Standardized evaluation of algorithms for computer-aided diagnosis of dementia based on structural MRI: the CADDementia challenge

Algorithms for computer-aided diagnosis of dementia based on structural MRI have demonstrated high performance in the literature, but are difficult to compare as different data sets and methodology were used for evaluation. In addition, it is unclear how the algorithms would perform on previously unseen data, and thus, how they would perform in clinical practice when there is no real opportunity to adapt the algorithm to the data at hand. To address these comparability, generalizability and clinical applicability issues, we organized a grand challenge that aimed to objectively compare algorithms based on a clinically representative multi-center data set. Using clinical practice as the starting point, the goal was to reproduce the clinical diagnosis. Therefore, we evaluated algorithms for multi-class classification of three diagnostic groups: patients with probable Alzheimer's disease, patients with mild cognitive impairment and healthy controls. The diagnosis based on clinical criteria was used as reference standard, as it was the best available reference despite its known limitations. For evaluation, a previously unseen test set was used consisting of 354 T1-weighted MRI scans with the diagnoses blinded. Fifteen research teams participated with a total of 29 algorithms. The algorithms were trained on a small training set (n=30) and optionally on data from other sources (e.g., the Alzheimer's Disease Neuroimaging Initiative, the Australian Imaging Biomarkers and Lifestyle flagship study of aging). The best performing algorithm yielded an accuracy of 63.0% and an area under the receiver-operating-characteristic curve (AUC) of 78.8%. In general, the best performances were achieved using feature extraction based on voxel-based morphometry or a combination of features that included volume, cortical thickness, shape and intensity. The challenge is open for new submissions via the web-based framework: http://caddementia.grand-challenge.org.

Brain region's relative proximity as marker for Alzheimer's disease based on structural MRI

BACKGROUND

Alzheimer's disease (AD) is a progressive, incurable neurodegenerative disease and the most common type of dementia. It cannot be prevented, cured or drastically slowed, even though AD research has increased in the past 5-10 years. Instead of focusing on the brain volume or on the single brain structures like hippocampus, this paper investigates the relationship and proximity between regions in the brain and uses this information as a novel way of classifying normal control (NC), mild cognitive impaired (MCI), and AD subjects.

METHODS

A longitudinal cohort of 528 subjects (170 NC, 240 MCI, and 114 AD) from ADNI at baseline and month 12 was studied. We investigated a marker based on Procrustes aligned center of masses and the percentile surface connectivity between regions. These markers were classified using a linear discriminant analysis in a cross validation setting and compared to whole brain and hippocampus volume.

RESULTS

We found that both our markers was able to significantly classify the subjects. The surface connectivity marker showed the best results with an area under the curve (AUC) at 0.877 (p<0.001), 0.784 (p<0.001), 0,766 (p<0.001) for NC-AD, NC-MCI, and MCI-AD, respectively, for the functional regions in the brain. The surface connectivity marker was able to classify MCI-converters with an AUC of 0.599 (p<0.05) for the 1-year period.

CONCLUSION

Our results show that our relative proximity markers include more information than whole brain and hippocampus volume. Our results demonstrate that our proximity markers have the potential to assist in early diagnosis of AD.

Remote Ischemic Perconditioning in Thrombolysed Stroke Patients: Randomized Study of Activating Endogenous Neuroprotection – Design and MRI Measurements

Background

Intravenous administration of alteplase is the only approved treatment for acute ischemic stroke. Despite the effectiveness of this treatment, 50% of patients suffer chronic neurological disability, which may in part be caused by ischemia-reperfusion injury. Remote ischemic perconditioning, performed as a transient ischemic stimulus by blood-pressure cuff inflation to an extremity, has proven effective in attenuating ischemia-reperfusion injury in animal models of stroke. Remote ischemic perconditioning increases myocardial salvage in patients undergoing acute revascularization for acute myocardial infarction. To clarify whether a similar benefit can be obtained in patients undergoing thrombolysis for acute stroke, we included patients from June 2009 to January 2011.

