Dataset Information

Comparing lesion segmentation methods in multiple sclerosis: Input from one manually delineated subject is sufficient for accurate lesion segmentation.

ABSTRACT:

Purpose

Accurate lesion segmentation is important for measurements of lesion load and atrophy in subjects with multiple sclerosis (MS). International MS lesion challenges show a preference of convolutional neural networks (CNN) strategies, such as nicMSlesions. However, since the software is trained on fairly homogenous training data, we aimed to test the performance of nicMSlesions in an independent dataset with manual and other automatic lesion segmentations to determine whether this method is suitable for larger, multi-center studies.

Methods

Manual lesion segmentation was performed in fourteen subjects with MS on sagittal 3D FLAIR images from a 3T GE whole-body scanner with 8-channel head coil. We compared five different categories of automated lesion segmentation methods for their volumetric and spatial agreement with manual segmentation: (i) unsupervised, untrained (LesionTOADS); (ii) supervised, untrained (LST-LPA and nicMSlesions with default settings); (iii) supervised, untrained with threshold adjustment (LST-LPA optimized for current data); (iv) supervised, trained with leave-one-out cross-validation on fourteen subjects with MS (nicMSlesions and BIANCA); and (v) supervised, trained on a single subject with MS (nicMSlesions). Volumetric accuracy was determined by the intra-class correlation coefficient (ICC) and spatial accuracy by Dice's similarity index (SI). Volumes and SI were compared between methods using repeated measures ANOVA or Friedman tests with post-hoc pairwise comparison.

Results

The best volumetric and spatial agreement with manual was obtained with the supervised and trained methods nicMSlesions and BIANCA (ICC absolute agreement >?0.968 and median SI?>?0.643) and the worst with the unsupervised, untrained method LesionTOADS (ICC absolute agreement?=?0.140 and median SI?=?0.444). Agreement with manual in the single-subject network training of nicMSlesions was poor for input with low lesion volumes (i.e. two subjects with lesion volumes ??3.0?ml). For the other twelve subjects, ICC varied from 0.593 to 0.973 and median SI varied from 0.535 to 0.606. In all cases, the single-subject trained nicMSlesions segmentations outperformed LesionTOADS, and in almost all cases it also outperformed LST-LPA.

Conclusion

Input from only one subject to re-train the deep learning CNN nicMSlesions is sufficient for adequate lesion segmentation, with on average higher volumetric and spatial agreement with manual than obtained with the untrained methods LesionTOADS and LST-LPA.

SUBMITTER: Weeda MM

PROVIDER: S-EPMC6861662 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Comparing lesion segmentation methods in multiple sclerosis: Input from one manually delineated subject is sufficient for accurate lesion segmentation.

Weeda M M MM Brouwer I I de Vos M L ML de Vries M S MS Barkhof F F Pouwels P J W PJW Vrenken H H

NeuroImage. Clinical 20191105

<h4>Purpose</h4>Accurate lesion segmentation is important for measurements of lesion load and atrophy in subjects with multiple sclerosis (MS). International MS lesion challenges show a preference of convolutional neural networks (CNN) strategies, such as nicMSlesions. However, since the software is trained on fairly homogenous training data, we aimed to test the performance of nicMSlesions in an independent dataset with manual and other automatic lesion segmentations to determine whether this m ...[more]

