Share this article on:

Preliminary structural MRI based brain classification of chronic pelvic pain: A MAPP network study

Bagarinao, Epifanioa; Johnson, Kevin A.a; Martucci, Katherine T.a; Ichesco, Ericb; Farmer, Melissa A.c; Labus, Jenniferd; Ness, Timothy J.f; Harris, Richardb; Deutsch, George; Apkarian, Vania A.c; Mayer, Emeran A.d; Clauw, Daniel J.b; Mackey, Seana,*

doi: 10.1016/j.pain.2014.09.002
Article

Summary A preliminary classifier of brain structure was identified in chronic pelvic pain using a support vector machine learning algorithm suggesting distributed regional gray matter increases.

ABSTRACT Neuroimaging studies have shown that changes in brain morphology often accompany chronic pain conditions. However, brain biomarkers that are sensitive and specific to chronic pelvic pain (CPP) have not yet been adequately identified. Using data from the Trans-MAPP Research Network, we examined the changes in brain morphology associated with CPP. We used a multivariate pattern classification approach to detect these changes and to identify patterns that could be used to distinguish participants with CPP from age-matched healthy controls. In particular, we used a linear support vector machine (SVM) algorithm to differentiate gray matter images from the 2 groups. Regions of positive SVM weight included several regions within the primary somatosensory cortex, pre-supplementary motor area, hippocampus, and amygdala were identified as important drivers of the classification with 73% overall accuracy. Thus, we have identified a preliminary classifier based on brain structure that is able to predict the presence of CPP with a good degree of predictive power. Our regional findings suggest that in individuals with CPP, greater gray matter density may be found in the identified distributed brain regions, which are consistent with some previous investigations in visceral pain syndromes. Future studies are needed to improve upon our identified preliminary classifier with integration of additional variables and to assess whether the observed differences in brain structure are unique to CPP or generalizable to other chronic pain conditions.

a Department of Anesthesiology, Perioperative and Pain Medicine, Division of Pain Medicine, Stanford University Medical Center, Stanford, CA, USA

b Department of Anesthesiology, Chronic Pain and Fatigue Research Center, University of Michigan, Ann Arbor, MI, USA

c Department of Physiology, Northwestern University, Feinberg School of Medicine, Chicago, IL, USA

d Gail and Gerald Oppenheimer Family Center for Neurobiology of Stress, Pain and Interoception Network (PAIN), David Geffen School of Medicine at UCLA, Los Angeles, CA, USA

e Department of Radiology, University of Alabama, Birmingham Medical Center, Birmingham, AL, USA

f Department of Anesthesiology, University of Alabama, Birmingham Medical Center, Birmingham, AL, USA

*Corresponding author at: Departments of Anesthesiology, Perioperative and Pain Medicine, Neurosciences, and Neurology, Stanford University School of Medicine, 1070 Arastradero Rd, Room 285, MC 5596, Palo Alto, CA 94304-1336, USA. Tel.: +1 (650) 498 6477; fax: +1 (650) 725 9642.

E-mail: smackey@stanford.edu

Article history: Received February 25, 2014; Received in revised form August 19, 2014; Accepted September 3, 2014.

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.

Back to Top | Article Outline

1. Introduction

Chronic pelvic pain (CPP) affects approximately 3% to 6% of women in the United States, affects quality of life, reduces work productivity, and adds financial burden to the health care system [9,30]. Individuals with CPP typically experience a combination of both pain (eg, constant and/or evoked) without specific etiology (eg, no observable pathology) localized in the pelvic region, and urinary symptoms (eg, urgency and increased frequency) by definition lasting for at least 3months. Similar to other chronic pain conditions, little is known about the mechanisms underlying CPP, and its pathogenesis is still not fully understood [30].

To counter the great lack of understanding and knowledge pertaining to the etiology and prognosis of CPP, a multisite study, the U.S. NIH MAPP Research Network, was formed (http://www.mappnetwork.org). Because of the lack of effective biomarkers for CPP, the Neuroimaging Working Group of the MAPP Research Network was formed with a goal of developing neuroimaging-based biomarkers for CPP.

Recent neuroimaging studies have shown that changes in brain morphology often accompany chronic pain conditions [3,4,7,15,31,49,56] especially affecting areas related to pain perception and modulation. Historically, these changes in regional gray matter (GM) volume have been detected using mass univariate methods such as voxel-based morphometry (VBM), where differences in group intensities are tested for statistical significance for each volume element (voxel) (reviewed in [5]). However, these differences in GM may ultimately reflect a distributed network of altered brain regions, and univariate techniques are often not able to capture the specific distributed spatial patterns. Newer approaches using multivariate pattern analysis (MVPA) are able to characterize the spatially distributed and covarying patterns of brain activity and structure and are yielding rich information about underlying conditions (reviewed in [38]). In addition to providing greater mechanistic information, MVPA approaches provide classification metrics such as accuracy, sensitivity, and specificity that can be useful in translating the results into a clinically useful biomarker.

One form of the MVPA approach uses support vector machine (SVM). SVM algorithms are effective in discriminating patterns of brain activity and structure that occur in neurodegenerative diseases and mental disorders [8,18]. We [11] and others [55] have recently demonstrated an SVM algorithm’s effectiveness to classify acute pain. We have also used an SVM algorithm to classify brain images from participants with chronic low back pain from that of healthy controls with high classification accuracy of 75%. Others have classified chronic low back pain using measures of brain functional connectivity [2,52].

