An Automated Method for Segmenting White Matter Lesions through Multi-Level Morphometric Feature Classification with Application to Lupus

Ultrasound-based measures in carpal tunnel syndrome: What to measure?

Carpal tunnel syndrome (CTS) is the most common peripheral entrapment neuropathy and is caused by compression of the median nerve (MN) at the level of transverse carpal ligament of the volar wrist. Radiomics is an advanced semi-automated image analysis method that is utilized to identify characteristics in the MN that can detect CTS with considerable reproducibility.

Libman-Sacks endocarditis and associated cerebrovascular disease: The role of medical therapy

BACKGROUND

Libman-Sacks endocarditis in patients with systemic lupus erythematosus (SLE) is commonly complicated with embolic cerebrovascular disease (CVD) or valve dysfunction for which high-risk valve surgery is frequently performed. However, the role of medical therapy alone for Libman-Sacks endocarditis and associated acute CVD remains undefined.

OBJECTIVE

To determine in this cross-sectional and longitudinal study if conventional anti-inflammatory and anti-thrombotic therapy may be an effective therapy in SLE patients with Libman-Sacks endocarditis and associated acute CVD.

METHODS AND MATERIALS

17 SLE patients with Libman-Sacks endocarditis detected by two-and-three-dimensional transesophageal echocardiography (TEE) and complicated with acute CVD [stroke/TIA, focal brain injury on MRI, or cognitive dysfunction] were treated with conventional anti-inflammatory and anti-thrombotic therapy for a median of 6 months and then underwent repeat TEE, transcranial Doppler, brain MRI, and neurocognitive testing for re-assessment of Libman-Sacks endocarditis and CVD.

RESULTS

Valve vegetations decreased in number, diameter, and area (all p ≤0.01); associated valve regurgitation significantly improved (p = 0.04), and valve thickening did not progress (p = 0.56). In 13 (76%) patients, valve vegetations or valve regurgitation resolved or improved in number and size or by ≥1 degree, respectively, as compared to 4 (24%) patients in whom vegetations or valve regurgitation persisted unchanged or increased in size or by ≥1 degree (p = 0.03). Also, cerebromicroembolism, lobar and global gray and white matter cerebral perfusion, ischemic brain lesion load, and neurocognitive dysfunction resolved or significantly improved (all p ≤0.04).

CONCLUSION

These preliminary data suggest that combined conventional anti-inflammatory and antithrombotic therapy may be an effective treatment for Libman-Sacks endocarditis and its associated CVD and may obviate the need for high-risk valve surgery.

Automatic brain lesion segmentation on standard magnetic resonance images: a scoping review

OBJECTIVES

Medical image analysis practices face challenges that can potentially be addressed with algorithm-based segmentation tools. In this study, we map the field of automatic MR brain lesion segmentation to understand the clinical applicability of prevalent methods and study designs, as well as challenges and limitations in the field.

DESIGN

Scoping review.

SETTING

Three databases (PubMed, IEEE Xplore and Scopus) were searched with tailored queries. Studies were included based on predefined criteria. Emerging themes during consecutive title, abstract, methods and whole-text screening were identified. The full-text analysis focused on materials, preprocessing, performance evaluation and comparison.

RESULTS

Out of 2990 unique articles identified through the search, 441 articles met the eligibility criteria, with an estimated growth rate of 10% per year. We present a general overview and trends in the field with regard to publication sources, segmentation principles used and types of lesions. Algorithms are predominantly evaluated by measuring the agreement of segmentation results with a trusted reference. Few articles describe measures of clinical validity.

CONCLUSIONS

The observed reporting practices leave room for improvement with a view to studying replication, method comparison and clinical applicability. To promote this improvement, we propose a list of recommendations for future studies in the field.

A systematic review of the applications of artificial intelligence and machine learning in autoimmune diseases

Autoimmune diseases are chronic, multifactorial conditions. Through machine learning (ML), a branch of the wider field of artificial intelligence, it is possible to extract patterns within patient data, and exploit these patterns to predict patient outcomes for improved clinical management. Here, we surveyed the use of ML methods to address clinical problems in autoimmune disease. A systematic review was conducted using MEDLINE, embase and computers and applied sciences complete databases. Relevant papers included "machine learning" or "artificial intelligence" and the autoimmune diseases search term(s) in their title, abstract or key words. Exclusion criteria: studies not written in English, no real human patient data included, publication prior to 2001, studies that were not peer reviewed, non-autoimmune disease comorbidity research and review papers. 169 (of 702) studies met the criteria for inclusion. Support vector machines and random forests were the most popular ML methods used. ML models using data on multiple sclerosis, rheumatoid arthritis and inflammatory bowel disease were most common. A small proportion of studies (7.7% or 13/169) combined different data types in the modelling process. Cross-validation, combined with a separate testing set for more robust model evaluation occurred in 8.3% of papers (14/169). The field may benefit from adopting a best practice of validation, cross-validation and independent testing of ML models. Many models achieved good predictive results in simple scenarios (e.g. classification of cases and controls). Progression to more complex predictive models may be achievable in future through integration of multiple data types.

