Clinical Validation of a Deep Learning Algorithm for Automated Coronary Artery Disease Detection and Classification Using a Heterogeneous Multivendor Coronary Computed Tomography Angiography Data Set.

PURPOSE: We sought to clinically validate a fully automated deep learning (DL) algorithm for coronary artery disease (CAD) detection and classification in a heterogeneous multivendor cardiac computed tomography angiography data set. MATERIALS AND METHODS: In this single-centre retrospective study, we included patients who underwent cardiac computed tomography angiography scans between 2010 and 2020 with scanners from 4 vendors (Siemens Healthineers, Philips, General Electrics, and Canon). Coronary Artery Disease-Reporting and Data System (CAD-RADS) classification was performed by a DL algorithm and by an expert reader (reader 1, R1), the gold standard. Variability analysis was performed with a second reader (reader 2, R2) and the radiologic reports on a subset of cases. Statistical analysis was performed stratifying patients according to the presence of CAD (CAD-RADS >0) and obstructive CAD (CAD-RADS ≥3). RESULTS: Two hundred ninety-six patients (average age: 53.66 ± 13.65, 169 males) were enrolled. For the detection of CAD only, the DL algorithm showed sensitivity, specificity, accuracy, and area under the curve of 95.3%, 79.7%, 87.5%, and 87.5%, respectively. For the detection of obstructive CAD, the DL algorithm showed sensitivity, specificity, accuracy, and area under the curve of 89.4%, 92.8%, 92.2%, and 91.1%, respectively. The variability analysis for the detection of obstructive CAD showed an accuracy of 92.5% comparing the DL algorithm with R1, and 96.2% comparing R1 with R2 and radiology reports. The time of analysis was lower using the DL algorithm compared with R1 (P < 0.001). CONCLUSIONS: The DL algorithm demonstrated robust performance and excellent agreement with the expert readers' analysis for the evaluation of CAD, which also corresponded with significantly reduced image analysis time.

AI-based, automated chamber volumetry from gated, non-contrast CT.

BACKGROUND: Accurate chamber volumetry from gated, non-contrast cardiac CT (NCCT) scans can be useful for potential screening of heart failure. OBJECTIVES: To validate a new, fully automated, AI-based method for cardiac volume and myocardial mass quantification from NCCT scans compared to contrasted CT Angiography (CCTA). METHODS: Of a retrospectively collected cohort of 1051 consecutive patients, 420 patients had both NCCT and CCTA scans at mid-diastolic phase, excluding patients with cardiac devices. Ground truth values were obtained from the CCTA scans. RESULTS: The NCCT volume computation shows good agreement with ground truth values. Volume differences [95% CI ] and correlation coefficients were: -9.6 [-45; 26] mL, r â€‹= â€‹0.98 for LV Total, -5.4 [-24; 13] mL, r â€‹= â€‹0.95 for LA, -8.7 [-45; 28] mL, r â€‹= â€‹0.94 for RV, -5.2 [-27; 17] mL, r â€‹= â€‹0.92 for RA, -3.2 [-42; 36] mL, r â€‹= â€‹0.91 for LV blood pool, and -6.7 [-39; 26] g, r â€‹= â€‹0.94 for LV wall mass, respectively. Mean relative volume errors of less than 7% were obtained for all chambers. CONCLUSIONS: Fully automated assessment of chamber volumes from NCCT scans is feasible and correlates well with volumes obtained from contrast study.

Coronary CT Angiography-derived Fractional Flow Reserve: Machine Learning Algorithm versus Computational Fluid Dynamics Modeling.

Purpose To compare two technical approaches for determination of coronary computed tomography (CT) angiography-derived fractional flow reserve (FFR)-FFR derived from coronary CT angiography based on computational fluid dynamics (hereafter, FFRCFD) and FFR derived from coronary CT angiography based on machine learning algorithm (hereafter, FFRML)-against coronary CT angiography and quantitative coronary angiography (QCA). Materials and Methods A total of 85 patients (mean age, 62 years ± 11 [standard deviation]; 62% men) who had undergone coronary CT angiography followed by invasive FFR were included in this single-center retrospective study. FFR values were derived on-site from coronary CT angiography data sets by using both FFRCFD and FFRML. The performance of both techniques for detecting lesion-specific ischemia was compared against visual stenosis grading at coronary CT angiography, QCA, and invasive FFR as the reference standard. Results On a per-lesion and per-patient level, FFRML showed a sensitivity of 79% and 90% and a specificity of 94% and 95%, respectively, for detecting lesion-specific ischemia. Meanwhile, FFRCFD resulted in a sensitivity of 79% and 89% and a specificity of 93% and 93%, respectively, on a per-lesion and per-patient basis (P = .86 and P = .92). On a per-lesion level, the area under the receiver operating characteristics curve (AUC) of 0.89 for FFRML and 0.89 for FFRCFD showed significantly higher discriminatory power for detecting lesion-specific ischemia compared with that of coronary CT angiography (AUC, 0.61) and QCA (AUC, 0.69) (all P < .0001). Also, on a per-patient level, FFRML (AUC, 0.91) and FFRCFD (AUC, 0.91) performed significantly better than did coronary CT angiography (AUC, 0.65) and QCA (AUC, 0.68) (all P < .0001). Processing time for FFRML was significantly shorter compared with that of FFRCFD (40.5 minutes ± 6.3 vs 43.4 minutes ± 7.1; P = .042). Conclusion The FFRML algorithm performs equally in detecting lesion-specific ischemia when compared with the FFRCFD approach. Both methods outperform accuracy of coronary CT angiography and QCA in the detection of flow-limiting stenosis.