As part of the MAPP Research Network, we captured a large set of structural magnetic resonance imaging (MRI) data from female participants with CPP and age-matched, female healthy controls. Our overall goals in this study were to identify a multivariate structural brain classifier coinciding with CPP using an SVM algorithm, and to assess the accuracy of this brain classifier in discriminating CPP from healthy controls. We also investigated relationships between brain structure in CPP vs symptom duration, symptom severity, and psychological measures.

Back to Top | Article Outline

2. Materials and methods

2.1. Participants

Data from the MAPP Research Network neuroimaging study were used [14]. These data were collected through a uniform and collaborative effort across multiple neuroimaging sites including Stanford University, University of California Los Angeles, University of Alabama at Birmingham, University of Michigan, Northwestern University, and University of Washington. All study participants provided signed written consent indicating that they understood and were willing to undergo the procedures of the study, which had been approved in accordance with the institutional review board at the participant’s site of involvement in the study.

The data set consisted of structural images from 33 female participants with CPP and 33 age- and gender-matched healthy controls. All participants were extensively characterized as part of the Trans-MAPP Research Network Protocol, which included an initial 4h visit (eligibility screening, 2 urine samples, a blood draw, and extensive online module questionnaires to provide phenotype data), and an MRI scan visit scheduled within 24h of the initial baseline visit [28]. Participants with CPP were also asked to continue with biweekly and bimonthly questionnaires, then return for a 6-month and 12-month in-person follow-up visit. The average (±standard deviation) age for the patient group was 39.51 (±12.09) years, while for the healthy control group was 38.95 (±11.64)years. For the patient group, the average symptom duration was 9.09 (±9.79)years, with the longest duration being 32.11years and the shortest 0.57years. Participants were included on the basis of having no other major somatic symptoms (eg, fibromyalgia, irritable bowel syndrome, chronic fatigue syndrome). However, of the participants included, 7 reported symptoms of vulvodynia and 8 reported symptoms of temporomandibular disorders. In terms of medications used, 6 participants were receiving opioid medications, 7 were receiving centrally acting pain medications (antidepressants, benzodiazepines, anticonvulsants, immunosuppressant drugs), 11 were receiving peripherally acting pain medications (eg, nonsteroidal anti-inflammatory drugs), and 9 participants were receiving no medications for pain.

Back to Top | Article Outline

2.2. Questionnaires

Several questionnaires were administered before the MRI scan session, including the McGill Pain Questionnaire (MPQ) and the Hospital Anxiety and Depression Scale (HADS). From the MPQ, a total sensory score was used to assess the correlation of pain intensity with GM density. From the HADS, anxiety and depression total scores were used to assess the correlation of psychological state with GM density.

Back to Top | Article Outline

2.3. Image acquisition

Trans-MAPP neuroimaging data was collected, quality controlled and archived according to multi site imaging procedures developed collaboratively between the MAPP Research Network, the UCLA PAIN repository, and the UCLA Laboratory of Neuroimaging. Detailed procedures and description of the repository are available elsewhere (http://painrepository.org/). Scanner-compatible acquisition parameters were developed, and all sites were required to complete and pass a site qualification including a set of pilot scans of a human volunteer; the initial scans were reviewed for quality control by the UCLA site, and recommendations and adjustments were made as necessary before commencement of study scans. After quality control approval, the same MRI scanner that was used for site qualification was used to acquire all subsequent subject scans at each site, along with the same 8-channel head coil throughout the duration of the study. Scanner upgrades required requalification of the scanners.

All structural T1-weighted images were acquired using a 3D MP-RAGE pulse sequence (Magnetization Prepared Rapid Acquisition Gradient Echo; Siemens) [35] or the equivalent FSPGR IR pulse sequence (Fast SPoiled GRadient Echo—Inversion Recovery; General Electric [GE]). The specific standardized parameters were as follows: TR=2200ms, TE=minimum, TI=750ms, flip angle=20 degrees, FoV=220×220mm, resolution=256×256, slices per volume=176, slice thickness=1mm, voxel size=0.86×0.86×1mm, slice selective inversion, phased array acceleration factor=2, phase encode direction=left–right and superior–inferior, orientation=axial–oblique parallel to the line between the anterior and posterior commissures.

Structural images were uploaded to a centralized database at UCLA’s PAIN repository, where they were run through a thorough automated quality control pipeline; scans that passed quality control standards were then available for download by any of the Trans-MAPP neuroimaging sites. Quality control standards required for inclusion of structural images included the following criteria: (1) compliance with acquisition protocol, (2) full brain coverage, (3) minimal motion (Gibbs ringing), (4) absence of flow/zipper, and (5) minor atrophy/vascular degeneration (all assessed qualitatively by trained expert at UCLA). Before data analysis, we at our site additionally conducted visual, nonautomated inspection and quality control of the structural MRI data.

Back to Top | Article Outline

2.4. Image preprocessing

We analyzed T1-weighted structural images from all subjects using SPM8 (Wellcome Trust Center for Neuroimaging; http://www.fil.ion.ucl.ac.uk/spm/software/spm8/) running on Matlab 7 (R2011b). The images were segmented into GM, white matter, and cerebrospinal fluid components using SPM8’s unified segmentation [6]. The segmentation was set to output component images in both native and Montreal Neurological Institute (MNI) template space. The normalized and modulated GM images were then spatially smoothed using an 8mm full width at half maximum Gaussian filter. These preprocessed GM images were then used in the SVM algorithm based analysis.

