A.M. Tiehuis

Progression of cerebral atrophy and white matter hyperintensities in patients with type 2 diabetes

OBJECTIVE Type 2 diabetes is associated with a moderate degree of cerebral atrophy and a higher white matter hyperintensity (WMH) volume. How these brain-imaging abnormalities evolve over time is unknown. The present study aims to quantify cerebral atrophy and WMH progression over 4 years in type 2 diabetes. RESEARCH DESIGN AND METHODS A total of 55 patients with type 2 diabetes and 28 age-, sex-, and IQ-matched control participants had two 1.5T magnetic resonance imaging scans with a 4-year interval. Volumetric measurements of total brain, peripheral cerebrospinal fluid (CSF), lateral ventricles, and WMH were performed with k-nearest neighbor–based probabilistic segmentation. All volumes were expressed as percentage of intracranial volume. Linear regression analyses, adjusted for age and sex, were performed to compare brain volumes between the groups and to identify determinants of volumetric change within the type 2 diabetic group. RESULTS At baseline, patients with type 2 diabetes had a significantly smaller total brain volume and larger peripheral CSF volume than control participants. In both groups, all volumes showed a significant change over time. Patients with type 2 diabetes had a greater increase in lateral ventricular volume than control participants (mean adjusted between-group difference in change over time [95% CI]: 0.11% in 4 years [0.00 to 0.22], P = 0.047). CONCLUSIONS The greater increase in lateral ventricular volume over time in patients with type 2 diabetes compared with control participants shows that type 2 diabetes is associated with a slow increase of cerebral atrophy over the course of years.

Diabetes Increases Atrophy and Vascular Lesions on Brain MRI in Patients With Symptomatic Arterial Disease

Background and Purpose— Diabetes type 2 (DM2) is associated with accelerated cognitive decline and structural brain abnormalities. Macrovascular disease has been described as a determinant for brain MRI changes in DM2, but little is known about the involvement of other DM2-related factors. Methods— Brain MRI was performed in 1043 participants (151 DM2) with symptomatic arterial disease. Brain volumes were obtained through automated segmentation. Results— Patients with arterial disease and DM2 had more global and subcortical brain atrophy (−1.20% brain/intracranial volume [95%CI −1.58 to −0.82], P<0.0005 and 0.20% ventricular/intracranial volume [0.05 to 0.34], P<0.01), larger WMH volumes (0.22 logtransformed volume [0.07 to 0.38], P<0.005), and more lacunar infarcts (OR 1.75 [1.13 to 2.69], P<0.01) than identical patients without DM2. In patients with DM2, high glucose levels (B−0.12% per mmol/L [−0.23 to −0.01], P<0.05) and diabetes duration (B−0.05% per year [−0.10 to −0.001], P<0.05) were associated with global brain atrophy. Conclusion— In patients with symptomatic arterial disease, DM2 has an added detrimental effect on the brain. In patients with DM2, hyperglycemia and diabetes duration contribute to brain atrophy.

Total cerebral blood flow, white matter lesions and brain atrophy: the SMART-MR study

We investigated whether total cerebral blood flow (CBF) was associated with brain atrophy, and whether this relation was modified by white matter lesions (WML). Within the Second Manifestations of ARTerial disease-magnetic resonance (SMART-MR) study, a prospective cohort study among patients with arterial disease, cross-sectional analyses were performed in 828 patients (mean age 58±10 years, 81% male) with quantitative flow, atrophy, and WML measurements on magnetic resonance imaging (MRI). Total CBF was measured with MR angiography and was expressed per 100 mL brain volume. Total brain volume and ventricular volume were divided by intracranial volume to obtain brain parenchymal fraction (BPF) and ventricular fraction (VF). Lower BPF indicates more global brain atrophy, whereas higher VF indicates more subcortical brain atrophy. Mean CBF was 52.0±10.2 mL/min per 100 mL, mean BPF was 79.2±2.9%, and mean VF was 2.03±0.96%. Linear regression analyses showed that lower CBF was associated with more subcortical brain atrophy, after adjusting for age, sex, vascular risk factors, intima-media thickness, and lacunar infarcts, but only in patients with moderate to severe WML (upper quartile of WML): Change in VF per s.d. decrease in CBF 0.18%, 95% CI: 0.02 to 0.34%. Our findings suggest that cerebral hypoperfusion in the presence of WML may be associated with subcortical brain atrophy.

