Search tips
Search criteria 


Logo of neurologyNeurologyAmerican Academy of Neurology
Neurology. 2009 May 26; 72(21): 1843–1849.
PMCID: PMC2690986

Brain and ventricular volumetric changes in frontotemporal lobar degeneration over 1 year



Measurement of volumetric changes with MR might be a useful surrogate endpoint for clinical trials in frontotemporal lobar degeneration (FTLD). Because there is only limited longitudinal imaging data currently available, we measured the rate of change over 1 year of whole brain volume (WBV) and ventricular volume (VV) in patients with FTLD.


Subjects with an FTLD cognitive syndrome were recruited from five centers using standard clinical diagnostic criteria for behavioral variant frontotemporal dementia (bvFTD), progressive nonfluent aphasia (PNFA), semantic dementia (SMD), and progressive logopenic aphasia. Structural brain imaging, using three-dimensional T1-weighted sequences at 1.5 teslas, and cognitive, behavioral, and functional assessments were performed at baseline and approximately 1 year later. The boundary shift integral algorithm was used to determine change in WBV and VV.


There were 76 patients (mean age 64 years; 41 men and 35 women) who had usable baseline and annual scans. The group-wise annualized change was −1.62% (SD 1.03, range +0.69 to −3.6) for WBV and 11.6% (SD 5.9, range −1.3 to 23.9) for VV. Rates of change were similar among bvFTD, PNFA, and SMD groups. Longitudinal changes in WBV and VV were correlated with decline on clinical global and cognitive measures.


Multicenter, serial measurements of whole brain volume (WBV) and ventricular volume (VV) from magnetic resonance scans were feasible in patients with frontotemporal lobar degeneration (FTLD). Using WBV or VV as outcome measures would require recruiting (at 80% power) 139 or 55 subjects per group to detect a small (25%) or medium-sized (40%) effect in a randomized, placebo-controlled trial of a putative agent for FTLD.


= Alzheimer disease;
= boundary shift integral;
= behavioral variant frontotemporal dementia;
= corticobasal degeneration;
= confidence interval;
= frontotemporal lobar degeneration;
= frontotemporal lobar degeneration modified Clinical Dementia Rating Scale;
= Mini-Mental State Examination;
= magnetic resonance;
= not significant;
= progressive logopenic aphasia;
= progressive nonfluent aphasia;
= progressive supranuclear palsy;
= semantic dementia;
= total intracranial volume;
= ventricular volume;
= whole brain volume.

Brain imaging might be a useful measure of disease progression in clinical trials in neurodegenerative diseases. For example, reductions in whole brain volume (WBV) have paralleled clinical progression in several Alzheimer disease (AD) clinical trials, including ones testing milameline,1 AN-1792,2 and donepezil.3 Because of the greater syndromic heterogeneity of frontotemporal lobar degenerations (FTLDs), structural neuroimaging might be particularly useful across different clinical syndromes that share a common molecular basis. There is only limited information on the magnitude of change in longitudinal studies of FTLD, however.4–7

We performed serial volumetric magnetic resonance (MR) scans in a group of patients with FTLD who were part of a multicenter, 1-year, simulated clinical trial. We analyzed the scans using automated methodology to measure WBV and ventricular volume (VV). The design of the parent study and the longitudinal cognitive, functional, and behavioral data8 have been reported.



Patients with an FTLD cognitive/behavioral syndrome were recruited from five academic medical centers over 2 years. Patients or their responsible family member provided written informed consent. The study was approved by the institutional review boards of all five clinical sites. The inclusion and exclusion criteria have been previously reported.8 Standard clinical diagnostic criteria were used for the syndromes of behavioral variant frontotemporal dementia (bvFTD), progressive nonfluent aphasia (PNFA), progressive logopenic aphasia (PLA), and semantic dementia (SMD). Because of a concern that patients with PLA might have AD pathology rather than an FTLD pathology,9–12 these subjects were excluded from group analyses. Subjects were required to have frontal, temporal, or insular atrophy or dysfunction by brain imaging. Structural brain imaging and cognitive, behavioral, and functional assessments were performed at baseline and approximately 1 year later.