Back to Top | Article Outline

2.5. SVM classification

For multivariate pattern classification, we used an SVM algorithm. SVM is a supervised learning algorithm, which requires training examples to generate a classification model for a given problem [42]. In particular, the SVM algorithm attempts to determine a separating hyperplane (decision boundary) optimizing the separation between 2 groups. After training, the decision boundary can be used to classify data sets not yet seen by the SVM algorithm. It can also be used to determine features that significantly contribute to the classification process. In the current context, SVMs were trained to classify GM images as either belonging to participants with CPP (assigned class value of +1) or healthy controls (class value of −1). Each voxel in the image was treated as a feature and each GM image represented a point in a high-dimensional feature space. Using the available data set, SVM algorithms were then trained to obtain a hyperplane optimally separating the images from the 2 groups. To minimize the dimensionality of the feature space and ensure that we were only investigating GM, we included only voxels in the GM image with values greater than 0.1. We used a linear SVM algorithm and set the regularization parameter to 1.

Decreases in GM density are known to occur with increasing age [36,40]. Therefore, because the ages among our groups were somewhat variable, and because we did not want age as a driving input for our classification, we regressed out age from the data before SVM classification analyses. Site-related differences in GM density were also regressed from the analysis through adjustment of mean GM intensity values as follows. For each voxel in a given image, a correction factor was applied [mean intensity over all GM images from the image’s site (site mean) minus mean intensity over all voxels from all images included in the analysis (global mean)].

In the absence of a proper test set to evaluate the classification performance of the trained SVM, we used a leave-one-out cross-validation (LOOCV) approach. The LOOCV approach provides a relatively unbiased estimate of the classifier’s true generalization performance [1]. In this approach, GM images from all but one subject (N−1 of the N subjects) are used for training the SVM. The trained SVM is then tested using the remaining image. This process is repeated until all the images are used as a test image. The LOOCV classification accuracy is calculated as the percentage of images that are correctly classified. The alternative approach of leave-pair-out cross-validation (LPOCV) was also tested, despite the limitation that patients and controls were not strictly age matched for each pair (age matching was ±5years for each pair and balanced so that group means were not significantly different). Because there were several possible pair combinations, the LPOCV classification was repeated 1000times, each time randomizing the pairing of subjects, and the mean classification accuracy was computed.

Aside from the classification accuracy, we also calculated the following classification measures: sensitivity=TP/(TP+FN), specificity=TN/(TN+FP), positive predictive value PPV=TP/(TP+FP), and negative predictive value NPV=TN/(TN+FN). TP was defined as the number of images correctly classified as belonging to the patient group (truly positive), TN was defined as the number of images correctly classified as belonging to the healthy control group (truly negative), FP was defined as the number of images from the healthy control group misclassified as belonging to the patient group (falsely positive), and FN was defined as the number of images from the patient group misclassified as belonging to the healthy control group (falsely negative). All of these additional values were assessed for significance level using previously described permutation tests [52].

For this analysis, we used a Matlab version of LIBSVM, a library for SVMs (http://www.csie.ntu.edu.tw/˜cjlin/libsvm; [13]).

Back to Top | Article Outline

2.6. Permutation test

We were also interested in the significance of each voxel’s contribution to the classification accuracy. We used a permutation test [34] with 5000 iterations to generate an SVM-derived significance map, a map showing regions that significantly contributed to the classification performance of the trained SVM. In brief, this nonparametric test involved randomly permuting the class labels and training an SVM for each permutation. This gives an estimate of the probability distribution of the SVM weight associated with each voxel under the null hypothesis of no relationship between class labels and the global structure of the training data. The P value is then calculated as the proportion of values in the null distribution that is greater or equal to the value obtained using the original (ie, nonpermuted) labels. Thus, the farther away the weight value of a voxel from the major mass of the distribution under null hypothesis (small P value), the more likely it is to be significantly predictive for the class label. The significance map was corrected for multiple comparisons using false discovery rate with q<0.05 and cluster size greater than 20 voxels. We used an SPM8 extension, xjview (http://www.alivelearn.net/xjview8), for the anatomical labeling of significant brain regions.

Using the same approach, we also estimated the significance of the classification accuracy and other classification measures described above. For each permutation, we computed the LOOCV/LPOCV accuracy, sensitivity, and specificity, among others. The P value associated with each measure was calculated as the proportion of values in the null distribution greater than the value obtained when using the nonpermuted labels.

Back to Top | Article Outline

2.7. Region of interest correlation analysis

Several regions of interest (ROIs) were extracted from the significance map. The values of the GM density from all voxels within each ROI were extracted from patient data sets and the mean computed. The correlation of the ROI’s mean density with other behavioral measures (symptom severity, pain duration, anxiety and depression) across all CPP patients was estimated. Previously, brain morphological changes were observed as correlating with duration of chronic pain symptoms, but only when the patient population was split into subgroups and separately assessed for short-term (<5years) or long-term (>5years) symptom duration [7]. Therefore, we similarly wished to investigate the relationship of our inferred GM density change with chronic pain duration at different stages of chronic pain. Due to our initially observed overall nonlinear relationship of duration and GM density, we used a systematic approach to characterize the correlations of duration subsets. To do this, we partitioned the range of symptom duration into 2 intervals—from 0 to x and x to 33 (greatest symptom duration in our population, years) and systematically varied x from 2.5, 5.0, 7.5, and 10. The correlation of GM density with symptom duration for each interval was then computed.

