C. Leavens

Golay Pulse Encoding for Microbubble Contrast Imaging in Ultrasound

We present a technique that uses Golay phase encoding, pulse inversion, and amplitude modulation (GPIAM) for microbubble contrast agent imaging with ultrasound. This technique improves the contrast-to-tissue ratio (CTR) by increasing the time-bandwidth product of the insonating waveforms. A nonlinear pulse compression algorithm is used to compress the signal energy upon receive. A 6.5-dB improvement in CTR was observed using an 8-chip GPIAM sequence compared to a conventional pulse-inversion amplitude-modulation sequence. The CTR improvement comes at the cost of a reduction in frame rate: GPIAM coding uses four input pulses whereas most contrast imaging sequences require two or three pulses. Our results showed that the microbubble response can be phase encoded and subsequently compressed using a nonlinear matched-filtering algorithm, in order to enhance the signal from the contrast agent, while maintaining resolution and suppressing the tissue signal.

Validation of automatic landmark identification for atlas-based segmentation for radiation treatment planning of the head-and-neck region

Manual contouring of target volumes and organs at risk in radiation therapy is extremely time-consuming, in particular for treating the head-and-neck area, where a single patient treatment plan can take several hours to contour. As radiation treatment delivery moves towards adaptive treatment, the need for more efficient segmentation techniques will increase. We are developing a method for automatic model-based segmentation of the head and neck. This process can be broken down into three main steps: i) automatic landmark identification in the image dataset of interest, ii) automatic landmark-based initialization of deformable surface models to the patient image dataset, and iii) adaptation of the deformable models to the patient-specific anatomical boundaries of interest. In this paper, we focus on the validation of the first step of this method, quantifying the results of our automatic landmark identification method. We use an image atlas formed by applying thin-plate spline (TPS) interpolation to ten atlas datasets, using 27 manually identified landmarks in each atlas/training dataset. The principal variation modes returned by principal component analysis (PCA) of the landmark positions were used by an automatic registration algorithm, which sought the corresponding landmarks in the clinical dataset of interest using a controlled random search algorithm. Applying a run time of 60 seconds to the random search, a root mean square (rms) distance to the ground-truth landmark position of 9.5 ± 0.6 mm was calculated for the identified landmarks. Automatic segmentation of the brain, mandible and brain stem, using the detected landmarks, is demonstrated.

Fast B-Flow Imaging: A Method for Improving Frame Rate in Golay Coded B-Flow Imaging

A technique for Golay coded B-flow imaging, called fast B-flow imaging, has been developed. This technique improves the frame rate of Golay coded B-flow imaging. In this technique, three instead of four input pulses are used to produce each scan line. A standard Golay pulse-pair is used as two of the three inputs, and pulse compression is performed upon receive returning the echoes from stationary (tissue) objects in the image. The third input is a repetition of one of the first two inputs. Upon receive, this pulse is cross correlated with an inverted copy of its input pulse. Addition of the cross-correlated signals produced from the identical input pulses results in the cancellation of the strong tissue echoes, and enables visualization of the weaker/moving blood echoes. Combining a small fraction of the tissue echoes with the weaker blood echoes allows both to be visualized in the same gray scale image. By using three instead of four input pulses, this technique can achieve a frame rate improvement of 33% compared with standard Golay coded B-flow imaging, with some loss in signal-to-noise ratio. The impact of axial and lateral motion on these techniques is examined. A quantitative comparison of both techniques is presented.

Serial 4DCT/4DPET imaging to predict and monitor response for locally-advanced non-small cell lung cancer chemo-radiotherapy

BACKGROUND AND PURPOSE A FDG-PET/CT image feature with optimal prognostic potential for locally-advanced non-small cell lung cancer (LA-NSCLC) patients has yet to be identified, and neither has the optimal time for FDG-PET/CT response assessment; furthermore, nodal features have been largely ignored in the literature. We propose to identify image features or imaging time point with maximal prognostic power. MATERIALS AND METHODS Consecutive consenting patients with LA-NSCLC receiving curative intent CRT were enrolled. 4DPET/4DCT scans were acquired 0, 2, 4, and 7 weeks during IMRT treatment. Eleven image features and their rates of change were recorded for each time point and tested for each of the possible outcome 2 years post CRT using the Kaplan-Meier method. RESULTS 32 consecutive patients were recruited, 27 completing all scans. Restricting analysis to 4DPET/4DCT features and rates of change with p < 0.005, several volume-based features and their rates of change reached significance. Image features involving nodal disease were the only ones associated with overall survival. CONCLUSIONS Several 4DPET/CT features and rates of change can reach significant association (p < 0.005) with outcomes, including overall survival, at many time points. The optimal time for adaptive CRT is therefore not constrained uniquely on imaging.