Grad-CAM Guided U-Net for MRI-based Pseudo-CT Synthesis.

In this paper, we address the task of image-to-image translation from MRI to CT domain. We propose a 2D U-Net-based deep learning approach for pseudo-CT synthesis that incorporates an additional Grad-CAM guided attention mechanism for superior image translation of bone regions. The suggested architecture consists of image-to-image translation and image classification modules. We first train our classifier to distinguish between MR and CT images. After that, we utilize it in combination with the Grad-CAM technique to provide additional guidance to our image-to-image translation network. We generate CT-class-specific localization maps for both CT and pseudo-CT images and then compare them. Thus, we force the image-to-image translation network to focus on relevant attributes of the CT class, such as bone structures, while learning to synthesize pseudo-CTs. The performance of the proposed approach is evaluated on the publicly available RIRE data set. Since MR and CT images in this data set are not correctly aligned with each other, we also briefly describe the applied image registration procedure. The experimental results are compared to the baseline U-Net model and demonstrate both qualitative and quantitative improvements, whereas significant performance gain is achieved for bone regions. Clinical Relevance- MRI-based pseudo-CT synthesis is essential for attenuation correction of PET in combined PET/MRI systems and plays a vital role in MRI-only radiotherapy planning. Accurate pseudo-CTs can prevent patients from harmful and unnecessary radiation exposure.

CHAOS Challenge - combined (CT-MR) healthy abdominal organ segmentation.

Segmentation of abdominal organs has been a comprehensive, yet unresolved, research field for many years. In the last decade, intensive developments in deep learning (DL) introduced new state-of-the-art segmentation systems. Despite outperforming the overall accuracy of existing systems, the effects of DL model properties and parameters on the performance are hard to interpret. This makes comparative analysis a necessary tool towards interpretable studies and systems. Moreover, the performance of DL for emerging learning approaches such as cross-modality and multi-modal semantic segmentation tasks has been rarely discussed. In order to expand the knowledge on these topics, the CHAOS - Combined (CT-MR) Healthy Abdominal Organ Segmentation challenge was organized in conjunction with the IEEE International Symposium on Biomedical Imaging (ISBI), 2019, in Venice, Italy. Abdominal organ segmentation from routine acquisitions plays an important role in several clinical applications, such as pre-surgical planning or morphological and volumetric follow-ups for various diseases. These applications require a certain level of performance on a diverse set of metrics such as maximum symmetric surface distance (MSSD) to determine surgical error-margin or overlap errors for tracking size and shape differences. Previous abdomen related challenges are mainly focused on tumor/lesion detection and/or classification with a single modality. Conversely, CHAOS provides both abdominal CT and MR data from healthy subjects for single and multiple abdominal organ segmentation. Five different but complementary tasks were designed to analyze the capabilities of participating approaches from multiple perspectives. The results were investigated thoroughly, compared with manual annotations and interactive methods. The analysis shows that the performance of DL models for single modality (CT / MR) can show reliable volumetric analysis performance (DICE: 0.98 ± 0.00 / 0.95 ± 0.01), but the best MSSD performance remains limited (21.89 ± 13.94 / 20.85 ± 10.63 mm). The performances of participating models decrease dramatically for cross-modality tasks both for the liver (DICE: 0.88 ± 0.15 MSSD: 36.33 ± 21.97 mm). Despite contrary examples on different applications, multi-tasking DL models designed to segment all organs are observed to perform worse compared to organ-specific ones (performance drop around 5%). Nevertheless, some of the successful models show better performance with their multi-organ versions. We conclude that the exploration of those pros and cons in both single vs multi-organ and cross-modality segmentations is poised to have an impact on further research for developing effective algorithms that would support real-world clinical applications. Finally, having more than 1500 participants and receiving more than 550 submissions, another important contribution of this study is the analysis on shortcomings of challenge organizations such as the effects of multiple submissions and peeking phenomenon.

Increasing the feasibility of minimally invasive procedures in type A aortic dissections: a framework for segmentation and quantification.

PURPOSE: Our goal is to provide precise measurements of the aortic dimensions in case of dissection pathologies. Quantification of surface lengths and aortic radii/diameters together with the visualization of the dissection membrane represents crucial prerequisites for enabling minimally invasive treatment of type A dissections, which always also imply the ascending aorta. METHODS: We seek a measure invariant to luminance and contrast for aortic outer wall segmentation. Therefore, we propose a 2D graph-based approach using phase congruency combined with additional features. Phase congruency is extended to 3D by designing a novel conic directional filter and adding a lowpass component to the 3D Log-Gabor filterbank for extracting the fine dissection membrane, which separates the true lumen from the false one within the aorta. RESULTS: The result of the outer wall segmentation is compared with manually annotated axial slices belonging to 11 CTA datasets. Quantitative assessment of our novel 2D/3D membrane extraction algorithms has been obtained for 10 datasets and reveals subvoxel accuracy in all cases. Aortic inner and outer surface lengths, determined within 2 cadaveric CT datasets, are validated against manual measurements performed by a vascular surgeon on excised aortas of the body donors. CONCLUSIONS: This contribution proposes a complete pipeline for segmentation and quantification of aortic dissections. Validation against ground truth of the 3D contour lengths quantification represents a significant step toward custom-designed stent-grafts.

Fully automatic extraction of human spine curve from MR images using methods of efficient intervertebral disk extraction and vertebra registration.

PURPOSE: A fully automatic method is proposed for extracting human spine curve which is required for gait modeling. By means of the gait modeling, origin of the gait pathology of patients could be found. METHODS: Our method is composed of two parts. The first part is the extraction of intervertebral disk positions where an efficient method is proposed. At the beginning of this part, all possible positions of intervertebral disks are located using a gradient-based method. Then, non-intervertebral disks are filtered out by a graph-based and an active shape model-based methods. In the second part, extracted disk positions are used by a vertebra registration method to segment spine vertebrae. Finally, spine curve is obtained by interpolating centers of segmented vertebrae using cubic spline. RESULTS: We tested our method with 13 MR data sets of patients. All disk positions of each MR data set were correctly extracted in the first part. The mean deviation of centers of segmented vertebrae that were obtained in the second part and used to interpolate spine curve was around 1.4 mm. CONCLUSIONS: Our method achieves a fully automatic extraction of the spine curve. The extraction of intervertebral disk positions in the first part of our method when compared to model-based methods and manual selection which were proposed in other papers is highly efficient. In the second part including the vertebra registration, a new similarity measurement method, which is used to guide the vertebra atlas fitting process, is proposed to solve the problem of changes in overlap. Through our experiment, results of spine curves are at a highly accurate level.