Review of automatic segmentation methods of multiple sclerosis white matter lesions on conventional magnetic resonance imaging

By Daniel García-Lorenzo, Simon J. Francis, Sridar Narayanan, Douglas L. Arnold, D. Collins. 2013

374 citations recorded by the source corpus.

Magnetic resonance (MR) imaging is often used to characterize and quantify multiple sclerosis (MS) lesions in the brain and spinal cord. The number and volume of lesions have been used to evaluate MS disease burden, to track the progression of the disease and to evaluate the effect of new pharmaceuticals in clinical trials. Accurate identification of MS lesions in MR images is extremely difficult due to variability in lesion location, size and shape in addition to anatomical variability between subjects. Since manual segmentation requires expert knowledge, is time consuming and is subject to intra- and inter-expert variability, many methods have been proposed to automatically segment lesions. The objective of this study was to carry out a systematic review of the literature to evaluate the state of the art in automated multiple sclerosis lesion segmentation. From 1240 hits found initially with PubMed and Google scholar, our selection criteria identified 80 papers that described an automatic lesion segmentation procedure applied to MS. Only 47 of these included quantitative validation with at least one realistic image. In this paper, we describe the complexity of lesion segmentation, classify the automatic MS lesion segmentation methods found, and review the validation methods applied in each of the papers reviewed. Although many segmentation solutions have been proposed, including some with promising results using MRI data obtained on small groups of patients, no single method is widely employed due to performance issues related to the high variability of MS lesion appearance and differences in image acquisition. The challenge remains to provide segmentation techniques that work in all cases regardless of the type of MS, duration of the disease, or MRI protocol, and this within a comprehensive, standardized validation framework. MS lesion segmentation remains an open problem.

Subfields: Computer Vision and Pattern Recognition, Molecular Biology, Pathology and Forensic Medicine

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Molecular Biology

Ajith Abraham

Citations in the corpus

Current Applications and Future Impact of Machine Learning in Radiology. (2018, neutral): “Other potential use cases of machine learning include line detection (43), prostate cancer detection at MRI (44–46), determination of coronary artery calcium score (47), or detection and segmentation of brain lesions (48,49).”
3D fully convolutional networks for subcortical segmentation in MRI: A large‐scale study (2016, background): “Accurate segmentation of subcortical brain structures is crucial to the study of a breadth of brain disorders, e.g., schizophrenia [1], Parkinson [2], autism [3], multiple-sclerosis lesions [4, 5], and to the assessment of structural brain abnormalities [6].”
Deep 3D Convolutional Encoder Networks With Shortcuts for Multiscale Feature Integration Applied to Multiple Sclerosis Lesion Segmentation (2016, methodology): “We used the following four measures to produce a comprehensive evaluation of segmentation accuracy as there is generally no single measure that is sufficient to capture all information relevant to the quality of a produced segmentation [12].”
Improving automated multiple sclerosis lesion segmentation with a cascaded 3D convolutional neural network approach (2017, methodology): “Although expert manual annotations of lesions is feasible in practice, this task is both timeconsuming and prone to inter-observer variability, which has been led progressively to the development of a wide number of automated lesion segmentation techniques (Lladó et al., 2012; García-Lorenzo et al.,”
Objective Evaluation of Multiple Sclerosis Lesion Segmentation using a Data Management and Processing Infrastructure (2018, neutral): “To simplify the clinician’s task, a large literature of automatic segmentation methods has been devised [4, 5, 6] with a large spectrum of algorithms from classical tissue intensity classification and lesion modeling to machine learning.”
Longitudinal multiple sclerosis lesion segmentation: Resource and challenge (2017, neutral): “…et al., 2013; Welti et al., 2001; Xie and Tao, 2011) with several review articles available that describe and evaluate the utility of these methods (García-Lorenzo et al., 2013; Lladó et al., 2012), though semi-automated approaches have also been reported (Udupa et al., 1997; Wu et al., 2006;…”
State-of-the-Art Methods for Brain Tissue Segmentation: A Review (2017, background): “Phantoms can be utilized to generate real images, but it is difficult to have dense ground truth for phantom images [118].”
A survey of MRI-based brain tumor segmentation methods (2014, background): “D MRI data, the segmentation of these data has become a challenging problem [108,109] .”