Aim and design

The aims of the study are: to estimate the effect of remote ischemic perconditioning as adjunctive therapy to intravenous alteplase of acute ischemic stroke within the 4½-h time window and to investigate the feasibility of remote ischemic perconditioning performed during transport to hospital in patients displaying symptoms of acute stroke. Patients are randomized to remote ischemic perconditioning in a single-blinded fashion during transportation to hospital. Only patients with magnetic resonance imaging-proven ischemic stroke, who subsequently are treated with intravenous alteplase, and in selected cases additional endovascular treatment, are finally included in the study.

Study outcomes

Primary end-point is penumbral salvage. Penumbra is defined as hypoperfused yet viable tissue identified as the mismatch between perfusion-weighted imaging and diffusion-weighted imaging lesion on magnetic resonance imaging scans. Primary outcome is a mismatch volume not progressing to infarction on one-month follow-up T2 fluid attenuated inversion recovery. Secondary end-points include: infarct growth (expansion of the diffusion-weighted imaging lesion) from baseline to the 24-h and one-month follow-up examination. Infarct growth inside and outside the acute perfusion-weighted imaging–diffusion-weighted imaging mismatch zone is quantified by use of coregistration. Clinical outcome after three-months. The influence of physical activity (Physical Activity Scale for the Elderly score) on effect of remote ischemic perconditioning. Feasibility of remote ischemic perconditioning in acute stroke patients.

Summary

This phase 3 trial is the first study in patients with acute ischemic stroke to evaluate the effect size of remote ischemic perconditioning as a pretreatment to intravenous alteplase, measured as penumbral salvage on multimodal magnetic resonance imaging and clinical outcome after three-months follow-up.

Texture-based analysis of COPD: a data-driven approach

This study presents a fully automatic, data-driven approach for texture-based quantitative analysis of chronic obstructive pulmonary disease (COPD) in pulmonary computed tomography (CT) images. The approach uses supervised learning where the class labels are, in contrast to previous work, based on measured lung function instead of on manually annotated regions of interest (ROIs). A quantitative measure of COPD is obtained by fusing COPD probabilities computed in ROIs within the lung fields where the individual ROI probabilities are computed using a k nearest neighbor (kNN ) classifier. The distance between two ROIs in the kNN classifier is computed as the textural dissimilarity between the ROIs, where the ROI texture is described by histograms of filter responses from a multi-scale, rotation invariant Gaussian filter bank. The method was trained on 400 images from a lung cancer screening trial and subsequently applied to classify 200 independent images from the same screening trial. The texture-based measure was significantly better at discriminating between subjects with and without COPD than were the two most common quantitative measures of COPD in the literature, which are based on density. The proposed measure achieved an area under the receiver operating characteristic curve (AUC) of 0.713 whereas the best performing density measure achieved an AUC of 0.598. Further, the proposed measure is as reproducible as the density measures, and there were indications that it correlates better with lung function and is less influenced by inspiration level.

Dissimilarity-based classification of anatomical tree structures

A novel method for classification of abnormality in anatomical tree structures is presented. A tree is classified based on direct comparisons with other trees in a dissimilarity-based classification scheme. The pair-wise dissimilarity measure between two trees is based on a linear assignment between the branch feature vectors representing those trees. Hereby, localized information in the branches is collectively used in classification and variations in feature values across the tree are taken into account. An approximate anatomical correspondence between matched branches can be achieved by including anatomical features in the branch feature vectors. The proposed approach is applied to classify airway trees in computed tomography images of subjects with and without chronic obstructive pulmonary disease (COPD). Using the wall area percentage (WA%), a common measure of airway abnormality in COPD, as well as anatomical features to characterize each branch, an area under the receiver operating characteristic curve of 0.912 is achieved. This is significantly better than computing the average WA%.

The expression and regulation of bone-acting cytokines in human peripheral adipose tissue in organ culture