Cognitive and behavioral assessments.

A battery of cognitive, functional, and behavioral assessment were performed baseline and then at the annual follow-up visit. These have been previously described.8 These included the modified Alzheimer's Disease Cooperative Study Clinical Global Impression of Change,13 an FTLD-specific Clinical Dementia Rating (FLTD-CDR)8 that included two additional domains: language and behavior and the Mini-Mental State Examination (MMSE).14

We constructed two cognitive test composites to express the neuropsychological test results in a concise manner. The composition and development of the composites are described in detail elsewhere.8 One composite, the executive composite, included six variables (Trails, backward digit span, Digit Symbol, Stroop, number cancellation, and total errors on Stroop interference test, Trails, and delayed recognition). The other, a language composite, included five variables (Boston Naming test, the two verbal fluency scores, Similarities, and immediate recall). We also constructed a global composite by combining the elements of both the executive and language composites.


MR images were acquired using three-dimensional volumetric T1-weighted imaging sequences at 1.5 teslas. The scanning parameters were as follows: A coronal volume with 124 contiguous partitions, each 1.6 mm in thickness, was acquired. The field of view was 24 × 24 cm, with 256 × 192 matrix, bandwidth 16 kHz, 25° flip angle, repetition time = 23 msec, and echo time = minimum full echo time. The boundary shift integral (BSI) algorithm was used to determine change in WBV and VV.15,16 Differences were calculated in pairwise fashion between the baseline and follow-up scans. After intensity normalization of both scans and spatial registration of the follow-up to the baseline scan, intensity differences between the two scans at the brain–CSF boundary were used to compute change in volume. The algorithm corrects drifts or discontinuities in gradient calibration between the baseline and follow up scans by using 9 degrees of freedom registration to a brain/skull composite target. The whole brain atrophy rate reflects shrinkage of the brain on the follow-up scan relative to the baseline scan from out-to-in at the cortical surface and from in-to-out at the ventricular surface. The ventricular atrophy rate was derived by creating a binary mask for each subject that selectively extracted ventricular change. The binary mask was an approximate area overlaying the ventricles within which the BSI was measured. Quality control testing in our laboratory shows that the nonparametric intraclass correlation coefficient for test–retest reproducibility of rate measurements from serial MRI scans with the BSI method was 0.91 for ventricle and 0.89 for brain.17

Total intracranial volume (TIV) was determined by tracing the margin of the inner table of the skull on consecutive slices from a T1-weighted sagittal sequence with 5-mm contiguous sections.18 We used the TIV value to calculate a baseline brain atrophy measure, by taking the ratio of WBV and TIV, for descriptive purposes.


The differences between baseline and follow-up scans were analyzed in each syndromic subgroup and in the group as a whole, after excluding subjects with PLA. The difference in WBV and VV between the first and second scans was expressed as a percent change from the first scan, and was then annualized by dividing by the interval between scans/365 days. Descriptive statistics for the baseline and change values for WBV and VV are given in tables 1 and 2. Spearman correlation coefficients were derived to describe the relationship between change in imaging features and change in cognitive and behavioral assessments. Linear regression analyses were also performed, with annual percent change in WBV and VV as the dependent variables, and FTLD-CDR and cognitive composites as independent variables, controlling for age, sex and diagnostic group (modeling with indicator variables), time between scans (years), and baseline MR value. To assess the adequacy of using annualized percent change as an outcome variable, Bland–Altman plots were constructed between the annualized percent change and the geometric mean of the MR data. Finally, we performed power calculations by assuming the use of a two-tailed t test for a two-group study, an α of 0.05, and a β of 0.80. We derived numbers of subjects per group after taking attrition into account.