Back to Top | Article Outline

2.8. Total GM volume

We also performed statistical analysis of the total GM volume. Total GM volume was computed by multiplying the value of each voxel by its total volume and summing over all voxels included. A 2-sided 2-sample t test was used to determine if the difference of mean total GM volume between the 2 groups was significant. Total whole brain GM volume values were also entered into an additional SVM analysis to test whether total GM volume could predict CPP from healthy controls.

Back to Top | Article Outline

3. Results

The trained SVM performed significantly better than chance in classifying the patient group from the healthy control group (Table 1). The classifier’s accuracy when using LOOCV was 72.73% (P=.039). The sensitivity was 69.70% (P=.067) and the specificity was 72.73% (P=.046). The positive and negative predictive values were 71.88% (P=.039) and 70.59% (P=.047), respectively. The classifier’s accuracy was slightly lower when using LPOCV at 69.36% (P=.017). The sensitivity was 70.22% (P=.013) and the specificity was 68.65% (P=.025). The positive and negative predictive values were 67.41% (P=.049) and 71.32% (P=.017), respectively. All following reported results are from the LOOCV analysis.

Table 1

Table 1

The classification of individual GM images was visualized by showing the distance of each GM image from the trained SVM’s separating hyperplane (Fig. 1A). Positive distance represents patient group (+1) classification, while negative distance represents healthy control group (−1). Participants with negative distance indicate false negatives, while healthy participants with positive distance are false positives.

Fig. 1.

Fig. 1.

Several regions of greater GM density were observed in CPP compared to the healthy control group. These regions contributed significantly to the classification performance of the trained SVM to discriminate between brain structure of CPP and healthy controls (Fig. 2). Regions of positive SVM weights included bilateral primary somatosensory cortex (S1), left pre-supplementary motor area (pre-SMA), bilateral hippocampus and left amygdala. All regions were corrected for multiple comparisons (false discovery rate (FDR), q<0.05) and cluster size greater than 20 voxels (Table 2).

Fig. 2.

Fig. 2.

Table 2

Table 2

Values of GM density within the patient group, from the identified discriminative ROIs (positive SVM weights, which suggest greater regional GM density in the CPP group), were correlated with patients’ behavioral measures from the questionnaire data. Pain intensity (MPQ, total sensory score) was significantly correlated with the weights of the left postcentral gyrus/paracentral lobule (r=0.37, P=.05), left pre-SMA (r=0.38, P=.05), and right S1 (r=0.55, P=.0015) (with age regressed out). Depression score (HADS) was significantly correlated with the weight of the right hippocampus (r=0.36, P=.04). Anxiety score (HADS) was not significantly correlated with any region.

As assessed in patients only, symptom duration was not linearly related to any of the identified regions of positive weights contributing to the SVM. However, when limiting the analysis to within specific ranges of symptom duration, some correlations emerged in 4 regions (Fig. 3). Specifically, bilateral S1 (both left and right S1 regions) showed a strong positive relationship with durations from 0 to 2.5years and a significant negative correlation with symptom durations greater than 5years. In contrast, the left hippocampus was negatively correlated with symptom durations of fewer than 10years. The left amygdala was significantly correlated with symptom duration of fewer than 7.5years (Table 3).

Fig. 3.

Fig. 3.

Table 3

Table 3

Total GM volume was significantly greater in CPP compared to healthy controls (P=.0042). SVM classification based on only total GM volume resulted in classification accuracy of 66.67% (with age regressed).

Back to Top | Article Outline

4. Discussion

We have identified a preliminary brain classifier associated with CPP. This GM classifier distinguishes individuals with CPP from matched healthy controls with an accuracy of 73%. On the basis of our findings, several brain regions were revealed to have positive weights contributing to the SVM classification of CPPs vs healthy controls. These regions included bilateral primary somatosensory cortex (S1), left pre-SMA, bilateral hippocampus, and left amygdala. The nature of positive weights of these regions suggests that patients with CPP may have increased GM density within these regions.

Previous morphological studies of chronic pain have revealed altered (typically decreased) GM density across numerous brain regions implicated in pain processing and perception (reviewed in [50]).

Increased GM density in S1 was observed in another chronic visceral pain syndrome (irritable bowel syndrome), both in terms of GM volume [27] and cortical thickness [23]. A region we identified within the primary somatosensory cortex (S1) was highly similar in location to a VBM study of women with CPP also collected through the MAPP Research Network [24]. Somatotopically, 2 regions that we observed in S1 were within a region activated during electrical stimulation of the human clitoris [32]. Our results of another S1 region (MNI: 44, −24, 56) closely corresponded to a region of S1 activity related to spontaneous pain ratings observed in a previous VBM study of males with CPP (MNI: 36, −28, 58) [20]. Importantly, patients included in our study had symptoms characteristic of interstitial cystitis, with pain in the lower abdomen accompanied by urinary symptoms (eg, increased urgency, increased frequency of urination, painful filling of the bladder). Also, the majority (85%) of participants with CPP also showed signs of pelvic floor dysfunction, which typically involves a hypertonic state of the pelvic floor musculature [12,39].

