Limits...
Relating multi-sequence longitudinal intensity profiles and clinical covariates in incident multiple sclerosis lesions.

Sweeney EM, Shinohara RT, Dewey BE, Schindler MK, Muschelli J, Reich DS, Crainiceanu CM, Eloyan A - Neuroimage Clin (2015)

Bottom Line: The proposed biomarker's ability to identify such effects is validated by two experienced clinicians (a neuroradiologist and a neurologist).We then relate the biomarker to the clinical information in a mixed model framework.Treatment with disease-modifying therapies (p < 0.01), steroids (p < 0.01), and being closer to the boundary of abnormal signal intensity (p < 0.01) are all associated with return of a voxel to an intensity value closer to that of normal-appearing tissue.

View Article: PubMed Central - PubMed

Affiliation: Department of Biostatistics, The Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD 21205, United States ; Translational Neuroradiology Unit, Division of Neuroimmunology and Neurovirology, National Institute of Neurological Disease and Stroke, National Institute of Health, Bethesda, MD 20892, United States.

ABSTRACT
The formation of multiple sclerosis (MS) lesions is a complex process involving inflammation, tissue damage, and tissue repair - all of which are visible on structural magnetic resonance imaging (MRI) and potentially modifiable by pharmacological therapy. In this paper, we introduce two statistical models for relating voxel-level, longitudinal, multi-sequence structural MRI intensities within MS lesions to clinical information and therapeutic interventions: (1) a principal component analysis (PCA) and regression model and (2) function-on-scalar regression models. To do so, we first characterize the post-lesion incidence repair process on longitudinal, multi-sequence structural MRI from 34 MS patients as voxel-level intensity profiles. For the PCA regression model, we perform PCA on the intensity profiles to develop a voxel-level biomarker for identifying slow and persistent, long-term intensity changes within lesion tissue voxels. The proposed biomarker's ability to identify such effects is validated by two experienced clinicians (a neuroradiologist and a neurologist). On a scale of 1 to 4, with 4 being the highest quality, the neuroradiologist gave the score on the first PC a median quality rating of 4 (95% CI: [4,4]), and the neurologist gave the score a median rating of 3 (95% CI: [3,3]). We then relate the biomarker to the clinical information in a mixed model framework. Treatment with disease-modifying therapies (p < 0.01), steroids (p < 0.01), and being closer to the boundary of abnormal signal intensity (p < 0.01) are all associated with return of a voxel to an intensity value closer to that of normal-appearing tissue. The function-on-scalar regression model allows for assessment of the post-incidence time points at which the covariates are associated with the profiles. In the function-on-scalar regression, both age and distance to the boundary were found to have a statistically significant association with the lesion intensities at some time point. The two models presented in this article show promise for understanding the mechanisms of tissue damage in MS and for evaluating the impact of treatments for the disease in clinical trials.

No MeSH data available.


Related in: MedlinePlus

SuBLIME and OASIS segmentations. Each column of the figure represents a different MRI study, starting at 98 days after baseline in the far left column and going until 343 days after baseline. A lesion is first identified in this area at 175 days. The first four rows show the longitudinal behavior of the FLAIR, T2, PD, and T1 sequences. The next rows show the SuBLIME segmentation of lesion incidence for each study and the OASIS segmentation of lesion presence in each study. The SuBLIME segmentation has been further divided into areas of edema and lesion.
© Copyright Policy - CC BY-NC-ND
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4660378&req=5

f0040: SuBLIME and OASIS segmentations. Each column of the figure represents a different MRI study, starting at 98 days after baseline in the far left column and going until 343 days after baseline. A lesion is first identified in this area at 175 days. The first four rows show the longitudinal behavior of the FLAIR, T2, PD, and T1 sequences. The next rows show the SuBLIME segmentation of lesion incidence for each study and the OASIS segmentation of lesion presence in each study. The SuBLIME segmentation has been further divided into areas of edema and lesion.

Mentions: SuBLIME segmentation of voxel-level lesion incidence and enlargement is a method for detecting voxels that are part of an area of new abnormal signal intensity between two MRI studies (Sweeney et al., 2013a). For each subject, we produce SuBLIME maps between the respective sets of consecutive MRI studies. We exclude all abnormal signal intensity areas that contained fewer than 27 voxels, as these areas could be artifact or noise. We then produce cross-sectional lesion segmentations using OASIS segmentation of abnormal signal presence (Sweeney et al., 2013b). As the signal from edema disappears rapidly from the MRI after lesion formation, we locate the incident abnormal signal voxels using SuBLIME, but only include the voxels that are detected by OASIS at the following study visit, as these voxels should not contain edema. Therefore, only voxels that have an MRI study within 40 days after SuBLIME detects the area of abnormal signal intensity, where the intensity remains in the OASIS maps, are considered as lesion tissue and used in this analysis, as by this time edema would subside. We use expert validation by a neuroradiologist and a neurologist, both with experience in MS imaging, to confirm that this method is identifying lesion tissue, which we describe in detail in the subsection Expert Validation. The figure below shows the SuBLIME segmentation for each study and the OASIS segmentation for each study, corresponding to Fig. 2 from the paper. The row corresponding to the SuBLIME segmentation is further divided into edema and lesion voxels using the method described above. Only voxels that are part of lesion tissue are used in the analysis (Fig. 8Fig. 8


Relating multi-sequence longitudinal intensity profiles and clinical covariates in incident multiple sclerosis lesions.