Of 107 patients who were enrolled at baseline, 103 (96%) successfully completed an MR scan at baseline. The other 4 had technically unacceptable scans (excessive motion artifact). At the follow-up visit, 79 (of the 90 patients who returned for the follow-up visit, 88%) had MR scans of sufficient quality to be analyzed. The current analysis set included 76 patients (mean age 64 years; 41 men and 35 women) who had usable MR scan pairs and had full clinical data. Of the 76 subjects, 34 had bvFTD, 17 had PNFA, 16 had SMD, and 9 had PLA. Table 1 shows demographic and clinical features of patients. Sixty-nine patients had follow-up scans performed within 11 to 16 months of the baseline scan; the remaining 7 had scans that were performed between 17 and 22 months after the baseline visit. Table 1 also shows the baseline imaging values.

Annual change values.

In the group as a whole (excluding PLA subjects), the annualized percent change was −1.6% for WBV and 11.6% for VV (table 2). These changes correspond to a mean annual decline in WBV of 21 mL and an increase in VV of 6.4 mL. The three syndromes—bvFTD, PNFA, and SMD—showed similar means, standard deviations, and ranges (all within-syndromic differences were p > 0.10). The PLA group exhibited slightly more atrophy than the other groups. Women had virtually identical annualized changes compared with men for both WBV (−1.73% vs −1.53%, not significant [NS]) and VV (13.1% vs 10.7%, NS). There was no correlation of age with either WBV or VV. Figures e-1 and e-2 on the Neurology® Web site at show the percent annualized changes for WBV and VV in relation to baseline values. There was also no relationship using Spearman correlation coefficients between annualized change values for the syndromes of bvFTD, PNFA, and SMD combined and either baseline volumes (WBV, r = 0.06, NS; VV, r = 0.12, NS) or baseline brain atrophy (WBV, r = −0.21, p = 0.07; VV, r = 0.15, NS).

We identified 15 normal volunteers (mean age 70 years [range 51–75 years]; 7 men) from Mayo Clinic Rochester Alzheimer's Disease Research Center with serial scans and found that their annual percent change in WBV was −0.39 ± 0.36% (range −1.01 to 0.24) and in VV was 2.34 ± 1.36% (range 5.66 to −0.35) (p < 0.001 compared with the patients with FTLD, for both WBV and VV). Table 2 shows the number of cases in our study group with annual percent change values on imaging that fell within 2 SDs of the normal range. Some patients, mainly among the bvFTD group, did not exceed the normal rate of brain volume loss or ventricular expansion.

Correlations with baseline levels and change in cognitive measures.

For the syndromes of bvFTD, PNFA, and SMD separately and as a group, baseline FTLD-CDR and WBV (r ≈ 0.3) or VV (r ≈ 0.3) were correlated (p < 0.01; for all except language and WBV, p = 0.03).

The results from the Bland–Altman analysis indicated that the annualized percent change was an adequate summary, in that there was no association between this transformation and the geometric mean of the MR data. Summaries from a series of linear regression analyses comparing change scores on the FTLD-CDR and the cognitive composites and changes in WBV and VV are shown in table 3. All but two reached a threshold of p = 0.006 when using the Bonferroni adjustment.

Table thumbnail
Table 3 Linear regression estimates describing the association between annualized percent changes in whole brain volume and ventricular volume for all patients with bvFTD, PNFA, and SMD and either FTLD-CDR or cognitive composites

Sample size considerations.

We performed sample size calculations (table 4) using the annual percent changes in WBV and VV and either small (25%) and medium (40%) effect sizes. We also took attrition of 26% into account to formulate the number of patients to be enrolled. VV exhibited somewhat more favorable estimates than WBV.