The differences in symptoms between our CPP patient population and those of previous morphological studies of CPP may account for differences in GM density findings. Specifically, one previous CPP study observed decreased GM density within the left middle frontal gyrus, right putamen, bilateral midcingulate cortex, right insular cortex, and left thalamus in a patient population with endometriosis and pain. However, these findings of regional decreased GM density may contrast with our present findings because they defined CPP as moderate to severe pelvic pain of 4 or more (on a 10-point scale) and excluded patients with interstitial cystitis [4]. Further, our patient population included only 9 individuals with endometriosis. In contrast to our present findings, and possibly due to sex-related differences in CPP, male CPP patients previously demonstrated decreased GM volume within the left anterior cingulate cortex [33]. Another VBM study of CPP in male subjects, revealed through a region of interest correlation analysis, showed a positive correlation of right insular cortex GM density with pain severity and a positive correlation of anterior cingulate cortex GM density with duration of pain (ie, pain chronicity) [20].

Positive weights within bilateral parahippocampal/hippocampal gyri and amygdala also significantly contributed to our classification of CPP. Complementary to our findings, greater GM density within the left amygdala was previously observed in a patient population of CPP and endometriosis [4]. However, decreased amygdala GM volume has been observed in irritable bowel syndrome [27] and in healthy individuals with increased visceral sensitivity [19]. Increased GM density within the parahippocampal gyrus and basal ganglia has also been observed in participants with provoked vestibulodynia [48]. Increased GM volume has also been observed in patients with primary dysmenorrhea in the right posterior hippocampus among several other brain regions of altered GM volume [51]. In contrast, other studies using VBM have shown less GM in the hippocampus in posttraumatic stress disorder [10,21,54] and chronic fatigue syndrome [44]. The amygdala and hippocampus, in addition to processes of emotion and memory, play direct roles in pain modulation and processing of anxiety, fear and aversive contents of pain [16,29,37,41,43,47]. Together, our and others’ observations of altered GM density within the hippocampus and amygdala in chronic pain states may reflect an altered state of pain modulation and emotional-regulatory aspects of chronic pain.

Back to Top | Article Outline

4.1. Correlations of GM density change with symptom duration

The range of symptom durations included in each study population could account for inconsistent findings of GM density change. In support of this concept, we observed that the degree of patients’ positive SVM weight within regions of our significance map was related to symptom duration, but only at certain stages of chronic pain. This suggests that the relationship between GM density and symptom duration may not be linear over time. Additionally, this relationship differs for distinct brain regions. For example, our results suggest that in S1, GM density may increase in the first few years of chronic pain; but as pain persists past the first initial years, the GM density may decrease—the longer the pain, the less the GM density. In contrast, in the left hippocampus we observed a negative correlation of GM density and symptom duration during the first 10years of pain but not after (potential decreases in GM density over time up to 10years), similar to a previous investigation [53]. In other chronic pain conditions, less GM density is observed in insular cortex, S1 and motor cortex; these differences are significantly correlated with pain duration, but only when the duration is greater than 5years [7]. However, importantly, these observations are not based on longitudinal studies, but rather on correlation results at a single time point. While it currently remains unclear whether chronic pain causes nonlinear morphological changes over time, emerging data from longitudinal studies indicate that this might be the case [22,45].

Back to Top | Article Outline

4.2. Enhanced potential of fMRI analyses by using multivariate methods

In the present study, through using SVM algorithms, we were able to classify between CPP and healthy controls based on brain morphological differences. Additionally, through our use of the SVM algorithm, we identified several brain regions of positive contributory weights, suggesting greater GM density in these regions in CPP. Some of these regions were in agreement with regions identified in a conventional VBM analysis also conducted through the MAPP Research Network [24]. Compared to the conventional univariate analysis, our SVM approach appeared to be more sensitive to identifying additional regions of morphological differences between CPP and healthy controls. This was somewhat expected due to known increases in power when using multivariate analyses [38] and because we previously observed increased sensitivity when using SVM to classify structural differences in chronic low back pain, compared to conventional VBM analysis [52].

Back to Top | Article Outline

4.3. Limitations

Brain morphological changes in chronic pain typically show decreases in GM density in brain regions related to aspects of pain processing (reviewed in [50]). In contrast, our observations of positive SVM weights indicate increased GM density within regions of S1, hippocampus and amygdala in CPP. Thus, changes in brain morphology may vary between different types of chronic pain, as noted previously [7]. For example, less total GM volume is observed in chronic low back pain and fibromyalgia [26], but not in participants with complex regional pain syndrome or knee osteoarthritis. However, regions of increased GM density are not entirely rare. For example, increased GM density has also been observed in striatal regions in fibromyalgia [46,52]. Differences in directionality of GM change across studies of chronic pain could be attributed to the differences in the various populations being studied, presence or absence of comorbidities, and medication usage.

We here describe a preliminary structural MRI classifier of CPP. Future characterization of a true brain classifier for CPP will require larger sample sizes, refined models and extensive model validation (eg, prospective testing and/or inclusion of independent training and test sets). Future analyses could also use more optimal analysis methods of modeled correction for nuisance features such as age [17,25]. Further, it is currently unknown whether these regions of greater GM density may be a specific feature of CPP and or generalizable to other chronic pain conditions.