Automated and Visual Scoring Methods of Cerebral White Matter Hyperintensities: Relation with Age and Cognitive Function

Background and Purpose: A reliable scoring method for ischemic cerebral white matter hyperintensities (WMH) will help to clarify the causes and consequences of these brain lesions. We compared an automated and two visual WMH scoring methods in their relations with age and cognitive function. Methods: MRI of the brain was performed on 154 participants of the Utrecht Diabetic Encephalopathy Study. WMH volumes were obtained with an automated segmentation method. Visual rating of deep and periventricular WMH (DWMH and PWMH) was performed with the Scheltens scale and the Rotterdam Scan Study (RSS) scale, respectively. Cognition was assessed with a battery of 11 tests. Results: Within the whole study group, the association with age was most evident for the automated measured WMH volume (β = 0.43, 95% CI = 0.29–0.57). With regard to cognition, automated measured WMH volume and Scheltens DWMH were significantly associated with information processing speed (β = –0.22, 95% CI = –0.40 to –0.06; β = –0.26, 95% CI = –0.42 to –0.10), whereas RSS PWMH were associated with attention and executive function (β = –0.19, 95% CI = –0.36 to –0.02). Conclusion: Measurements of WMH with an automated quantitative segmentation method are comparable with visual rating scales and highly suitable for use in future studies to assess the relationship between WMH and subtle impairments in cognitive function.

Metabolic Syndrome, Prediabetes, and Brain Abnormalities on MRI in Patients With Manifest Arterial Disease: The SMART-MR Study

OBJECTIVE Metabolic syndrome (MetS) is a cluster of cardiovascular risk factors leading to atherosclerosis and diabetes. Diabetes is associated with both structural and functional abnormalities of the brain. MetS, even before diabetes is diagnosed, may also predispose to cerebral changes, probably through shared mechanisms. We examined the association of MetS with cerebral changes in patients with manifest arterial disease. RESEARCH DESIGN AND METHODS Cross-sectional data on MetS and brain MRI were available in 1,232 participants with manifest arterial disease (age 58.6 ± 10.1 years; 37% MetS). Volumes of brain tissue, ventricles, and white matter hyperintensities (WMH) were obtained by automated segmentation and expressed relative to intracranial volume. Infarcts were distinguished into lacunar and nonlacunar infarcts. RESULTS The presence of MetS (n = 451) was associated with smaller brain tissue volume (B −0.72% [95% CI −0.97, −0.47]), even in the subgroup of patients without diabetes (B −0.42% [95% CI −0.71, −0.13]). MetS was not associated with an increased occurrence of WMH or cerebral infarcts. Impaired glucose metabolism, abdominal obesity, and elevated triglycerides were individual components associated with smaller brain volume. Obesity and hypertriglyceridemia remained associated with smaller brain volume when patients with diabetes were excluded. Hypertension was associated with an increased occurrence of WMH and infarcts. CONCLUSIONS In patients with manifest arterial disease, presence of MetS is associated with smaller brain volume, even in patients without diabetes. Screening for MetS and treatment of its individual components, in particular, hyperglycemia, hypertriglyceridemia, and obesity, may prevent progression of cognitive aging in patients with MetS, even in a prediabetic stage.

Cognitive dysfunction and its clinical and radiological determinants in patients with symptomatic arterial disease and diabetes.

