Automatic segmentation of thrombosed aortic dissection in post-operative CT-angiography images

PURPOSE

The thrombus in the false lumen (FL) of aortic dissection (AD) patients is a meaningful indicator to determine aortic remodeling but difficult to measure in clinic. In this study, a novel segmentation strategy based on deep learning was proposed to automatically extract the thrombus in the FL in post-operative computed tomography angiography (CTA) images of AD patients, which provided an efficient and convenient segmentation method with high accuracy.

METHODS

A two-step segmentation strategy was proposed. Each step contained a convolutional neural network (CNN) to segment the aorta and the thrombus, respectively. In the first step, a CNN was used to obtain the binary segmentation mask of the whole aorta. In the second step, another CNN was introduced to segment the thrombus. The results of the first step were used as additional input to the second step to highlight the aorta in the complex background. Moreover, skip connection attention refinement (SAR) modules were designed and added in the second step to improve the segmentation accuracy of the thrombus details by efficiently using the low-level features.

RESULTS

The proposed method provided accurate thrombus segmentation results (0.903 ± 0.062 in dice score, 0.828 ± 0.092 in Jaccard index, and 2.209 ± 2.945 in 95% Hausdorff distance), which showed improvement compared to the methods without prior information (0.846 ± 0.085 in dice score) and the method without SAR (0.899 ± 0.060 in dice score). Moreover, the proposed method achieved 0.967 ± 0.029 and 0.948 ± 0.041 in dice score of true lumen (TL) and patent FL (PFL) segmentation, respectively, indicating the excellence of the proposed method in the segmentation task of the overall aorta.

CONCLUSIONS

A novel CNN-based segmentation framework was proposed to automatically obtain thrombus segmentation for thrombosed AD in post-operative CTA images, which provided a useful tool for further application of thrombus-related indicators in clinical and research application.

Space-time-regulated imaging analyzer for smart coagulation diagnosis

The development of intelligent blood coagulation diagnoses is awaited to meet the current need for large clinical time-sensitive caseloads due to its efficient and automated diagnoses. Herein, a method is reported and validated to realize it through artificial intelligence (AI)-assisted optical clotting biophysics (OCB) properties identification. The image differential calculation is used for precise acquisition of OCB properties with elimination of initial differences, and the strategy of space-time regulation allows on-demand space time OCB properties identification and enables diverse blood function diagnoses. The integrated applications of smartphones and cloud computing offer a user-friendly automated analysis for accurate and convenient diagnoses. The prospective assays of clinical cases (n = 41) show that the system realizes 97.6%, 95.1%, and 100% accuracy for coagulation factors, fibrinogen function, and comprehensive blood coagulation diagnoses, respectively. This method should enable more low-cost and convenient diagnoses and provide a path for potential diagnostic-markers finding.

•

An ultraportable optofluidic analyzer empowers convenient coagulation diagnoses

•

The system enables optical clotting biophysics (OCB) properties acquisition and process

•

Coagulation function diagnoses uses intelligent OCB properties identification

•

Space-time regulation of OCB properties endow it capability to diverse diagnoses

OTO-Net: An Automated MRA Image Segmentation Network for Intracranial Aneurysms

Intracranial aneurysms are local dilations of the cerebral blood vessels; people with intracranial aneurysms have a high risk to cause bleeding in the brain, which is related to high mortality and morbidity rates. Accurate detection and segmentation of intracranial aneurysms from Magnetic Resonance Angiography (MRA) images are essential in the clinical routine. Manual annotations used to assess the intracranial aneurysms on MRA images are substantial interobserver variability for both aneurysm detection and assessment of aneurysm size and growth. Many prior automated segmentation works have focused their efforts on tackling the problem, but there is still room for performance improvement due to the significant variability of lesions in the location, size, structure, and morphological appearance. To address these challenges, we propose a novel One-Two-One Fully Convolutional Networks (OTO-Net) for intracranial aneurysms automated segmentation in MRA images. The OTO-Net uses full convolution to achieve intracranial aneurysms automated segmentation through the combination of downsampling, upsampling, and skip connection. In addition, loss ensemble is used as the objective function to steadily improve the backpropagation efficiency of the network structure during the training process. We evaluated the proposed OTO-Net on one public benchmark dataset and one private dataset. Our proposed model can achieve the automated segmentation accuracy with 98.37% and 97.86%, average surface distances with 1.081 and 0.753, dice similarity coefficients with 0.9721 and 0.9813, and Hausdorff distance with 0.578 and 0.642 on these two datasets, respectively.