Back to Top | Article Outline

4.4. Conclusions

We have identified a preliminary structural brain based classifier representative of CPP. The classifier was comprised of several distributed regions of positive SVM weight that contributed to our SVM algorithm including S1, pre-SMA, the hippocampus, and amygdala. While previous studies have typically observed decreased GM density in chronic pain, the regions we identified suggest regional increased GM density in CPP. Ultimately, the good classification accuracy observed in our study suggests the significance of these regions in distinguishing participants with CPP from healthy controls and serve as a preliminary potential biomarker for CPP. While a structural brain classifier of CPP as a definitive marker of disease would have significant ethical and legal implications, our results are preliminary and further investigation is warranted. Through the combination of SVM analyses of brain structure and function with additional genotype and phenotype biomarkers, we may ultimately define strong predictors of treatment response, define subgroups of chronic pain syndromes and develop tailored and systematic therapy for the individual patient.

Back to Top | Article Outline

Conflict of interest

The authors report no conflict of interest.

Back to Top | Article Outline

Acknowledgments

Thanks to Cody Ashe-McNally for his technical expertise in coordinating and running the cross-site quality control of all MAPP Research Network neuroimaging data. Special thanks to Jeff Alger for his expertise as a physicist at UCLA in oversight and coordination of the multisite collection of neuroimaging data. Funding for the MAPP Research Network was obtained under a cooperative agreement from National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health (NIH) (DK82370, DK82342, DK82315, DK082344, DK082325, DK082345, DK082333, and DK082316). This study was also supported by an additional NIH grant (K24 DA029262) and the Redlich Pain Research Endowment. The authors declare no competing financial interests.