Table thumbnail
Table 4 Sample size (per group) estimates for enrollment in clinical trials in which brain volume or ventricular volume are outcome measures, for small (25%) or conservative medium (40%) effect sizes in a t test


We have shown that it was feasible to perform quantitative brain imaging in patients with FTLD over a 1-year interval. If a patient could be successfully scanned at the baseline visit, there was a nearly 90% probability that a follow-up scan could be completed. Serial measurements of WBV and VV in FTLD syndromes showed that there was a readily detectable change over that interval. The different clinical syndromes in the FTLD spectrum showed roughly similar quantitative changes. There was a moderate amount of individual variability in both baseline measurements and the amount of annual change. The declines in brain volume and concomitant increases in ventricular size were correlated with the declines in MMSE, FTLD-CDR, and the cognitive composites. If one measurement were to be designated as primary, VV seemed to be the more efficient measure based on the observed mean differences and their variances.

Prior studies of FTLD included fewer subjects, but the range of annual change in WBV and VV were similar to the values reported here. A study from London, UK,6 found that the annual rate of WBV loss in a group of 30 patients with FTLD of various clinical subtypes was 3.15 ± 2.08%, slightly larger in magnitude and variation than our finding of 1.67 ± 1.04%. Another recent study from London in 21 patients with SMD found the rate of change in WBV to be 39.6 ± 31.9 mL and in VV to be 8.9 ± 4.4 mL.7 Our values were somewhat smaller for WBV, 21.8 ± 14.4 mL, but similar for VV, 7.4 ± 4.0 mL. Some of the difference in the rate of brain volume loss could have arisen because of differences in inclusion criteria or clinical severity. However, simple correlational analyses in our cohort provided little support for an impact of baseline disease severity, as measured by baseline brain atrophy or brain volume, on rates of change in WBV and VV in our group of patients. A prior study of a small group of patients with pathologically proved FTLD with ubiquitin inclusions from our laboratory has shown that the rate of WBV seemed to accelerate over time.4 If patients with an earlier stage of disease indeed have lower rates of brain loss, trialists need to take that into account in designing future trials in FTLD. Patients who are at the earliest stages of the disease may be preferred for conceptual reasons, but larger numbers would need to be recruited to accommodate a smaller decline in brain volume over time. Our inclusion criteria required that patients have focal atrophy or focal hypometabolism on imaging at baseline to avoid patients with nonprogressive conditions. Had we included patients with normal imaging, we almost certainly would have found smaller rates of brain volume loss and ventricular expansion.19 Nonetheless, there were patients, particularly among the bvFTD group, who experienced brain volume loss and ventricular expansion that was within 2 SDs of normal controls from our laboratory.

Sample size calculations using our estimates of change and their variability show that imaging features seem to be more efficient than clinical features,8 comparing equivalent effect sizes. The sample size needed to detect a conservative moderate effect was 99 subjects per group for the FTLD-CDR and 69 for the global cognitive composite,8 whereas the equivalent sample sizes for WBV and VV were 41 and 27 subjects. Because there is no experience with a successful therapy in FTLD, it is impossible to know what kind of therapeutic effect sizes are feasible. Moreover, it is also not known whether a successful therapy would have the same magnitude of effect on clinical and imaging measures. It seems plausible that a therapy that produced a small change in brain volume, for example, could produce a moderate or large effect on a clinical marker. In contrast, it seems unlikely that the opposite would occur.

Prior longitudinal imaging studies of patients with dementia and autopsy-proved corticobasal degeneration (CBD),5 a tauopathy related to FTLD, showed that the rate of WBV loss and ventricular expansion exceeded the rates we show here for the FTLD cognitive syndromes. In five CBD patients, the rate of WBV loss was 2.3% per year, and the VV expansion was 16.2% per year. In contrast, rates for patients with dementia and progressive supranuclear palsy (PSP) were 1% for WBV and 10.9% for VV.5 Another study of PSP noted a rate for WBV loss of 1.2%.20 Uncertainty about the range of severity of cases in other studies makes it difficult to draw conclusions about differential atrophy rates across different syndromes in the FTLD-tauopathy spectrum, but if CBD patients were to be included in an FTLD clinical trial, our sample size estimates would be appropriate but conservative. For PSP, larger sample sizes would be needed.

