clough, N. Byrne, I. Oksuz, V. A. Zimmer, J. A. Schnabel, A. P. King, "A Topological Loss Function for Deep-Learning based Image Segmentation using Persistent Homology", IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020. (paper)

J. R. Clough, I. Oksuz, N. Byrne, J. A. Schnabel, A. P. King, "Explicit Topological Priors for Deep-Learning Based Image Segmentation Using Persistent Homology", Proceedings IPMI, 2019. (paper)

segmentationbarcode.png

Persistence barcode for a predicted segmentation

segmentation_topo.png

Examples of a topology-informed loss function fixing topological errors in short axis cardiac MR segmentation

segmentationspleen.png

Examples of automated spleen segmentation and length measurement

Automated assessment of spleen size from ultrasound

Patients suffering from Sickle Cell Disease can suffer from a range of complications including abnormal enlargement of the spleen. Ultrasound forms part of the clinical pipeline for spleen assessment, but measuring its size in ultrasound images is a subjective process. Furthermore, in many parts of the developing world there is a lack of skills for performing reliable spleen assessments. We investigated the use of deep learning to automate this process, reaching almost human-level performance.

Publications

Z. Yuan, E. Puyol-Antón, H. Jogeesvaran, N. Smith, B. Inusa, A. P. King, "Deep Learning-based Quality-controlled Spleen Assessment From Ultrasound Images", Biomedical Signal Processing and Control, 76:103724, 2022. (paper)

 

Z. Yuan, E. Puyol-Antón, H. Jogeesvaran, C. Reid, B. Inusa, A. P. King, "Deep Learning for Automatic Spleen Length Measurement in Sickle Cell Disease Patients", Proceedings MICCAI ASMUS, 2020. (paper)

Segmentation of the Vocal Tract From Speech MRI

MR is increasingly playing a role in speech analyis - dynamic MR images of patients acquired during speech can be used for planning and monitoring of treatment for speech problems such as velopharyngeal insufficiency. In this project our ultimate aim is to construct dynamic 3D speech models of patients for informing treatment decisions. Initial work has investigated the use of deep learning for full segmentation of the articulators in the vocal tract from 2D MR, resulting in the first fully-automatic model demonstrated for this task.

Publications:

M. Ruthven, M. E. Miquel, A. P. King, "Deep-learning-based segmentation of the vocal tract and articulators in real-time magnetic resonance images of speech", Computer Methods and Programs in Biomedicine, 2020. (paper)

Automated segmentation of the vocal tract articulators from dynamic MR images: left - ground truth, middle - deep learning based segmentation, right - after post-processing.

bottom of page