Marc-André Weber

The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)

In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences. Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients - manually annotated by up to four raters - and to 65 comparable scans generated using tumor image simulation software. Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%-85%), illustrating the difficulty of this task. We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously. Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements. The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource.

Clinical long-term outcome, technical success, and cost analysis of radiofrequency ablation for the treatment of osteoblastomas and spinal osteoid osteomas in comparison to open surgical resection

ObjectiveTo assess the clinical success and costs of computed tomography (CT)-guided radiofrequency ablation (RFA) of osteoblastoma (OB) and spinal osteoid osteoma (OO).Materials and methodsNineteen patients with OB and eight patients with spinal OO were treated with CT-guided RFA. The OBs were localized in the extremities (n = 10), the vertebral column (n = 2), and (juxta-)articular (n = 7). Dedicated procedural techniques included three-dimensional CT-guided access planning in all cases, overlapping RFA needle positions (median, two positions; range, 1–6 RF-electrode positions) within the OB nidus (multiple ablation technique, n = 15), and thermal protection in case of adjacent neural structure in four spinal OO. The data of eight operated OB and ten operated spinal OO patients were used for comparison. Long-term success was assessed by clinical examination and using a questionnaire sent to all operated and RFA-treated patients including visual analogue scales (VAS) regarding the effect of RFA on severity of pain and limitations of daily activities (0–10, with 0 = no pain/limitation up to 10 = maximum or most imaginable pain/limitation).ResultsAll patients had a clear and persistent pain reduction until the end of follow-up. The mean VAS score for all spinal OO patients and all OB patients treated either with RFA or with surgical excision significantly decreased for severity of pain at night, severity of pain during the day, and both for limitations of daily and of sports activities.ConclusionsRFA is an efficient method for treating OB and spinal OO and should be regarded as the first-line therapy after interdisciplinary individual case discussion.

Sensitivity of whole-body CT and MRI versus projection radiography in the detection of osteolyses in patients with monoclonal plasma cell disease

PURPOSE To compare sensitivity of whole-body Computed Tomography (wb-CT) and whole-body Magnetic Resonance Imaging (wb-MRI) with Projection Radiography (PR) regarding each methods ability to detect osteolyses in patients with monoclonal plasma cell disease. PATIENTS AND METHODS The bone status of 171 patients was evaluated. All patients presented with multiple myeloma (MM) of all stages, monoclonal gammopathy of unknown significance (MGUS) or solitary plasmacytoma. Two groups were formed. Group A consisted of 52 patients (26 females, 26 males) with an average age of 62 years (range, 45-89 years) who received, both, PR and wb-CT as part of their diagnostic work-up. Group B comprised 119 patients (58 females, 61 males) averaging 57 years of age (range, 20-80 years) who received, both, PR and wb-MRI. Two experienced radiologists were blinded regarding the disease status and assessed the number and location of osteolyses in consensus. A distinction was made between axial and extra-axial lesions. RESULTS In group A, wb-CT revealed osteolyses in 12 patients (23%) that were not detected in PR. CT was superior in detecting lesions in patients with osteopenia and osteoporosis. Compared with PR, wb-CT was significantly more sensitive in detecting osteolyses than PR (p<0.001). This was particularly true for axial lesions. Additionally, CT revealed clinically relevant incidental findings in 33 patients (63%). In group B, wb-MRI revealed lesions in 19 patients (16%) that were not detected in PR. All lesions detected by PR were also detected by wb-MRI and wb-CT. Wb-MRI and wb-CT are each superior to PR in detecting axial lesions. CONCLUSION Wb-CT can detect 23% more focal lesions than PR, especially in the axial skeleton. Therefore, this imaging method should be preferred over PR in the diagnostic work-up and staging of patients with monoclonal plasma cell disease.

Cloud-Based Evaluation of Anatomical Structure Segmentation and Landmark Detection Algorithms: VISCERAL Anatomy Benchmarks

Variations in the shape and appearance of anatomical structures in medical images are often relevant radiological signs of disease. Automatic tools can help automate parts of this manual process. A cloud-based evaluation framework is presented in this paper including results of benchmarking current state-of-the-art medical imaging algorithms for anatomical structure segmentation and landmark detection: the VISCERAL Anatomy benchmarks. The algorithms are implemented in virtual machines in the cloud where participants can only access the training data and can be run privately by the benchmark administrators to objectively compare their performance in an unseen common test set. Overall, 120 computed tomography and magnetic resonance patient volumes were manually annotated to create a standard Gold Corpus containing a total of 1295 structures and 1760 landmarks. Ten participants contributed with automatic algorithms for the organ segmentation task, and three for the landmark localization task. Different algorithms obtained the best scores in the four available imaging modalities and for subsets of anatomical structures. The annotation framework, resulting data set, evaluation setup, results and performance analysis from the three VISCERAL Anatomy benchmarks are presented in this article. Both the VISCERAL data set and Silver Corpus generated with the fusion of the participant algorithms on a larger set of non-manually-annotated medical images are available to the research community.