The humoral cross-talk between bone and fat is an area of increasing interest. We investigated the expression and regulation of the bone-acting cytokines; bone morphogenetic protein 2 (BMP2), connective tissue growth factor (CTGF), osteoprotegerin (OPG), and transforming growth factor beta (TGFB1). Subcutaneous adipose tissue was aspirated from lean, healthy women. Tissue samples were incubated with interleukin 1-β (IL1-β), tumor necrosis factor-α (TNF-α), cortisol, troglitazone, IL1-β + troglitazone, or vehicle. Gene expression in the adipose tissue was analyzed using qPCR and protein levels in the incubation media were analyzed using ELISA. OPG expression and secretion was diminished by 40.8% and 43.1% respectively, by cortisol, and OPG expression was diminished by 67.5% by troglitazone (p<0.05). The proinflammatory cytokines IL1-β and TNF-α significantly increased the expression of CTGF (p<0.05) by 65.1% and 101.3%, respectively, and the expression and secretion of OGP by 62.3-165.8% (p<0.05). This interleukin 1-β mediated increase in CTGF- and OPG expression and secretion was ameliorated by troglitazone. Troglitazone and related drugs are known to have adverse effects on bone. We suggest that this could be mediated via altered cytokine production in adipose tissue. Moreover, obese individuals have a low-grade inflammation in their adipose tissue and have higher bone mineral density than lean individuals. We suggest that this inflammation may increase the expression and secretion of OPG and CTGF and thereby increase BMD. In conclusion, bone acting cytokines are produced in the adipose tissue and may affect bone through endocrine mechanisms.

Systemic thrombolytic therapy alone and in combination with mechanical revascularization in acute ischemic stroke in two children

Thrombolytic therapy is not recommended for acute ischemic stroke (AIS) in patients under the age of 18 and published experience is limited. In this case report, we describe two children treated with systemic thrombolytic therapy. One child received additional mechanical revascularization and achieved a good clinical outcome. The differences in the fibrinolytic system and the different etiology of AIS in childhood may limit a simple extrapolation of the adult guidelines for systemic thrombolytic therapy. Acute multimodal imaging to clarify the etiology of AIS might help to select the most appropriate treatment modality.

Localization of white-matter lesions and effect of vascular risk factors in late-onset major depression

BACKGROUND

Several studies suggest that patients with late-onset major depression (MD) have an increased load of cerebral white-matter lesions (WMLs) compared with age-matched controls. Vascular risk factors such as hypertension and smoking may confound such findings. Our aim was to investigate the association between the localization and load of WMLs in late-onset MD with respect to vascular risk factors.

METHOD

We examined 22 consecutive patients with late-onset first-episode MD and 22 age- and gender-matched controls using whole-brain magnetic resonance imaging (MRI). The localization, number and volume of WMLs were compared between patients and controls, while testing the effect of vascular risk factors.

RESULTS

Among subjects with one or more WMLs, patients displayed a significantly higher WML density in two white-matter tracts: the left superior longitudinal fasciculus and the right frontal projections of the corpus callosum. These tracts are part of circuitries essential for cognitive and emotional functions. Analyses revealed no significant difference in the total number and volume of WMLs between groups. Patients and controls showed no difference in vascular risk factors, except for smoking. Lesion load was highly correlated with smoking.

CONCLUSIONS

Our results indicate that lesion localization rather than lesion load differs between patients with late-onset MD and controls. Increased lesion density in regions associated with cognitive and emotional functions may be crucial in late-onset MD, and vascular risk factors such as smoking may play an important role in the pathophysiology of late-onset MD, consistent with the vascular depression hypothesis.

Quantitative analysis of pulmonary emphysema using local binary patterns

We aim at improving quantitative measures of emphysema in computed tomography (CT) images of the lungs. Current standard measures, such as the relative area of emphysema (RA), rely on a single intensity threshold on individual pixels, thus ignoring any interrelations between pixels. Texture analysis allows for a much richer representation that also takes the local structure around pixels into account. This paper presents a texture classification-based system for emphysema quantification in CT images. Measures of emphysema severity are obtained by fusing pixel posterior probabilities output by a classifier. Local binary patterns (LBP) are used as texture features, and joint LBP and intensity histograms are used for characterizing regions of interest (ROIs). Classification is then performed using a k nearest neighbor classifier with a histogram dissimilarity measure as distance. A 95.2% classification accuracy was achieved on a set of 168 manually annotated ROIs, comprising the three classes: normal tissue, centrilobular emphysema, and paraseptal emphysema. The measured emphysema severity was in good agreement with a pulmonary function test (PFT) achieving correlation coefficients of up to |r| = 0.79 in 39 subjects. The results were compared to RA and to a Gaussian filter bank, and the texture-based measures correlated significantly better with PFT than did RA.