Correlation of neurocognitive function and brain lesion load on magnetic resonance imaging in systemic lupus erythematosus

Neurocognitive dysfunction and brain injury on magnetic resonance imaging (MRI) are common in patients with systemic lupus erythematosus (SLE) and are associated with increased morbidity and mortality. However, brain MRI is expensive, is restricted by payers, and requires high expertise. Neurocognitive assessment is an easily available, safe, and inexpensive clinical tool that may select patients needing brain MRI. In this cross-sectional and controlled study, 76 SLE patients (69 women, age 37 ± 12 years) and 26 age and gender-matched healthy subjects (22 women, age 34 ± 11 years) underwent assessment of attention, memory, processing speed, executive function, motor function, and global neurocognitive function. All subjects underwent brain MRI with T1-weighted, fluid-attenuated inversion recovery (FLAIR), and diffusion-weighted imaging. Hemispheric and whole brain lesion load in cm³ were determined using semi-automated methods. Neurocognitive z-scores in all clinical domains were significantly lower and whole brain and right and left hemispheres brain lesion load were significantly greater in patients than in controls (all p ≤ 0.02). There was significant correlation between neurocognitive z-scores in all domains and whole brain lesion load: processing speed (r = - 0.46; p < 0.0001), attention (r = - 0.42; p < 0.001), memory (r = - 0.40; p = 0.0004), executive function (r = - 0.25; p = 0.03), motor function (r = - 0.25; p = 0.05), and global neurocognitive function (r = - 0.38; p = 0.006). Similar correlations were found for brain hemisphere lesion loads (all p ≤ 0.05). These correlations were strengthened when adjusted for glucocorticoid therapy and SLE disease activity index. Finally, global neurocognitive z-score and erythrosedimentation rate were the only independent predictors of whole brain lesion load (both p ≤ 0.007). Neurocognitive measures and brain lesion load are worse in SLE patients than in controls. In SLE patients, neurocognitive z-scores correlate negatively with and independently predict brain lesion load. Therefore, neurocognitive testing may be an effective clinical tool to select patients needing brain MRI.

Automated Detection of Lupus White Matter Lesions in MRI

Brain magnetic resonance imaging provides detailed information which can be used to detect and segment white matter lesions (WML). In this work we propose an approach to automatically segment WML in Lupus patients by using T1w and fluid-attenuated inversion recovery (FLAIR) images. Lupus WML appear as small focal abnormal tissue observed as hyperintensities in the FLAIR images. The quantification of these WML is a key factor for the stratification of lupus patients and therefore both lesion detection and segmentation play an important role. In our approach, the T1w image is first used to classify the three main tissues of the brain, white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF), while the FLAIR image is then used to detect focal WML as outliers of its GM intensity distribution. A set of post-processing steps based on lesion size, tissue neighborhood, and location are used to refine the lesion candidates. The proposal is evaluated on 20 patients, presenting qualitative, and quantitative results in terms of precision and sensitivity of lesion detection [True Positive Rate (62%) and Positive Prediction Value (80%), respectively] as well as segmentation accuracy [Dice Similarity Coefficient (72%)]. Obtained results illustrate the validity of the approach to automatically detect and segment lupus lesions. Besides, our approach is publicly available as a SPM8/12 toolbox extension with a simple parameter configuration.

Advanced and Conventional Magnetic Resonance Imaging in Neuropsychiatric Lupus

Neuropsychiatric lupus is a major diagnostic challenge, and a main cause of morbidity and mortality in patients with systemic lupus erythematosus (SLE). Magnetic resonance imaging (MRI) is, by far, the main tool for assessing the brain in this disease. Conventional and advanced MRI techniques are used to help establishing the diagnosis, to rule out alternative diagnoses, and recently, to monitor the evolution of the disease. This review explores the neuroimaging findings in SLE, including the recent advances in new MRI methods.

Libman-Sacks endocarditis and embolic cerebrovascular disease

OBJECTIVES

The aim of this study was to determine whether Libman-Sacks endocarditis is a pathogenic factor for cerebrovascular disease (CVD) in systemic lupus erythematosus (SLE).

BACKGROUND

A cardioembolic pathogenesis of SLE CVD manifested as: 1) neuropsychiatric systemic lupus erythematosus (NPSLE), including stroke and transient ischemic attacks (TIA); 2) neurocognitive dysfunction; and 3) magnetic resonance imaging of focal brain lesions has not been established.