Deep learning based fully automatic segmentation of the left ventricular endocardium and epicardium from cardiac cine MRI

Background

The segmentation of cardiac medical images is a crucial step for calculating clinical indices such as wall thickness, ventricular volume, and ejection fraction.

Methods

In this study, we introduce a method named LsUnet that combines multi-channel, fully convolutional neural network, and annular shape level-set methods for efficiently segmenting cardiac cine magnetic resonance (MR) images. In this method, the multi-channel deep learning algorithm is applied to train the segmentation task to extract the left ventricle (LV) endocardial and epicardial contours. Next, the segmentation contours from the multi-channel deep learning method are incorporated into a level-set formulation, which is dedicated explicitly to detecting annular shapes to assure the segmentation's accuracy and robustness.

Results

The proposed automatic approach was evaluated on 95 volumes (total 1,076 slices, ~80% as for training datasets, ~20% 2D as for testing datasets). This combined multi-channel deep learning and annular shape level-set segmentation method achieved high accuracy with average Dice values reaching 92.15% and 95.42% for LV endocardium and epicardium delineation, respectively, in comparison to the reference standard (the manual segmentation).

Conclusions

A novel method for fully automatic segmentation of the LV endocardium and epicardium from different MRI datasets is presented. The proposed workflow is accurate and robust compared to the reference and other state-of-the-art methods.

Left ventricular geometry during unloading and the end-systolic pressure volume relationship: Measurement with a modified real-time MRI-based method in normal sheep

The left ventricular (LV) end-systolic (ES) pressure volume relationship (ESPVR) is the cornerstone of systolic LV function analysis. We describe a 2D real-time (RT) MRI-based method (RTPVR) with separate software tools for 1) semi-automatic level set-based shape prior method (LSSPM) of the LV, 2) generation of synchronized pressure area loops and 3) calculation of the ESPVR. We used the RTPVR method to measure ventricular geometry, ES pressure area relationship (ESPAR) and ESPVR during vena cava occlusion (VCO) in normal sheep. 14 adult sheep were anesthetized and underwent measurement of LV systolic function. Ten of the 14 sheep underwent RTMRI and eight of the 14 underwent measurement with conductance catheter; 4 had both RTMRI and conductance measurements. 2D cross sectional RTMRI were performed at apex, mid-ventricle and base levels during separate VCOs. The Dice similarity coefficient was used to compare LSSPM and manual image segmentation and thus determine LSSPM accuracy. LV cross-sectional area, major and minor axis length, axis ratio, major axis orientation angle and ESPAR were measured at each LV level. ESPVR was calculated with a trapezoidal rule. The Dice similarity coefficient between LSSPM and manual segmentation by two readers was 87.31±2.51% and 88.13±3.43%. All cross sections became more elliptical during VCO. The major axis orientation shifted during VCO but remained in the septo-lateral direction. LV chamber obliteration at the apical level occurred during VCO in 7 of 10 sheep that underwent RTMRI. ESPAR was non-linear at all levels. Finally, ESPVR was non-linear because of apical collapse. ESPVR measured by conductance catheter (E_ES,Index = 2.23±0.66 mmHg/ml/m²) and RT (E_ES,Index = 2.31±0.31 mmHg/ml/m²) was not significantly different. LSSPM segmentation of 2D RT MRI images is accurate and allows calculation of LV geometry, ESPAR and ESPVR during VCO. In the future, RTPVR will facilitate determination of regional systolic material parameters underlying ESPVR.

Extending Cardiac Functional Assessment with Respiratory-Resolved 3D Cine MRI

This study aimed to develop a cardiorespiratory-resolved 3D magnetic resonance imaging (5D MRI: x-y-z-cardiac-respiratory) approach based on 3D motion tracking for investigating the influence of respiration on cardiac ventricular function. A highly-accelerated 2.5-minute sparse MR protocol was developed for a continuous acquisition of cardiac images through multiple cardiac and respiratory cycles. The heart displacement along respiration was extracted using a 3D image deformation algorithm, and this information was used to cluster the acquired data into multiple respiratory phases. The proposed approach was tested in 15 healthy volunteers (7 females). Cardiac function parameters, including the end-systolic volume (ESV), end-diastolic volume (EDV), stroke volume (SV), and ejection fraction (EF), were measured for the left and right ventricle in both end-expiration and end-inspiration. Although with the proposed 5D cardiac MRI, there were no significant differences (p > 0.05, t-test) between end-expiration and end-inspiration measurements of the cardiac function in volunteers, incremental respiratory motion parameters that were derived from 3D motion tracking, such as the depth, expiration and inspiration distribution, correlated (p < 0.05, correlation coefficient, Mann-Whitney) with those volume-based parameters of cardiac function and varied between genders. The obtained initial results suggested that this new approach allows evaluation of cardiac function during specific respiratory phases. Thus, it can enable investigation of effects related to respiratory variability and better assessment of cardiac function for studying respiratory and/or cardiac dysfunction.