Reward System Dysfunction as a Neural Substrate of Symptom Expression Across the General Population and Patients With Schizophrenia

Dysfunctional patterns of activation in brain reward networks have been suggested as a core element in the pathophysiology of schizophrenia. However, it remains unclear whether this dysfunction is specific to schizophrenia or can be continuously observed across persons with different levels of nonclinical and clinical symptom expression. Therefore, we sought to investigate whether the pattern of reward system dysfunction is consistent with a dimensional or categorical model of psychosis-like symptom expression. 23 patients with schizophrenia and 37 healthy control participants with varying levels of psychosis-like symptoms, separated into 3 groups of low, medium, and high symptom expression underwent event-related functional magnetic resonance imaging while performing a Cued Reinforcement Reaction Time task. We observed lower activation in the ventral striatum during the expectation of high vs no reward to be associated with higher symptom expression across all participants. No significant difference between patients with schizophrenia and healthy participants with high symptom expression was found. However, connectivity between the ventral striatum and the medial orbitofrontal cortex was specifically reduced in patients with schizophrenia. Dysfunctional local activation of the ventral striatum depends less on diagnostic category than on the degree of symptom expression, therefore showing a pattern consistent with a psychosis continuum. In contrast, aberrant connectivity in the reward system is specific to patients with schizophrenia, thereby supporting a categorical view. Thus, the results of the present study provide evidence for both continuous and discontinuous neural substrates of symptom expression across patients with schizophrenia and the general population.

Soft tissue tumors in adults: ESSR-approved guidelines for diagnostic imaging

Soft tissue sarcomas are rare, but early, accurate diagnosis with subsequent appropriate treatment is crucial for the clinical outcome. The ESSR guidelines are intended to help radiologists in their decision-making and support discussion among clinicians who deal with patients with suspected or proven soft tissue tumors. Potentially malignant lesions recognized by ultrasound should be referred for magnetic resonance imaging (MRI), which also serves as a preoperative local staging modality, with specific technical requirements and mandatory radiological report elements. Radiography may add information about matrix calcification and osseous involvement. Indeterminate lesions, or lesions in which therapy is dependent on histology results, should be biopsied. For biopsy, we strongly recommend referral to a specialist sarcoma center, where an interdisciplinary tumor group, with a specialized pathologist, radiologist, and the surgeon are involved. In sarcoma, a CT scan of the chest is mandatory. Additional staging modalities are entity-specific. There are no evidence-based recommendations for routine follow-up in surgically treated sarcomas. However, we would recommend regular follow-up with intervals dependent on tumor grade, for 10 years after the initial diagnosis.

Dynamic Contrast-Enhanced Sonography and Dynamic Contrast-Enhanced Magnetic Resonance Imaging for Preoperative Diagnosis of Infected Nonunions

Bone regeneration depends on perfusion of the fracture tissue, whereby hypervascularity is associated with infection, which itself causes nonunions. To date, nonunion perfusion has not been assessed with contrast‐enhanced sonography. The aim of this study was to evaluate the potential of contrast‐enhanced sonography in the analysis of nonunion tissue perfusion.

Ultrasound vs. MRI in the assessment of rotator cuff structure prior to shoulder arthroplasty

BACKGROUND/AIMS We compared the accuracy of US to 3 T Tesla MRI for the detection of rotator cuff and long biceps tendon pathologies before joint replacement. METHODS 45 patients were prospectively included. RESULTS For the supraspinatus tendon, the accuracy of US when using MRI as reference was 91.1%. For the infraspinatus tendon, the accuracy with MRI as reference was 84.4%. The subscapularis tendon was consistently assessed by US and MRI in 35/45 patients (accuracy 77.8%). For the long biceps tendon the accuracy was 86.7%. CONCLUSION US detection of rotator cuff and biceps tendon integrity is comparable to MRI and should be preferred in revision cases.

Delayed gadolinium‐enhanced MRI of cartilage (dGEMRIC) and T2 mapping at 3T MRI of the wrist: Feasibility and clinical application

To assess the feasibility of delayed gadolinium‐enhanced magnetic resonance imaging (MRI) of cartilage (dGEMRIC) and T2 mapping for biochemical imaging of the wrist at 3T.

Groin Pain in Athletes.

Groin pain in athletes is one of the most difficult to treat clinical entities in sports medicine. The reasons are the amount of differential diagnoses, complexity of pathophysiologic causes and the long time of limited participation in sport. In order to maximize efficient treatment, thorough diagnostics and a clear therapeutic regimen are crucial. To succeed with this issue, a close cooperation between physicians and radiologists is mandatory. MRI is gold standard in the diagnostic work-up of the principal differential diagnoses, such as muscle tears, avulsion injuries, stress fractures, tears of acetabular labrum, and osteitis pubis. The article gives a comprehensive overview of the special anatomy and biomechanics of the pubic region and of typical MRI findings in athletes with groin pain. The use of dedicated imaging protocols is also discussed.