Back to Top | Article Outline

References

[1] Airola A, Pahikkala T, Waegeman W, De Baets B, Salakoski T. An experimental comparison of cross-validation techniques for estimating the area under the ROC curve. Comput Stat Data Anal 2011;55:1828-1844.
[2] Apkarian AV, Baliki MN, Farmer MA. Predicting transition to chronic pain. Curr Opin Neurol 2013;26:360-367.
[3] Apkarian AV, Sosa Y, Sonty S, Levy RM, Harden RN, Parrish TB, Gitelman DR. Chronic back pain is associated with decreased prefrontal and thalamic gray matter density. J Neurosci 2004;24:10410-10415.
[4] As-Sanie S, Harris RE, Napadow V, Kim J, Neshewat G, Kairys A, Williams D, Clauw DJ, Schmidt-Wilcke T. Changes in regional gray matter volume in women with chronic pelvic pain: a voxel-based morphometry study. PAIN® 2012;153:1006-1014.
[5] Ashburner J, Friston KJ. Voxel-based morphometry—the methods. Neuroimage 2000;11:805-821.
[6] Ashburner J, Friston KJ. Unified segmentation. Neuroimage 2005;26:839-851.
[7] Baliki MN, Schnitzer TJ, Bauer WR, Apkarian AV. Brain morphological signatures for chronic pain. PLoS ONE 2011;6:e26010.
[8] Bendfeldt K, Klöppel S, Nichols TE, Smieskova R, Kuster P, Traud S, Mueller-Lenke N, Naegelin Y, Kappos L, Radue EW, Borgwardt SJ. Multivariate pattern classification of gray matter pathology in multiple sclerosis. Neuroimage 2012;60:400-408.
[9] Berry SH, Elliott MN, Suttorp M, Bogart LM, Stoto MA, Eggers P, Nyberg L, Clemens JQ. Prevalence of symptoms of bladder pain syndrome/interstitial cystitis among adult females in the United States. J Urol 2011;186:540-544.
[10] Bremner JD, Randall P, Scott TM, Bronen RA, Seibyl JP, Southwick SM, Delaney RC, McCarthy G, Charney DS, Innis RB. MRI-based measurement of hippocampal volume in patients with combat-related posttraumatic stress disorder. Am J Psychiatry 1995;152:973-981.
[11] Brown JE, Chatterjee N, Younger J, Mackey S. Towards a physiology-based measure of pain: patterns of human brain activity distinguish painful from non-painful thermal stimulation. PLoS ONE 2011; 6.
[12] Butrick CW. Interstitial cystitis and chronic pelvic pain: new insights in neuropathology, diagnosis, and treatment. Clin Obstet Gynecol 2003;46:811-823.
[13] Chang CC, Lin CJ. LIBSVM: a library for support vector machines. ACM Trans Intell Syst Technol 2011; 2.
[14] Clemens JQ, Mullins C, Kusek JW, Kirkali Z, Mayer EA, Rodríguez LV, Klumpp DJ, Schaeffer AJ, Kreder KJ, Buchwald D, Andriole GL, Lucia MS, Landis JR, Clauw DJ., MAPP Research Network Study Group. The MAPP research network: a novel study of urologic chronic pelvic pain syndromes. BMC Urol 2014;14:57.
[15] Davis KD, Pope G, Chen J, Kwan CL, Crawley AP, Diamant NE. Cortical thinning in IBS: implications for homeostatic, attention, and pain processing. Neurology 2008;70:153-154.
[16] Derbyshire SW, Jones AK, Gyulai F, Clark S, Townsend D, Firestone LL. Pain processing during three levels of noxious stimulation produces differential patterns of central activity. PAIN® 1997;73:431-445.
[17] Dukart J, Schroeter ML, Mueller K. Age correction in dementia—matching to a healthy brain. PLoS ONE 2011;6:e22193.
[18] Ecker C, Rocha-Rego V, Johnston P, Mourao-Miranda J, Marquand A, Daly EM, Brammer MJ, Murphy C, Murphy DG., MRC AIMS Consortium. Investigating the predictive value of whole-brain structural MR scans in autism: a pattern classification approach. Neuroimage 2010;49:44-56.
[19] Elsenbruch S, Schmid J, Kullmann JS, Kattoor J, Theysohn N, Forsting M, Kotsis V. Visceral sensitivity correlates with decreased regional gray matter volume in healthy volunteers: a voxel-based morphometry study. PAIN® 2014;155:244-249.
[20] Farmer MA, Chanda ML, Parks EL, Baliki MN, Apkarian AV, Schaeffer AJ. Brain functional and anatomical changes in chronic prostatitis/chronic pelvic pain syndrome. J Urol 2011;186:117-124.
[21] Gurvits TV, Shenton ME, Hokama H, Ohta H, Lasko NB, Gilbertson MW, Orr SP, Kikinis R, Jolesz FA, McCarley RW, Pitman RK. Magnetic resonance imaging study of hippocampal volume in chronic, combat-related posttraumatic stress disorder. Biol Psychiatry 1996;40:1091-1099.
[22] Gwilym SE, Filippini N, Douaud G, Carr AJ, Tracey I. Thalamic atrophy associated with painful osteoarthritis of the hip is reversible after arthroplasty a longitudinal voxel-based morphometric study. Arthritis Rheum 2010;62:2930-2940.
[23] Jiang Z, Dinov ID, Labus J, Shi Y, Zamanyan A, Gupta A, Ashe-McNalley C, Hong JY, Tillisch K, Toga AW, Mayer EA. Sex-related differences of cortical thickness in patients with chronic abdominal pain. PLoS ONE 2013;8:e73932.
[24] Kairys AE, Schmidt-Wilcke T, Puiu T, Ichesco E, Labus JS, Martucci K, Farmer MA, Ness TJ, Deutsch G, Mayer EA, Mackey S, Apkarian AV, Maravilla K, Clauw DJ, Harris RE. Increased brain gray matter in the primary somatosensory cortex is associated with increased pain and mood disturbance in interstitial cystitis/painful bladder syndrome patients. J Urol 2014. http://dx.doi.org/10.1016/j.juro.2014.08.042 pii: S0022-5347(14)04213-X, [Epub ahead of print].
[25] Koikkalainen J, Pölönen H, Mattila J, van Gils M, Soininen H, Lötjönen J., Alzheimer’s Disease Neuroimaging Initiative. Improved classification of Alzheimer’s disease data via removal of nuisance variability. PLoS ONE 2012;7:e31112.
[26] Kuchinad A, Schweinhardt P, Seminowicz DA, Wood PB, Chizh BA, Bushnell MC. Accelerated brain gray matter loss in fibromyalgia patients: premature aging of the brain? J Neurosci 2007;27:4004-4007.
[27] Labus JS, Dinov ID, Jiang Z, Ashe-McNalley C, Zamanyan A, Shi Y, Hong JY, Gupta A, Tillisch K, Ebrat B, Hobel S, Gutman BA, Joshi S, Thompson PM, Toga AW, Mayer EA. Irritable bowel syndrome in female patients is associated with alterations in structural brain networks. PAIN® 2014;155:137-149.