The correlations between changes in imaging features and changes in clinical measures provide support for the use of brain imaging as a surrogate outcome measure in clinical trials of FTLD. However, the key qualification of imaging as a surrogate marker—demonstration that a treatment-induced change in imaging correlates with treatment-induced changes in cognition and behavior—cannot be addressed with this study. Because only approximately 25% of the variance in WBV changes is shared with the FTLD-CDR change score, it is not a certainty that changes in brain imaging will correlate with treatment effects in therapeutic trials.

In comparison with AD,21 the amount of annual change in brain volume and VV seems to be greater in our sample of subjects with FTLD. Among patients with “fast-progressing” AD,21 the annualized percent change in WBV was 1.4%, whereas the mean change in our patients with FTLD was 1.6%. Similarly, the annualized percent change in VV in the patients with fast-progressing AD21 was 6.4%, compared with 11.6% in our patients with FTLD. This observation must be tempered by the possibility that recruitment criteria for different studies may lead to enrollment of subjects at different stages of their disease. Our patients with FTLD may be at a point in their disease where loss of brain volume is high, whereas the patients with AD, drawn from the Mayo Alzheimer's Disease Research Center, may have been at some milder level. Taking the differences between patients with FTLD and patients with AD at face value, however, suggests that the higher rate of brain atrophy in FTLD makes this disorder a more favorable one in which to use brain imaging as a clinical trial outcome measure.

The amount of change among patients with FTLD is far greater than that seen in normal elderly subjects in our own group or in published series (0.5% for WBV20,21 and 1.7% for VV21). Despite our attempts to identify patients with progressive disease, some of our patients with FTLD had small changes in WBV and VV that were in the range seen in the normal controls. We cannot say at this point whether subjects with FTLD and slower rates of brain volume loss have different pathologic processes than the other patients.

A weakness of using WBV and VV, as opposed to regional volumetric measures, is that the whole brain measures may be more insensitive to disease-related brain deterioration. SMD may be the most dramatic example of an FTLD characterized by regionally specific anterior temporal lobe atrophy.10,22,23 The recent study that used manual segmentation of the temporal lobe for a longitudinal study of patients with SMD found that the value of the mean change/SD (Cohen d) was 1.24 for WBV but 2.33 for left temporal lobe volume.7 These values indicate that the use of temporal lobe volumetric measurements would have far more power to detect change than WBV, at least in patients with SMD. A similar regional predilection for bvFTD or PNFA might also favor regional measurements over global ones. Measurement approaches are more complicated in bvFTD, however, because the regional involvement might involve any or all of cortices in orbital frontal, anterior cingulate, anterior insula, or anterior temporal locations bilaterally.24,25 The situation in PNFA is equally challenging for regional volumetry because the inferior frontal and insular regions particularly involved in PNFA10,26 would be difficult to delineate reliably by manual tracing. However, syndrome-specific regions, once known, could be embedded in a template that could then be used for automated regional morphometric measurements. Finally, there remains the problem of combining, in a clinical trial, patients with identical molecular pathologies who happen to have different clinical and anatomic signatures. With a variety of brain areas potentially affected, regional measurement perhaps could be individualized at entry into the trial. However, that might increase rather than decrease between subject variability. Further analyses are needed to determine which approach, whole brain or regional measurements, has the greatest reliability, precision, and effect size magnitude.


Statistical analyses were performed by N. Mercaldo under the supervision of D. Knopman.