BACKGROUND AND AIMS Both vascular disease and diabetes type 2 (DM2) decrease cognitive functioning in elderly people. It is uncertain if DM2 affects cognition independent of vascular disease. In patients with symptomatic arterial disease, we studied the effect of DM2 on cognition and identified clinical and radiological determinants for impaired cognition in patients with DM2. METHODS 766 patients (mean age 58.8+/-9.5 years; 108 DM2) with symptomatic arterial disease underwent neuropsychological testing. In 542 patients (77 DM2), volumes of brain tissue, ventricles and white matter lesions were obtained by segmentation of brain MR images. Infarcts were distinguished into small (lacunar) or large (cortical or subcortical). RESULTS Patients with arterial disease and DM2 performed worse on neuropsychological tests compared to similar patients without DM2 (adjusted composite z-score: beta -0.14 [-0.25 to -0.02]). Insulin treatment, systolic and diastolic blood pressures were significantly associated with cognition in patients with DM2. Large infarcts, global and cortical atrophy on MRI were independently associated with cognition in patients with DM2. CONCLUSION The presence of DM2 in patients with symptomatic arterial disease is associated with decreased cognitive functioning. Insulin treatment, high blood pressure, brain atrophy and large infarcts were determinants for cognitive dysfunction in patients with DM2 and arterial disease.

Semi-automatic classification of textures in thoracic CT scans.

The textural patterns in the lung parenchyma, as visible on computed tomography (CT) scans, are essential to make a correct diagnosis in interstitial lung disease. We developed one automatic and two interactive protocols for classification of normal and seven types of abnormal lung textures. Lungs were segmented and subdivided into volumes of interest (VOIs) with homogeneous texture using a clustering approach. In the automatic protocol, VOIs were classified automatically by an extra-trees classifier that was trained using annotations of VOIs from other CT scans. In the interactive protocols, an observer iteratively trained an extra-trees classifier to distinguish the different textures, by correcting mistakes the classifier makes in a slice-by-slice manner. The difference between the two interactive methods was whether or not training data from previously annotated scans was used in classification of the first slice. The protocols were compared in terms of the percentages of VOIs that observers needed to relabel. Validation experiments were carried out using software that simulated observer behavior. In the automatic classification protocol, observers needed to relabel on average 58% of the VOIs. During interactive annotation without the use of previous training data, the average percentage of relabeled VOIs decreased from 64% for the first slice to 13% for the second half of the scan. Overall, 21% of the VOIs were relabeled. When previous training data was available, the average overall percentage of VOIs requiring relabeling was 20%, decreasing from 56% in the first slice to 13% in the second half of the scan.

Optimization Strategies for Interactive Classification of Interstitial Lung Disease Textures

For computerized analysis of textures in interstitial lung disease, manual annotations of lung tissue are necessary. Since making these annotations is labor-intensive, we previously proposed an interactive annotation framework. In this framework, observers iteratively trained a classifier to distinguish the different texture types by correcting its classification errors. In this work, we investigated three ways to extend this approach, in order to decrease the amount of user interaction required to annotate all lung tissue in a CT scan. First, we conducted automatic classification experiments to test how data from previously annotated scans can be used for classification of the scan under consideration. We compared the performance of a classifier trained on data from one observer, a classifier trained on data from multiple observers, a classifier trained on consensus training data, and an ensemble of classifiers, each trained on data from different sources. Experiments were conducted without and with texture selection. In the former case, training data from all 8 textures was used. In the latter, only training data from the texture types present in the scan were used, and the observer would have to indicate textures contained in the scan to be analyzed. Second, we simulated interactive annotation to test the effects of (1) asking observers to perform texture selection before the start of annotation, (2) the use of a classifier trained on data from previously annotated scans at the start of annotation, when the interactive classifier is untrained, and (3) allowing observers to choose which interactive or automatic classification results they wanted to correct. Finally, various strategies for selecting the classification results that were presented to the observer were considered. Classification accuracies for all possible interactive annotation scenarios were compared. Using the best performing protocol, in which observers select the textures that should be distinguished in the scan and in which they can choose which classification results to use for correction, a median accuracy of 88% was reached. The results obtained using this protocol were significantly better than results obtained with other interactive or automatic classification protocols.