Improved real-time segmentation of Intravascular Ultrasound images using coordinate-aware fully convolutional networks.

Segmentation of Intravascular Ultrasound (IVUS) images into Lumen and Media (interior and exterior) artery vessel walls is highly clinically relevant in the diagnosis and treatment of cardiovascular diseases such as atherosclerosis. When fused with position data, such segmentations also play a key role in reconstructing 3D representations of arteries. Automated segmentation in real-time is known to be a difficult image analysis problem, primarily due to artefacts commonly present in IVUS ultrasound images such as shadows, guide-wire effects, and side-branches. An additional challenge is the limited amount of expert labelled IVUS data, which limits the application of many well-performing deep learning models from other domains. To exploit the circular layered structure of the artery in B-Mode images, we propose a multi-class fully convolutional semantic segmentation network based on a minimal U-Net architecture augmented with learned translation dependence in the polar domain. The coordinate awareness in the multi-class segmentation allows the model to exploit relative spatial context about the interior and exterior vessel walls which are simply separable in polar coordinates. After training on 109 expert-labelled examples, our model significantly outperforms the state-of-the art in terms of mean Jaccard Measure (0.91 vs. 0.89) and Hausdorff distance (0.32 mm vs. 0.48 mm) on Media segmentation, and reaches equivalent performance on Lumen segmentation when evaluated on a standard publicly available dataset of 326 IVUS B-Mode images captured by 20 Mhz ultrasound probes. Using an order of magnitude fewer trainable parameters than the previous state-of-the-art, our model runs over 50 times faster and is able to execute in only 3 ms on a common GPU, achieving both leading accuracy and practical real-time performance.

Estimation of central aortic blood pressure: a systematic meta-analysis of available techniques.

BACKGROUND: Central aortic blood pressure (cBP) is often promoted to be a superior predictor of cardiovascular risk compared to brachial blood pressure, and brachial-central pulse pressure amplification is also suggested as prognostic. Several devices and techniques, each purporting to estimate cBP, have entered commercial use. The interchangeability of cBP measurements between devices and the influence of disease states on central to brachial pulse pressure amplification remain unclear. The useful measurement of cBP in clinical trials is dependent on clarification of these issues. METHOD: We performed a systematic meta-analysis of studies reporting cBP between 2000 and 2012. Studies were included if both central and brachial SBPs (cSBP and bSBP) were reported. Studies were categorized by technique and according to the prevalent disease state with the bSBP - cSBP difference calculated. Random-effects modeling (inverse variance weighted approach) was used to estimate the pooled mean difference associated with each technique. RESULTS: Of the 164 eligible studies, the SphygmoCor device was most commonly reported (110 studies), with direct carotid applanation second-most utilized (31 studies). In 30 included invasive cohorts, the measured cSBP did not differ significantly from the oscillometric bSBP recorded [mean difference 4.19  mmHg, 95% confidence interval (CI) -4.13 to 12.51], whereas mean differences of 12.77  mmHg (95% CI 11.93, 13.60) and 8.83  mmHg (95% CI 7.86, 9.79) were obtained with the SphygmoCor and carotid applanation estimates of cSBP, respectively (both P < 0.05). Conversely, the reported mean cSBP-to-bSBP differences measured across various disease states with SphygmoCor did not differ significantly. CONCLUSION: This meta-analysis suggests that noninvasive cBP estimation is device/technique-dependent. Consequently, caution is advisable in applying these devices and techniques across clinical studies.