D.S.K. has served on a Data Safety Monitoring Board for Sanofi-Aventis Pharmaceuticals and received grant support from R01AG023195, U01AG06786, P50AG16574, Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program, Elan Pharmaceuticals, and Forest Pharmaceuticals. C.R.J. has served as a consultant to GE Healthcare. He received grant support from R01AG023195, P50 AG16574, U01 AG06786, R01 AG11378, and AG024904 and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program. J.H.K. received grant support from R01AG023195, P50-AG023501, P01 AG19724, NINDS HHSN271200623661C, and NIA AG22983. B.F.B. has received a consulting fee from GE Healthcare and received grant support from R01AG023195, U01AG06786, P50AG16574, the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program, and Myriad Pharmaceuticals. R.J.C. has served as a consultant to Medivation and received grant support from R01AG023195, P30AG19610, and R01MH57899. N.R.G.-R. has served as a consultant to Codman and received grant support from R01AG023195, R01AG06656, K01NS054722, U24AG26395, R01AG015866, U01AG24904, P50AG16574, R01AG015922, the Robert H. Smith Foundation, Myriad Pharmaceuticals, Elan Pharmaceutical Research, and Pfizer Pharmaceuticals. M.F.M. received grant support from the Veteran's Administration, NIH grants R01AG023195, P01AG25831, R01AG26938, and P50AG016570, and Forest Laboratories. B.L.M. has served on speakers' bureaus for Pfizer and Novartis and received grant support from P50-AG023501 and P01 AG19724. N.D.M. was supported by U01 AG016976 (National Alzheimer's Coordinating Center).

Supplementary Material

[Data Supplement]


Address correspondence and reprint requests to Dr. David Knopman, Department of Neurology, Mayo Clinic College of Medicine, 200 First St. SW, Rochester MN 55905 ude.oyam@namponk

Supplemental data at

Supported by National Institute on Aging studies R01-AG023195, R01-AG11378, P50-AG 16574 (Mayo Alzheimer's Disease Research Center), P30-AG19610 (Arizona Alzheimer Disease Center), P50-AG016570 (UCLA Alzheimer Disease Research Center), P50-AG023501 (UCSF Alzheimer Disease Research Center), and U01 AG016976 (National Alzheimer's Coordinating Center).

Disclosures: Author disclosures are provided at the end of the article.

Portions of this work were presented at the American Academy of Neurology, Chicago, IL, April 23, 2008.

Received December 8, 2008. Accepted in final form February 20, 2009.