PMID: 31734527

Similar Datasets

Project description:Magnetic resonance imaging (MRI) can be used to detect lesions in the brains of multiple sclerosis (MS) patients and is essential for diagnosing the disease and monitoring its progression. In practice, lesion load is often quantified by either manual or semi-automated segmentation of MRI, which is time-consuming, costly, and associated with large inter- and intra-observer variability. We propose OASIS is Automated Statistical Inference for Segmentation (OASIS), an automated statistical method for segmenting MS lesions in MRI studies. We use logistic regression models incorporating multiple MRI modalities to estimate voxel-level probabilities of lesion presence. Intensity-normalized T1-weighted, T2-weighted, fluid-attenuated inversion recovery and proton density volumes from 131 MRI studies (98 MS subjects, 33 healthy subjects) with manual lesion segmentations were used to train and validate our model. Within this set, OASIS detected lesions with a partial area under the receiver operating characteristic curve for clinically relevant false positive rates of 1% and below of 0.59% (95% CI; [0.50%, 0.67%]) at the voxel level. An experienced MS neuroradiologist compared these segmentations to those produced by LesionTOADS, an image segmentation software that provides segmentation of both lesions and normal brain structures. For lesions, OASIS out-performed LesionTOADS in 74% (95% CI: [65%, 82%]) of cases for the 98 MS subjects. To further validate the method, we applied OASIS to 169 MRI studies acquired at a separate center. The neuroradiologist again compared the OASIS segmentations to those from LesionTOADS. For lesions, OASIS ranked higher than LesionTOADS in 77% (95% CI: [71%, 83%]) of cases. For a randomly selected subset of 50 of these studies, one additional radiologist and one neurologist also scored the images. Within this set, the neuroradiologist ranked OASIS higher than LesionTOADS in 76% (95% CI: [64%, 88%]) of cases, the neurologist 66% (95% CI: [52%, 78%]) and the radiologist 52% (95% CI: [38%, 66%]). OASIS obtains the estimated probability for each voxel to be part of a lesion by weighting each imaging modality with coefficient weights. These coefficients are explicit, obtained using standard model fitting techniques, and can be reused in other imaging studies. This fully automated method allows sensitive and specific detection of lesion presence and may be rapidly applied to large collections of images.

Project description:The recent global outbreak and spread of coronavirus disease (COVID-19) makes it an imperative to develop accurate and efficient diagnostic tools for the disease as medical resources are getting increasingly constrained. Artificial intelligence (AI)-aided tools have exhibited desirable potential; for example, chest computed tomography (CT) has been demonstrated to play a major role in the diagnosis and evaluation of COVID-19. However, developing a CT-based AI diagnostic system for the disease detection has faced considerable challenges, which is mainly due to the lack of adequate manually-delineated samples for training, as well as the requirement of sufficient sensitivity to subtle lesions in the early infection stages. In this study, we developed a dual-branch combination network (DCN) for COVID-19 diagnosis that can simultaneously achieve individual-level classification and lesion segmentation. To focus the classification branch more intensively on the lesion areas, a novel lesion attention module was developed to integrate the intermediate segmentation results. Furthermore, to manage the potential influence of different imaging parameters from individual facilities, a slice probability mapping method was proposed to learn the transformation from slice-level to individual-level classification. We conducted experiments on a large dataset of 1202 subjects from ten institutes in China. The results demonstrated that 1) the proposed DCN attained a classification accuracy of 96.74% on the internal dataset and 92.87% on the external validation dataset, thereby outperforming other models; 2) DCN obtained comparable performance with fewer samples and exhibited higher sensitivity, especially in subtle lesion detection; and 3) DCN provided good interpretability on the loci of infection compared to other deep models due to its classification guided by high-level semantic information. An online CT-based diagnostic platform for COVID-19 derived from our proposed framework is now available.