Assessment of dementia in nursing home residents by nurses and assistants: criteria validity and determinants

OBJECTIVES

To describe the criterion validity of nursing home staff's assessment of organic disorder compared with ICD-10 criteria, and to identify determinants of staff assessment of organic disorder.

METHOD

Two hundred and eighty-eight residents were diagnosed using the GMS-AGECAT. Nursing staff members were interviewed about the residents' activities of Daily Living, behavioural problems, orientation in surroundings and communication skills, and asked if the resident had an organic disorder. Multiple logistic regression was used to select the items that most strongly determined staff assessment of organic disorder.

RESULTS

Sixty-two per cent of the residents were diagnosed by GMS-AGECAT as having organic disorder, 78% of these were correctly identified by the staff. Whether analysed among residents with or without organic disorder, or in the total group of residents, the staff assessment of the presence of organic disorder depended on a limited set of behavioural characteristics of the resident, namely 'going to the toilet in inappropriate places', 'saying things that do not make sense' and impairment in orientation.

CONCLUSIONS

Staff comprehension of organic disorder resulted in over- as well as under-labelling of residents, a tendency that will affect communication with medical personnel and may lead to inadequate or wrong medical treatment and to negative performance as well as negative role expectations in everyday life in nursing homes.

Determinants for the use of psychotropics among nursing home residents

PURPOSE

To characterise the prescription pattern of psychotropics in Danish nursing homes and to identify diagnostic, behavioural, cognitive and performance characteristics associated with prevalent psychotropic drug use.

METHODS

Prescribed daily medication was recorded from nurses' files. Based on the Anatomical Therapeutical Chemical (ATC) classification index, psychotropics were categorised into neuroleptics, benzodiazepines and antidepressants. Two hundred and eighty-eight residents were diagnosed using the GMS-AGECAT. One hundred and eighteen staff members were interviewed about the residents's Activities of Daily Living (ADL), behavioural problems (Nursing Home Behavior Problem Scale), orientation, communication skills and if the resident had any psychiatric disorder. Multiple logistic regression was used to select the items that determined the use of psychotropics.

RESULTS

Fifty-six percent of the residents received a psychotropic, 21% received neuroleptics, 38% received benzodiazepines and 24% received antidepressants. In the multivariate analysis, staff assessment of the resident's mental health was a determinant for the use of all types of specific psychotropics, whereas a GMS-AGECAT diagnosis only determined the use of neuroleptics. Behavioural problems were a determinant for the use of neuroleptics and the use of benzodiazepines irrespective of the psychiatric diagnosis of the resident. Use of antidepressants was associated with male gender and increasing age.

CONCLUSIONS

Staff perceptions of psychiatric morbidity and norms have a greater impact on the prescription of psychotropics than standardised clinical criteria.

Determinants for the use of psychotropics among nursing home residents

PURPOSE

To characterise the prescription pattern of psychotropics in Danish nursing homes and to identify diagnostic, behavioural, cognitive and performance characteristics associated with prevalent psychotropic drug use.

METHODS

Prescribed daily medication was recorded from nurses' files. Based on the Anatomical Therapeutical Chemical (ATC) classification index, psychotropics were categorised into neuroleptics, benzodiazepines and antidepressants. Two hundred and eighty-eight residents were diagnosed using the GMS-AGECAT. One hundred and eighteen staff members were interviewed about the residents's Activities of Daily Living (ADL), behavioural problems (Nursing Home Behavior Problem Scale), orientation, communication skills and if the resident had any psychiatric disorder. Multiple logistic regression was used to select the items that determined the use of psychotropics.

RESULTS

Fifty-six percent of the residents received a psychotropic, 21% received neuroleptics, 38% received benzodiazepines and 24% received antidepressants. In the multivariate analysis, staff assessment of the resident's mental health was a determinant for the use of all types of specific psychotropics, whereas a GMS-AGECAT diagnosis only determined the use of neuroleptics. Behavioural problems were a determinant for the use of neuroleptics and the use of benzodiazepines irrespective of the psychiatric diagnosis of the resident. Use of antidepressants was associated with male gender and increasing age.

CONCLUSIONS

Staff perceptions of psychiatric morbidity and norms have a greater impact on the prescription of psychotropics than standardised clinical criteria.