[28] Landis JR, Williams DA, Lucia MS, Clauw DJ, Naliboff BD, Robinson NA, van Bokhoven A, Sutcliffe S, Schaeffer AJ, Rodriguez LV, Mayer EA, Lai HH, Krieger JN, Kreder KJ, Afari N, Andriole GL, Bradley CS, Griffith JW, Klumpp DJ, Hong BA, Lutgendorf SK, Buchwald D, Yang CC, Mackey S, Pontari MA, Hanno P, Kusek JW, Mullins C, Clemens JQ., MAPP Research Network Study Group. The MAPP research network: design, patient characterization and operations. BMC Urol 2014;14:58.
[29] Lathe R. Hormones and the hippocampus. J Endorinol 2001;169:205-231.
[30] Mathias SD, Kuppermann M, Liberman RF, Lipschutz RC, Steege JF. Chronic pelvic pain: prevalence, health-related quality of life, and economic correlates. Obstet Gynecol 1996;87:321-327.
[31] May A. Chronic pain may change the structure of the brain. PAIN® 2008;137:7-15.
[32] Michels L, Mehnert U, Boy S, Schurch B, Kollias S. The somatosensory representation of the human clitoris: an fMRI study. Neuroimage 2010;49:177-184.
[33] Mordasini L, Weisstanner C, Rummel C, Thalmann GN, Verma RK, Wiest R, Kessler TM. Chronic pelvic pain syndrome in men is associated with reduction of relative gray matter volume in the anterior cingulate cortex compared to healthy controls. J Urol 2012;188:2233-2237.
[34] Mourão-Miranda J, Bokde AL, Born C, Hampel H, Stetter M. Classifying brain states and determining the discriminating activation patterns: support vector machine on functional MRI data. Neuroimage 2005;28:980-995.
[35] Mugler JP 3rd, Brookeman JR. Three-dimensional magnetization-prepared rapid gradient-echo imaging (3D MP RAGE). Magn Reson Med 1990;15:152-157.
[36] Murphy DG, DeCarli C, Schapiro MB, Rapoport SI, Horwitz B. Age-related differences in volumes of subcortical nuclei, brain matter, and cerebrospinal fluid in healthy men as measured with magnetic resonance imaging. Arch Neurol 1992;49:839-845.
[37] Neugebauer V, Li W, Bird GC, Han JS. The amygdala and persistent pain. Neuroscientist 2004;10:221-234.
[38] Orrù G, Pettersson-Yeo W, Marquand AF, Sartori G, Mechelli A. Using support vector machine to identify imaging biomarkers of neurological and psychiatric disease: a critical review. Neurosci Biobehav Rev 2012;36:1140-1152.
[39] Peters KM, Carrico DJ, Kalinowski SE, Ibrahim IA, Diokno AC. Prevalence of pelvic floor dysfunction in patients with interstitial cystitis. Urology 2007;70:16-18.
[40] Pfefferbaum A, Mathalon DH, Sullivan EV, Rawles JM, Zipursky RB, Lim KO. A quantitative magnetic resonance imaging study of changes in brain morphology from infancy to late adulthood. Arch Neurol 1994;51:874-887.
[41] Ploghaus A, Narain C, Beckmann CF, Clare S, Bantick S, Wise R, Matthews PM, Rawlins JN, Tracey I. Exacerbation of pain by anxiety is associated with activity in a hippocampal network. J Neurosci 2001;21:9896-9903.
[42] Pontil M, Verri A. Properties of support vector machines. Neural Comput 1998;10:955-974.
[43] Prado WA, Roberts MHT. An assessment of the antinociceptive and aversive effects of stimulating identified sites in the rat-brain. Brain Res 1985;340:219-228.
[44] Puri BK, Jakeman PM, Agour M, Gunatilake KD, Fernando KA, Gurusinghe AI, Treasaden IH, Waldman AD, Gishen P. Regional grey and white matter volumetric changes in myalgic encephalomyelitis (chronic fatigue syndrome): a voxel-based morphometry 3 T MRI study. Br J Radiol 2012;1015:e270-e273.
[45] Rodriguez-Raecke R, Niemeier A, Ihle K, Ruether W, May A. Brain gray matter decrease in chronic pain is the consequence and not the cause of pain. J Neurosci 2009;29:13746-13750.
[46] Schmidt-Wilcke T, Luerding R, Weigand T, Jürgens T, Schuierer G, Leinisch E, Bogdahn U. Striatal grey matter increase in patients suffering from fibromyalgia—a voxel-based morphometry study. PAIN® 2007;132:S109-S116.
[47] Schneider F, Habel U, Holthusen H, Kessler C, Posse S, Müller-Gärtner HW, Arndt JO. Subjective ratings of pain correlate with subcortical-limbic blood flow: an fMRI study. Neuropsychobiology 2001;43:175-185.
[48] Schweinhardt P, Kuchinad A, Pukall CF, Bushnell MC. Increased gray matter density in young women with chronic vulvar pain. PAIN® 2008;140:411-419.
[49] Seminowicz DA, Labus JS, Bueller JA, Tillisch K, Naliboff BD, Bushnell MC, Mayer EA. Regional gray matter density changes in brains of patients with irritable bowel syndrome. Gastroenterology 2010;139:e2.
[50] Smallwood RF, Laird AR, Ramage AE, Parkinson AL, Lewis J, Clauw DJ, Williams DA, Schmidt-Wilcke T, Farrell MJ, Eickhoff SB, Robin DA. Structural brain anomalies and chronic pain: a quantitative meta-analysis of gray matter volume. J Pain 2013;14:663-675.
[51] Tu CH, Niddam DM, Chao HT, Chen LF, Chen YS, Wu YT, Yeh TC, Lirng JF, Hsieh JC. Brain morphological changes associated with cyclic menstrual pain. PAIN® 2010;150:462-468.
[52] Ung H, Brown JE, Johnson KA, Younger J, Hush J, Mackey S. Multivariate classification of structural MRI data detects chronic low back pain. Cereb Cortex 2014;24:1037-1044.
[53] Valet M, Gündel H, Sprenger T, Sorg C, Mühlau M, Zimmer C, Henningsen P, Tölle TR. Patients with pain disorder show gray-matter loss in pain-processing structures: a voxel-based morphometric study. Psychosom Med 2009;71:49-56.
[54] Villarreal G, Hamilton DA, Petropoulos H, Driscoll I, Rowland LM, Griego JA, Kodituwakku PW, Hart BL, Escalona R, Brooks WM. Reduced hippocampal volume and total white matter volume in posttraumatic stress disorder. Biol Psychiatry 2002;52:119-125.
[55] Wager TD, Atlas LY, Lindquist MA, Roy M, Woo CW, Kross E. An fMRI-based neurologic signature of physical pain. N Engl J Med 2013;368:1388-1397.
[56] Younger JW, Shen YF, Goddard G, Mackey SC. Chronic myofascial temporomandibular pain is associated with neural abnormalities in the trigeminal and limbic systems. PAIN® 2010;149:222-228.
Keywords:

Gray matter density; Machine learning; Support vector machine; SVM; UCPPS

© 2014 International Association for the Study of Pain