Automatic MR Spinal Cord Segmentation by Hybrid Residual Attention-Aware Convolutional Neural Networks and Learning Rate Optimization on Real World Data

Gómez, A. Bueno; Bosch, Ignacio; Polo, J. Carreres

doi:10.1007/978-3-030-64610-3_19

Cited by 1 publication

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further to a proof of concept study developed with a preliminary methodology in a reduced cohort [28], in this extended work we aim to present a novel approach for automated cervical spinal cord segmentation from MR, based on hybrid residual attention-aware mechanisms that, together with a focal loss function [29] with the Tversky-index [30] as main metric, addresses the problems of unbalanced annotation and MR imaging preprocessing. We evaluate the performance of this methodology by its application to a dataset of real-world MR images (3D-T1 weighted images acquired in a 3T system) acquired in patients suffering from MS.…”

Section: Introductionmentioning

confidence: 99%

Automated Cervical Spinal Cord Segmentation in Real-World MRI of Multiple Sclerosis Patients by Optimized Hybrid Residual Attention-Aware Convolutional Neural Networks

et al. 2022

Self Cite

View full text Add to dashboard Cite

Magnetic resonance (MR) imaging is the most sensitive clinical tool in the diagnosis and monitoring of multiple sclerosis (MS) alterations. Spinal cord evaluation has gained interest in this clinical scenario in recent years but, unlike the brain, there is a more limited choice of algorithms to assist spinal cord segmentation. Our goal was to investigate and develop an automatic MR cervical cord segmentation method, enabling automated and seamless spinal cord atrophy assessment and setting the stage for the development of an aggregated algorithm for the extraction of lesion-related imaging biomarkers. The algorithm was developed using a real-world MR imaging dataset of 121 MS patients (96 cases used as a training dataset and 25 cases as a validation dataset). Transversal, 3D-T1 weighted gradient echo MR images (TE/TR/FA=1.7-2.7ms/5.6-8.2ms/12°) were acquired in a 3T system (SignaHD, GEHC) as standard of care in our clinical practice. Experienced radiologists supervised the manual labelling, which was considered the ground-truth. The 2D convolutional neural network consisted of a hybrid residual attentionaware segmentation method trained to delineate the cervical spinal cord. The training was conducted using a focal loss function, based on the Tversky index to address label imbalance, and an automatic optimal learning rate finder. Our automated model provided an accurate segmentation, achieving a validation DICE coefficient of 0.904±0.101 compared with the manual delineation. An automatic method for cervical spinal cord segmentation on T1-weighted MR images was successfully implemented. It will have direct implications serving as the first step for accelerating the process for MS staging and follow-up through imaging biomarkers.

show abstract

Section: Introductionmentioning

confidence: 99%