A machine-learning approach for computation of fractional flow reserve from coronary computed tomography.

Fractional flow reserve (FFR) is a functional index quantifying the severity of coronary artery lesions and is clinically obtained using an invasive, catheter-based measurement. Recently, physics-based models have shown great promise in being able to noninvasively estimate FFR from patient-specific anatomical information, e.g., obtained from computed tomography scans of the heart and the coronary arteries. However, these models have high computational demand, limiting their clinical adoption. In this paper, we present a machine-learning-based model for predicting FFR as an alternative to physics-based approaches. The model is trained on a large database of synthetically generated coronary anatomies, where the target values are computed using the physics-based model. The trained model predicts FFR at each point along the centerline of the coronary tree, and its performance was assessed by comparing the predictions against physics-based computations and against invasively measured FFR for 87 patients and 125 lesions in total. Correlation between machine-learning and physics-based predictions was excellent (0.9994, P < 0.001), and no systematic bias was found in Bland-Altman analysis: mean difference was -0.00081 ± 0.0039. Invasive FFR ≤ 0.80 was found in 38 lesions out of 125 and was predicted by the machine-learning algorithm with a sensitivity of 81.6%, a specificity of 83.9%, and an accuracy of 83.2%. The correlation was 0.729 (P < 0.001). Compared with the physics-based computation, average execution time was reduced by more than 80 times, leading to near real-time assessment of FFR. Average execution time went down from 196.3 ± 78.5 s for the CFD model to ∼2.4 ± 0.44 s for the machine-learning model on a workstation with 3.4-GHz Intel i7 8-core processor.

Coronary CT angiography derived fractional flow reserve: Methodology and evaluation of a point of care algorithm.

BACKGROUND: Recently several publications described the diagnostic value of coronary CT angiography (coronary CTA) derived fractional flow reserve (CTA-FFR). For a recently introduced on-site CTA-FFR application, detailed methodology and factors potentially affecting performance have not yet been described. OBJECTIVE: To provide a methodological background for an on-site CTA-FFR application and evaluate the effect of patient and acquisition characteristics. METHODS: The on-site CTA-FFR application utilized a reduced-order hybrid model applying pressure drop models within stenotic regions. In 116 patients and 203 vessels the diagnostic performance of CTA-FFR was investigated using invasive FFR measurements as a reference. The effect of several potentially relevant factors on CTA-FFR was investigated. RESULTS: 90 vessels (44%) had a hemodynamically relevant stenosis according to invasive FFR (threshold ≤0.80). The overall vessel-based sensitivity, specificity and accuracy of CTA-FFR were 88% (CI 95%:79-94%), 65% (55-73%) and 75% (69-81%). The specificity was significantly lower in the presence of misalignment artifacts (25%, CI: 6-57%). A non-significant reduction in specificity from 74% (60-85%) to 48% (26-70%) was found for higher coronary artery calcium scores. Left ventricular mass, diabetes mellitus and large vessel size increased the discrepancy between invasive FFR and CTA-FFR values. CONCLUSIONS: On-site calculation of CTA-FFR can identify hemodynamically significant CAD with an overall per-vessel accuracy of 75% in comparison to invasive FFR. The diagnostic performance of CTA-FFR is negatively affected by misalignment artifacts. CTA-FFR is potentially affected by left ventricular mass, diabetes mellitus and vessel size.

Image-based computational models for TAVI planning: from CT images to implant deployment.