METHODS

A 6-year study of 30 patients with acute NPSLE (27 women, 38 ± 12 years of age), 46 age- and sex-matched SLE controls without NPSLE (42 women, 36 ± 12 years of age), and 26 age- and sex-matched healthy controls (22 women, 34 ± 11 years of age) who underwent clinical and laboratory evaluations, transesophageal echocardiography, carotid duplex ultrasound, transcranial Doppler ultrasound, neurocognitive testing, and brain magnetic resonance imaging/magnetic resonance angiography. Patients with NPSLE were re-evaluated after 4.5 months of therapy. All patients were followed clinically for a median of 52 months.

RESULTS

Libman-Sacks vegetations (87%), cerebromicroembolism (27% with 2.5 times more events per hour), neurocognitive dysfunction (60%), and cerebral infarcts (47%) were more common in NPSLE than in SLE (28%, 20%, 33%, and 0%) and healthy controls (8%, 0%, 4%, and 0%, respectively) (all p ≤ 0.009). Patients with vegetations had 3 times more cerebromicroemboli per hour, lower cerebral blood flow, more strokes/TIA and overall NPSLE events, neurocognitive dysfunction, cerebral infarcts, and brain lesion load than those without (all p ≤ 0.01). Libman-Sacks vegetations were independent risk factors of NPSLE (odds ratio [OR]: 13.4; p < 0.001), neurocognitive dysfunction (OR: 8.0; p = 0.01), brain lesions (OR: 5.6; p = 0.004), and all 3 outcomes combined (OR: 7.5; p < 0.001). Follow-up re-evaluations in 18 of 23 (78%) surviving patients with NPSLE demonstrated improvement of vegetations, microembolism, brain perfusion, neurocognitive dysfunction, and lesion load (all p ≤ 0.04). Finally, patients with vegetations had reduced event-free survival time to stroke/TIA, cognitive disability, or death (p = 0.007).

CONCLUSIONS

The presence of Libman-Sacks endocarditis in patients with SLE was associated with a higher risk for embolic CVD. This suggests that Libman-Sacks endocarditis may be a source of cerebral emboli.

Morphologic, distributional, volumetric, and intensity characterization of periventricular hyperintensities

BACKGROUND AND PURPOSE

White matter hyperintensities are characteristic of old age and identifiable on FLAIR and T2-weighted MR imaging. They are typically separated into periventricular or deep categories. It is unclear whether the innermost segment of periventricular white matter hyperintensities is truly abnormal or is imaging artifacts.

MATERIALS AND METHODS

We used FLAIR MR imaging from 665 community-dwelling subjects 72-73 years of age without dementia. Periventricular white matter hyperintensities were visually allocated into 4 categories: 1) thin white line; 2) thick rim; 3) penetrating toward or confluent with deep white matter hyperintensities; and 4) diffuse ill-defined, labeled as "subtle extended periventricular white matter hyperintensities." We measured the maximum intensity and width of the periventricular white matter hyperintensities, mapped all white matter hyperintensities in 3D, and investigated associations between each category and hypertension, stroke, diabetes, hypercholesterolemia, cardiovascular disease, and total white matter hyperintensity volume.

RESULTS

The intensity patterns and morphologic features were different for each periventricular white matter hyperintensity category. Both the widths (r = 0.61, P < .001) and intensities (r = 0.51, P < .001) correlated with total white matter hyperintensity volume and with each other (r = 0.55, P < .001) for all categories with the exception of subtle extended periventricular white matter hyperintensities, largely characterized by evidence of erratic, ill-defined, and fragmented pale white matter hyperintensities (width: r = 0.02, P = .11; intensity: r = 0.02, P = .84). The prevalence of hypertension, hypercholesterolemia, and neuroradiologic evidence of stroke increased from periventricular white matter hyperintensity categories 1 to 3. The mean periventricular white matter hyperintensity width was significantly larger in subjects with hypertension (mean difference = 0.5 mm, P = .029) or evidence of stroke (mean difference = 1 mm, P < .001). 3D mapping revealed that periventricular white matter hyperintensities were discontinuous with deep white matter hyperintensities in all categories, except only in particular regions in brains with category 3.

CONCLUSIONS

Periventricular white matter hyperintensity intensity levels, distribution, and association with risk factors and disease suggest that in old age, these are true tissue abnormalities and therefore should not be dismissed as artifacts. Dichotomizing periventricular and deep white matter hyperintensities by continuity from the ventricle edge toward the deep white matter is possible.

The National Alliance for Medical Image Computing, a roadmap initiative to build a free and open source software infrastructure for translational research in medical image analysis

The National Alliance for Medical Image Computing (NA-MIC), is a multi-institutional, interdisciplinary community of researchers, who share the recognition that modern health care demands improved technologies to ease suffering and prolong productive life. Organized under the National Centers for Biomedical Computing 7 years ago, the mission of NA-MIC is to implement a robust and flexible open-source infrastructure for developing and applying advanced imaging technologies across a range of important biomedical research disciplines. A measure of its success, NA-MIC is now applying this technology to diseases that have immense impact on the duration and quality of life: cancer, heart disease, trauma, and degenerative genetic diseases. The targets of this technology range from group comparisons to subject-specific analysis.