Sweeney EM, Shinohara RT, Dewey BE, Schindler MK, Muschelli J, Reich DS, Crainiceanu CM, Eloyan A - Neuroimage Clin (2015)

SuBLIME and OASIS segmentations. Each column of the figure represents a different MRI study, starting at 98 days after baseline in the far left column and going until 343 days after baseline. A lesion is first identified in this area at 175 days. The first four rows show the longitudinal behavior of the FLAIR, T2, PD, and T1 sequences. The next rows show the SuBLIME segmentation of lesion incidence for each study and the OASIS segmentation of lesion presence in each study. The SuBLIME segmentation has been further divided into areas of edema and lesion.
© Copyright Policy - CC BY-NC-ND
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4660378&req=5

f0040: SuBLIME and OASIS segmentations. Each column of the figure represents a different MRI study, starting at 98 days after baseline in the far left column and going until 343 days after baseline. A lesion is first identified in this area at 175 days. The first four rows show the longitudinal behavior of the FLAIR, T2, PD, and T1 sequences. The next rows show the SuBLIME segmentation of lesion incidence for each study and the OASIS segmentation of lesion presence in each study. The SuBLIME segmentation has been further divided into areas of edema and lesion.
Mentions: SuBLIME segmentation of voxel-level lesion incidence and enlargement is a method for detecting voxels that are part of an area of new abnormal signal intensity between two MRI studies (Sweeney et al., 2013a). For each subject, we produce SuBLIME maps between the respective sets of consecutive MRI studies. We exclude all abnormal signal intensity areas that contained fewer than 27 voxels, as these areas could be artifact or noise. We then produce cross-sectional lesion segmentations using OASIS segmentation of abnormal signal presence (Sweeney et al., 2013b). As the signal from edema disappears rapidly from the MRI after lesion formation, we locate the incident abnormal signal voxels using SuBLIME, but only include the voxels that are detected by OASIS at the following study visit, as these voxels should not contain edema. Therefore, only voxels that have an MRI study within 40 days after SuBLIME detects the area of abnormal signal intensity, where the intensity remains in the OASIS maps, are considered as lesion tissue and used in this analysis, as by this time edema would subside. We use expert validation by a neuroradiologist and a neurologist, both with experience in MS imaging, to confirm that this method is identifying lesion tissue, which we describe in detail in the subsection Expert Validation. The figure below shows the SuBLIME segmentation for each study and the OASIS segmentation for each study, corresponding to Fig. 2 from the paper. The row corresponding to the SuBLIME segmentation is further divided into edema and lesion voxels using the method described above. Only voxels that are part of lesion tissue are used in the analysis (Fig. 8Fig. 8

Bottom Line: The proposed biomarker's ability to identify such effects is validated by two experienced clinicians (a neuroradiologist and a neurologist).We then relate the biomarker to the clinical information in a mixed model framework.Treatment with disease-modifying therapies (p < 0.01), steroids (p < 0.01), and being closer to the boundary of abnormal signal intensity (p < 0.01) are all associated with return of a voxel to an intensity value closer to that of normal-appearing tissue.

View Article: PubMed Central - PubMed

Affiliation: Department of Biostatistics, The Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD 21205, United States ; Translational Neuroradiology Unit, Division of Neuroimmunology and Neurovirology, National Institute of Neurological Disease and Stroke, National Institute of Health, Bethesda, MD 20892, United States.

ABSTRACT
The formation of multiple sclerosis (MS) lesions is a complex process involving inflammation, tissue damage, and tissue repair - all of which are visible on structural magnetic resonance imaging (MRI) and potentially modifiable by pharmacological therapy. In this paper, we introduce two statistical models for relating voxel-level, longitudinal, multi-sequence structural MRI intensities within MS lesions to clinical information and therapeutic interventions: (1) a principal component analysis (PCA) and regression model and (2) function-on-scalar regression models. To do so, we first characterize the post-lesion incidence repair process on longitudinal, multi-sequence structural MRI from 34 MS patients as voxel-level intensity profiles. For the PCA regression model, we perform PCA on the intensity profiles to develop a voxel-level biomarker for identifying slow and persistent, long-term intensity changes within lesion tissue voxels. The proposed biomarker's ability to identify such effects is validated by two experienced clinicians (a neuroradiologist and a neurologist). On a scale of 1 to 4, with 4 being the highest quality, the neuroradiologist gave the score on the first PC a median quality rating of 4 (95% CI: [4,4]), and the neurologist gave the score a median rating of 3 (95% CI: [3,3]). We then relate the biomarker to the clinical information in a mixed model framework. Treatment with disease-modifying therapies (p < 0.01), steroids (p < 0.01), and being closer to the boundary of abnormal signal intensity (p < 0.01) are all associated with return of a voxel to an intensity value closer to that of normal-appearing tissue. The function-on-scalar regression model allows for assessment of the post-incidence time points at which the covariates are associated with the profiles. In the function-on-scalar regression, both age and distance to the boundary were found to have a statistically significant association with the lesion intensities at some time point. The two models presented in this article show promise for understanding the mechanisms of tissue damage in MS and for evaluating the impact of treatments for the disease in clinical trials.

No MeSH data available.


Related in: MedlinePlus