1. Jack CR Jr, Slomkowski M, Gracon S, et al. MRI as a biomarker of disease progression in a therapeutic trial of milameline for AD. Neurology 2003;60:253–260. [PMC free article] [PubMed]
2. Fox NC, Black RS, Gilman S, et al. Effects of Aβ immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology 2005;64:1563–1572. [PubMed]
3. Jack CR Jr, Petersen RC, Grundman M, et al. Longitudinal MRI findings from the vitamin E and donepezil treatment study for MCI. Neurobiol Aging 2008;29:1285–1295. [PMC free article] [PubMed]
4. Whitwell JL, Jack CR Jr, Pankratz VS, et al. Rates of brain atrophy over time in autopsy-proven frontotemporal dementia and Alzheimer disease. Neuroimage 2008;39:1034–1040. [PMC free article] [PubMed]
5. Whitwell JL, Jack CR Jr, Parisi JE, et al. Rates of cerebral atrophy differ in different degenerative pathologies. Brain 2007;130:1148–58. [PMC free article] [PubMed]
6. Chan D, Fox NC, Jenkins R, Scahill RI, Crum WR, Rossor MN. Rates of global and regional cerebral atrophy in AD and frontotemporal dementia. Neurology 2001;57:1756–1763. [PubMed]
7. Rohrer JD, McNaught E, Foster J, et al. Tracking progression in frontotemporal lobar degeneration: serial MRI in semantic dementia. Neurology 2008;71:1445–1451. [PubMed]
8. Knopman DS, Kramer JH, Boeve BF, et al. Development of methodology for conducting clinical trials in frontotemporal lobar degeneration. Brain 2008;131:2957–2968. [PMC free article] [PubMed]
9. Galton CJ, Patterson K, Xuereb JH, Hodges JR. Atypical and typical presentations of Alzheimer's disease: a clinical, neuropsychological, neuroimaging and pathological study of 13 cases. Brain 2000;123:484–498. [PubMed]
10. Gorno-Tempini ML, Dronkers NF, Rankin KP, et al. Cognition and anatomy in three variants of primary progressive aphasia. Ann Neurol 2004;55:335–346. [PMC free article] [PubMed]
11. Knibb JA, Xuereb JH, Patterson K, Hodges JR. Clinical and pathological characterization of progressive aphasia. Ann Neurol 2006;59:156–165. [PubMed]
12. Josephs KA, Whitwell JL, Duffy JR, et al. Progressive aphasia secondary to Alzheimer disease vs FTLD pathology. Neurology 2008;70:25–34. [PMC free article] [PubMed]
13. Schneider LS, Olin JT, Doody RS, et al. Validity and reliability of the Alzheimer's Disease Cooperative Study–Clinical Global Impression of Change: the Alzheimer's Disease Cooperative Study. Alzheimer Dis Assoc Disord 1997;11 (suppl 2):S22–S32. [PubMed]
14. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–198. [PubMed]
15. Freeborough PA, Fox NC. The boundary shift integral: an accurate and robust measure of cerebral volume changes from registered repeat MRI. IEEE Trans Med Imag 1997;15:623–629. [PubMed]
16. Gunter JL, Shiung MM, Manduca A, Jack CR Jr. Methodological considerations for measuring rates of brain atrophy. JMRI 2003;18:16–24. [PMC free article] [PubMed]
17. Jack CR Jr, Shiung MM, Weigand SD, et al. Brain atrophy rates predict subsequent clinical conversion in normal elderly and amnestic MCI. Neurology 2005;65:1227–1231. [PMC free article] [PubMed]
18. Jack CR Jr, Petersen RC, O'Brien PC, Tangalos EG. MR-based hippocampal volumetry in the diagnosis of Alzheimer's disease. Neurology 1992;42:183–188. [PubMed]
19. Kipps CM, Nestor PJ, Dawson CE, Mitchell J, Hodges JR. Measuring progression in frontotemporal dementia: implications for therapeutic interventions. Neurology 2008;70:2046–2052. [PubMed]
20. Paviour DC, Price SL, Jahanshahi M, Lees AJ, Fox NC. Longitudinal MRI in progressive supranuclear palsy and multiple system atrophy: rates and regions of atrophy. Brain 2006;129:1040–1049. [PubMed]
21. Jack CR Jr, Shiung MM, Gunter JL, et al. Comparison of different MRI brain atrophy rate measures with clinical disease progression in AD. Neurology 2004;62:591–600. [PMC free article] [PubMed]
22. Chan D, Fox NC, Scahill RI, et al. Patterns of temporal lobe atrophy in semantic dementia and Alzheimer's disease. Ann Neurol 2001;49:433–442. [PubMed]
23. Galton CJ, Patterson K, Graham K, et al. Differing patterns of temporal atrophy in Alzheimer's disease and semantic dementia. Neurology 2001;57:216–225. [PubMed]
24. Rosen HJ, Gorno-Tempini ML, Goldman WP, et al. Patterns of brain atrophy in frontotemporal dementia and semantic dementia. Neurology 2002;58:198–208. [PubMed]
25. Brambati SM, Renda NC, Rankin KP, et al. A tensor based morphometry study of longitudinal gray matter contraction in FTD. Neuroimage 2007;35:998–1003. [PMC free article] [PubMed]
26. Josephs KA, Duffy JR, Strand EA, et al. Clinicopathological and imaging correlates of progressive aphasia and apraxia of speech. Brain 2006;129:1385–1398. [PMC free article] [PubMed]

Articles from Neurology are provided here courtesy of American Academy of Neurology