Fully automatic segmentation of 4D MRI for cardiac functional measurements

PURPOSE

Segmentation of cardiac medical images, an important step in measuring cardiac function, is usually performed either manually or semiautomatically. Fully automatic segmentation of the left ventricle (LV), the right ventricle (RV) as well as the myocardium of three-dimensional (3D) magnetic resonance (MR) images throughout the entire cardiac cycle (four-dimensional, 4D), remains challenging. This study proposes a deformable-based segmentation methodology for efficiently segmenting 4D (3D + t) cardiac MR images.

METHODS

The proposed methodology first used the Hough transform and the local Gaussian distribution method (LGD) to segment the LV endocardial contours from cardiac MR images. Following this, a novel level set-based shape prior method was applied to generate the LV epicardial contours and the RV boundary.

RESULTS

This automatic image segmentation approach has been applied to studies on 17 subjects. The results demonstrated that the proposed method was efficient compared to manual segmentation, achieving a segmentation accuracy with average Dice values of 88.62 ± 5.47%, 87.35 ± 7.26%, and 82.63 ± 6.22% for the LV endocardial, LV epicardial, and RV contours, respectively.

CONCLUSIONS

We have presented a method for accurate LV and RV segmentation. Compared to three existing methods, the proposed method can successfully segment the LV and yield the highest Dice value. This makes it an option for clinical assessment of the volume, size, and thickness of the ventricles.

Towards optimal flow diverter porosity for the treatment of intracranial aneurysm

PURPOSE

Low-porosity endovascular stents, known as flow diverters (FDs), have been proposed as an effective and minimally invasive treatment for sidewall intracranial aneurysms (IAs). Although it has been reported that the efficacy of a FD is substantially influenced by its porosity, clinical doctors would clearly prefer to do their interventions optimally based on refined quantitative data. This study focuses on the association between the porosity configurations and the FD efficacy, in order to provide practical data to help the clinical doctors optimize the interventions.

METHOD

Numerical simulations in fluid dynamics were performed using four patient-specific IA geometries, pulsatile velocity profiles and braided fully resolved FDs. The variation of velocity and wall shear stress within the IAs, were investigated in this study. Lattice Boltzmann method (LBM) was used to solve the main challenge centered on the diversity of spatial scales since the typical diameter of struts of FDs is only 25μm while the artery normally can be larger by a hundred times.

RESULTS

Numerical simulations revealed that the blood flow within IA sac was substantially reduced when the porosity is less than 86%. In particular, the flow condition within each IA sac is favorite to initialize thrombus formation when porosity is less than 70%.

CONCLUSION

Our study suggests the existence of a porosity threshold below which the efficacy of a FD will be sufficient for the patients to initialize the thrombus formation. Therefore, by estimating the porosity of FD on patient-specific information, it may be potentially to predict whether or the blood flow condition will successfully become prothrombotic after the FD intervention.

Segmentation of lumen and outer wall of abdominal aortic aneurysms from 3D black-blood MRI with a registration based geodesic active contour model

Segmentation of the geometric morphology of abdominal aortic aneurysm is important for interventional planning. However, the segmentation of both the lumen and the outer wall of aneurysm in magnetic resonance (MR) image remains challenging. This study proposes a registration based segmentation methodology for efficiently segmenting MR images of abdominal aortic aneurysms. The proposed methodology first registers the contrast enhanced MR angiography (CE-MRA) and black-blood MR images, and then uses the Hough transform and geometric active contours to extract the vessel lumen by delineating the inner vessel wall directly from the CE-MRA. The proposed registration based geometric active contour is applied to black-blood MR images to generate the outer wall contour. The inner and outer vessel wall are then fused presenting the complete vessel lumen and wall segmentation. The results obtained from 19 cases showed that the proposed registration based geometric active contour model was efficient and comparable to manual segmentation and provided a high segmentation accuracy with an average Dice value reaching 89.79%.