COMPUTER-AIDED DIAGNOSIS AND VISUALIZATION BASED ON CLUSTERING AND INDEPENDENT COMPONENT ANALYSIS FOR BREAST MRI.

Computer-aided diagnosis and simultaneous visualization based on independent component analysis and clustering are integrated in an intelligent system for the evaluation of small mammographic lesions in breast MRI. These techniques are tested on biomedical time-series representing breast MRI scans and enable the extraction of spatial and temporal features of dynamic MRI data stemming from patients with confirmed lesion diagnosis. By revealing regional properties of contrast-agent uptake characterized by subtle differences of signal amplitude and dynamics, these methods provide both a set of prototypical time-series and a corresponding set of cluster assignment maps which further provide a segmentation with regard to identification and regional subclassification of pathological breast tissue lesions. Both approaches lead to an increase of the diagnostic accuracy of MRI mammography by improving the sensitivity without reduction of specificity.

[Cardiac arrest following blunt chest injury. Emergency thoracotomy without ifs or buts?]. / Kreislaufstillstand nach stumpfem Thoraxtrauma. Notfallthorakotomie ohne Wenn und Aber?

In German-speaking countries, most serious thoracic injuries are attributable to the impact of blunt force; they are the second most frequent result of injury after head injury in polytrauma patients with multiple injuries. Almost one in every three polytraumatized patients with significant chest injury develops acute lung failure, and one in every four, acute circulatory failure. The acute circulatory arrest following serious chest injury involves a high mortality rate, and in most cases it reflects a tension pneumothorax, cardiac tamponade, or hemorrhagic shock resulting from injury to the heart or one of the large vessels close to it. Brisk drainage of tension pneumothorax and adequate volume restoration are therefore particularly important in resuscitation of multiply traumatized patients, as are rapid resuscitative thoracotomy to allow direct heart massage, drainage of pericardial tamponade, and control of hemorrhage. However the probability of survival described in the literature is very low for patients sustaining severe chest trauma with acute cardiac arrest. The case report presented here describes a female polytrauma patient who suffered an acute cardiac arrest following cardiac tamponade after admission in the emergency department and who survived without neurological deficits after an emergency thoracotomy. Selections from the topical literature can help the treating physician in the emergency department in making decisions on whether an emergency thoracotomy is indicated after a blunt chest injury and on the procedure itself.

Cluster analysis of dynamic cerebral contrast-enhanced perfusion MRI time-series.

We performed neural network clustering on dynamic contrast-enhanced perfusion magnetic resonance imaging time-series in patients with and without stroke. Minimal-free-energy vector quantization, self-organizing maps, and fuzzy c-means clustering enabled self-organized data-driven segmentation with respect to fine-grained differences of signal amplitude and dynamics, thus identifying asymmetries and local abnormalities of brain perfusion. We conclude that clustering is a useful extension to conventional perfusion parameter maps.

Detection of suspicious lesions in dynamic contrast enhanced MRI data.

Dynamic contrast-enhanced magnet resonance imaging (DCE-MRI) has become an important source of information to aid breast cancer diagnosis. Nevertheless, next to the temporal sequence of 3D volume data from the DCE-MRI technique, the radiologist commonly adducts information from other modalities for his final diagnosis. Thus, the diagnosis process is time consuming and tools are required to support the human expert. We investigate an automatic approach that detects the location and delineates the extent of suspicious masses in multi-temporal DCE-MRI data sets. It applies the state-of-the-art support vector machine algorithm to the classification of the short-time series associated with each voxel. The ROC analysis shows an increased specificity in contrast to standard evaluations techniques.

[Volumetric MRI for evaluation of regional pattern and progressin of neocortical degeneration in Alzheimer's disease]. / MR-Volumetrie zur Darstellung von Verteilung und zeitlicher Abfolge neokortikaler Degeneration bei Morbus Alzheimer.

PURPOSE: Volumetric analysis of the corpus callosum and hippocampus using MRI in Alzheimer's disease (AD) to evaluate the regional pattern and progression of neocortical neurodegeneration. METHODS: In subsequent studies we investigated patients with AD and healthy controls. Volumetry was based on MRI-data from a sagittal 3D T1w-gradient echo sequence. The corpus callosum (CC) was measured in a midsagittal slice, and subdivided into 5 subregions. Volumetry of the hippocampus/amygdala-formation (HAF) was performed by segmentation in coronary reoriented slices. RESULTS: In AD patients we found a significant atrophy in the rostrum und splenium of CC. The atrophy was correlated with the severity of dementia, but no correlation was found with the load of white matter lesions. In comparison with (18)FDG-PET, we found a significant correlation of regional CC-atrophy with the regional decline of cortical glucose metabolism. A ROC-analysis demonstrated no significant differences in the diagnostic accuracy of HAF volumetry and regional CC volumetry of the splenium (region C5) even in mild stages of dementia. CONCLUSION: Regional atrophy of CC can be used as a marker of neocortical degeneration even in early stages of dementia in AD.

Quantitative comparison of automatic and interactive methods for MRI-SPECT image registration of the brain based on 3-dimensional calculation of error.

UNLABELLED: A wide range of techniques for registration of medical images has been devised in recent years. The aim of this study is to quantify the overall spatial registration error of 3 different methods for image registration: interactive matching, surface matching, and uniformity index matching as described by Woods. METHODS: MRI and ethylcysteinate dimer-SPECT images of the brain were registered for 15 patients. The matching error was assessed by determining intra- and interobserver variability of registrations. Quantification of the registration error was based on the mean spatial distance of 5000 voxels between 2 image positions. The mean position after repeated registrations in each patient was used as the gold standard. To evaluate the coherence of the 3 different registration methods, intermethod variability was determined. RESULTS: Interactive matching showed an intraobserver/interobserver variability of 1.5+/-0.3 mm/1.6+/-0.3 mm (mean +/- SD). The time demand for this method was 11+/-5 min. Surface matching revealed a variability of 2.6+/-1.1 mm/3.8+/-1.0 mm and a time demand of 26+/-12 min. Reproducibility of Woods' algorithm was 2.2+/-0.8 mm with a time demand of 9+/-3 min. In 4 of the 15 cases, Woods' method failed. The mean deviation between all 3 methods was 2.3+/-0.8 mm. CONCLUSION: With a suitable user interface, interactive matching had the lowest registration error. The influence of subjectivity was shown to be negligible. Therefore, interactive matching is our preferred technique for image fusion of the brain.