Project description:PurposeProstate cancer is multifocal. However, there often exists a single dominant focus in the gland responsible for driving the biology of the disease. Dose escalation to the dominant lesion is a proposed strategy to increase tumor control. We applied radiobiological modeling to evaluate the dosimetric feasibility and benefit of dominant intraprostatic lesion simultaneous in-field boosts (DIL-SIB) to the gross tumor volume (GTV), defined using a novel molecular positron emission tomography (PET) probe (18F-DCFPyL) directed against prostate specific membrane antigen (PSMA).Methods and materialsPatients with clinically localized, biopsy-proven prostate cancer underwent preoperative [18F]-DCFPyL PET/computed tomography (CT). DIL-SIB plans were generated by importing the PET/CT into the RayStation treatment planning system. GTV-PET for the DIL-SIB was defined by the highest %SUVmax (percentage of maximum standardized uptake value) that generated a biologically plausible volume. Volumetric arc-based plans incorporating prostate plus DIL-SIB treatment were generated. Tumor control probability (TCP) and normal tissue complication probability (NTCP) with fractionation schemes and boost doses specified in the FLAME (Investigate the Benefit of a Focal Lesion Ablative Microboost in Prostate Cancer; NCT01168479), PROFIT (Prostate Fractionated Irradiation Trial; NCT00304759), PACE (Prostate Advances in Comparative Evidence; NCT01584258), and hypoFLAME (Hypofractionated Focal Lesion Ablative Microboost in prostatE Cancer 2.0; NCT02853110) protocols were compared.ResultsComparative DIL-SIB plans for 6 men were generated from preoperative [18F]-DCFPyL PET/CT. Median boost GTV volume was 1.015 cm3 (0.42-1.83 cm3). Median minimum (D99%) DIL-SIB dose for F35BS, F20BS, F5BS, and F5BSH were 97.3 Gy, 80.8 Gy, 46.5 Gy, and 51.5Gy. TCP within the GTV ranged from 84% to 88% for the standard plan and 95% to 96% for the DIL-SIB plans. Within the rest of the prostate, TCP ranged from 89% to 91% for the standard plans and 90% to 92% for the DIL-SIB plans. NTCP for the rectum NTCP was similar for the DIL-SIB plans (0.3%-2.7%) compared with standard plans (0.7%-2.6%). Overall, DIL-SIB plans yielded higher uncomplicated TCP (NTCP, 90%-94%) versus standard plans (NTCP, 83%-85%).ConclusionsPSMA PET provides a novel approach to define GTV for SIB-DIL dose escalation. Work is ongoing to validate PSMA PET-delineated GTV through correlation to coregistered postprostatectomy digitized histopathology.

Project description:Background and purposeWM lesion segmentation is often performed with the use of subjective rating scales because manual methods are laborious and tedious; however, automated methods are now available. We compared the performance of total lesion volume grading computed by use of an automated WM lesion segmentation algorithm with that of subjective rating scales and expert manual segmentation in a cohort of subjects with type 2 diabetes.Materials and methodsStructural T1 and FLAIR MR imaging data from 50 subjects with diabetes (age, 67.7 ± 7.2 years) and 50 nondiabetic sibling pairs (age, 67.5 ± 9.4 years) were evaluated in an institutional review board-approved study. WM lesion segmentation maps and total lesion volume were generated for each subject by means of the Statistical Parametric Mapping (SPM8) Lesion Segmentation Toolbox. Subjective WM lesion grade was determined by means of a 0-9 rating scale by 2 readers. Ground-truth total lesion volume was determined by means of manual segmentation by experienced readers. Correlation analyses compared manual segmentation total lesion volume with automated and subjective evaluation methods.ResultsCorrelation between average lesion segmentation and ground-truth total lesion volume was 0.84. Maximum correlation between the Lesion Segmentation Toolbox and ground-truth total lesion volume (ρ = 0.87) occurred at the segmentation threshold of k = 0.25, whereas maximum correlation between subjective lesion segmentation and the Lesion Segmentation Toolbox (ρ = 0.73) occurred at k = 0.15. The difference between the 2 correlation estimates with ground-truth was not statistically significant. The lower segmentation threshold (0.15 versus 0.25) suggests that subjective raters overestimate WM lesion burden.ConclusionsWe validate the Lesion Segmentation Toolbox for determining total lesion volume in diabetes-enriched populations and compare it with a common subjective WM lesion rating scale. The Lesion Segmentation Toolbox is a readily available substitute for subjective WM lesion scoring in studies of diabetes and other populations with changes of leukoaraiosis.

Dataset Information

Comparing lesion segmentation methods in multiple sclerosis: Input from one manually delineated subject is sufficient for accurate lesion segmentation.

Purpose

Methods

Results

Conclusion

Publications

Comparing lesion segmentation methods in multiple sclerosis: Input from one manually delineated subject is sufficient for accurate lesion segmentation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets