Efficient multilevel brain tumor segmentation with integrated bayesian model classification

We present a new method for automatic segmentation of heterogeneous image data that takes a step toward bridging the gap between bottom-up affinity-based segmentation methods and top-down generative model based approaches. The main contribution of the paper is a Bayesian formulation for incorporating soft model assignments into the calculation of affinities, which are conventionally model free. We integrate the resulting model-aware affinities into the multilevel segmentation by weighted aggregation algorithm, and apply the technique to the task of detecting and segmenting brain tumor and edema in multichannel magnetic resonance (MR) volumes. The computationally efficient method runs orders of magnitude faster than current state-of-the-art techniques giving comparable or improved results. Our quantitative results indicate the benefit of incorporating model-aware affinities into the segmentation process for the difficult case of glioblastoma multiforme brain tumor.

A customizable MR brain imaging atlas of structure and function for decision support

We present a MR brain atlas for structure and function (diffusion weighted images). The atlas is customizable for contrast and orientation to match the current patient images. In addition, the atlas also provides normative values of MR parameters (T1, T2 and ADC values). The atlas is designed on informatics principles to provide context sensitive decision support at the time of primary image interpretation. Additional support for diagnostic interpretation is provided by a list of expert created most relevant 'Image Finding Descriptors' that will serve as cues to the user. The architecture of the atlas module is integrated into the image workflow of a radiology department to provide support at the time of primary diagnosis.

Disease specific intelligent pre-fetch and hanging protocol for diagnostic neuroradiology workstations

Clinical data sets for neuroradiological cases can be quite large. A typical brain tumor patient at UCLA will undergo 8-10 separate studies over a 2 year period, each study will produce 60-100 magnetic resonance (MR) images. Gathering and sorting through a patient s imaging events during the course of treatment can be both overwhelming and time consuming. The purpose of this research is to develop an intelligent pre-fetch and hanging protocol that automatically gathers the relevant prior examinations from a picture archiving, and communication systems (PACS) archive and sends the pertinent historical images to the diagnostic display station where the new examination is subsequently read out. The intelligent hanging protocol describes the type of layout and sequence for image display. We have developed a classification scheme to organize the pertinent patient information to selectively pre-fetch and intelligently present the images to review brain tumor cases for a diagnostic neuroradiology workstation.

MR imaging of cervical spine motion with HASTE

The HASTE (half-Fourier acquisition single-shot turbo spin-echo) technique delivers images with T2-weighting in about half a second and could be ideal for fast dynamic studies when T2-weighting is needed. We evaluated cardiac-triggered HASTE to study cervical spine flexion/extension. The cervical spines of ten asymptomatic volunteers were studied during flexion/extension motion on a 1.5 Tesla imager using a cardiac triggered version of the HASTE technique. Midline sagittal images were acquired every 2 to 3 s during neck flexion and extension. Image quality was compared to traditional T2-weighted Turbo spin-echo. The study duration per flexion/ extension was typically less than 20 seconds and well tolerated. The cardiac-gated T2-weighted HASTE images compared favorably to the traditional T2-weighted TSE images in quality and overall anatomic detail. Range of motion averaged: flexion 30 degrees (range 8 degrees -48 degrees) and extension 23 degrees (range 0 degrees -57 degrees ). Greatest motion occurred in the lower cervical spine (C4-C7). At the intervertebral discs the canal diameter, anterior and posterior CSF spaces were widest in neutral position and decreased with flexion and extension. Therefore, Cardiac-gated T2 HASTE sequences provide diagnostic and time-efficient dynamic MR images of cervical spine motion.