Transcatheter aortic valve implantation (TAVI) is becoming the standard choice of care for non-operable patients suffering from severe aortic valve stenosis. As there is no direct view or access to the affected anatomy, accurate preoperative planning is crucial for a successful outcome. The most important decision during planning is selecting the proper implant type and size. Due to the wide variety in device sizes and types and non-circular annulus shapes, there is often no obvious choice for the specific patient. Most clinicians base their final decision on their previous experience. As a first step towards a more predictive planning, we propose an integrated method to estimate the aortic apparatus from CT images and compute implant deployment. Aortic anatomy, which includes aortic root, leaflets and calcifications, is automatically extracted using robust modeling and machine learning algorithms. Then, the finite element method is employed to calculate the deployment of a TAVI implant inside the patient-specific aortic anatomy. The anatomical model was evaluated on 198 CT images, yielding an accuracy of 1.30 +/- 0.23 mm. In eleven subjects, pre- and post-TAVI CT images were available. Errors in predicted implant deployment were of 1.74 +/- 0.40 mm in average and 1.32 mm in the aortic valve annulus region, which is almost three times lower than the average gap of 3 mm between consecutive implant sizes. Our framework may thus constitute a surrogate tool for TAVI planning.

Three-dimensional visualisation improves understanding of surgical liver anatomy.

OBJECTIVES: Three-dimensional (3-D) representation is thought to improve understanding of complex spatial interactions and is being used more frequently in diagnostic and therapeutic procedures. It has been suggested that males benefit more than females from 3-D presentations. There have been few randomised trials to confirm these issues. We carried out a randomised trial, based on the identification of complex surgical liver anatomy, to evaluate whether 3-D presentation has a beneficial impact and if gender differences were evident. METHODS: A computer-based teaching module (TM) was developed to test whether two-dimensional (2-D) computed tomography (CT) images or 3-D presentations result in better understanding of liver anatomy. Following a PowerPoint lecture, students were randomly selected to participate in computer-based testing which used either 2-D images presented as consecutive transversal slices, or one of two 3-D variations. In one of these the vessel tree of portal and hepatic veins was shown in one colour (3-D) and in the other the two vessel systems were coloured differently (3-Dc). Participants were asked to answer 11 medical questions concerning surgical anatomy and four questions on their subjective assessment of the TM. RESULTS: Of the 160 Year 4 and 5 medical students (56.8% female) who participated in this prospective randomised trial, students exposed to 3-D presentation performed significantly better than those exposed to 2-D images (p < 0.001). Comparison of the number of correct answers revealed no significant differences between the 3-D and 3-Dc modalities p > 0.1). Male students gave significantly more correct answers in the 3-D and 3-Dc modalities than female students (p < 0.03). The gender difference observed in both 3-D modalities was not evident in the 2-D group (p = 0.21). CONCLUSIONS: This study showed that 3-D imaging significantly improved the identification of complex surgical liver anatomy. Male students benefited significantly more than female students from 3-D presentations. Use of colour in 3-D presentation did not improve student performance.

Estimation of pulmonary motion in healthy subjects and patients with intrathoracic tumors using 3D-dynamic MRI: initial results.

OBJECTIVE: To estimate a new technique for quantifying regional lung motion using 3D-MRI in healthy volunteers and to apply the technique in patients with intra- or extrapulmonary tumors. MATERIALS AND METHODS: Intraparenchymal lung motion during a whole breathing cycle was quantified in 30 healthy volunteers using 3D-dynamic MRI (FLASH [fast low angle shot] 3D, TRICKS [time-resolved interpolated contrast kinetics]). Qualitative and quantitative vector color maps and cumulative histograms were performed using an introduced semiautomatic algorithm. An analysis of lung motion was performed and correlated with an established 2D-MRI technique for verification. As a proof of concept, the technique was applied in five patients with non-small cell lung cancer (NSCLC) and 5 patients with malignant pleural mesothelioma (MPM). RESULTS: The correlation between intraparenchymal lung motion of the basal lung parts and the 2D-MRI technique was significant (r = 0.89, p < 0.05). Also, the vector color maps quantitatively illustrated regional lung motion in all healthy volunteers. No differences were observed between both hemithoraces, which was verified by cumulative histograms. The patients with NSCLC showed a local lack of lung motion in the area of the tumor. In the patients with MPM, there was global diminished motion of the tumor bearing hemithorax, which improved significantly after chemotherapy (CHT) (assessed by the 2D- and 3D-techniques) (p < 0.01). Using global spirometry, an improvement could also be shown (vital capacity 2.9 +/- 0.5 versus 3.4 L +/- 0.6, FEV1 0.9 +/- 0.2 versus 1.4 +/- 0.2 L) after CHT, but this improvement was not significant. CONCLUSION: A 3D-dynamic MRI is able to quantify intraparenchymal lung motion. Local and global parenchymal pathologies can be precisely located and might be a new tool used to quantify even slight changes in lung motion (e.g. in therapy monitoring, follow-up studies or even benign lung diseases).

Does preoperative analysis of intrahepatic venous anastomoses improve the surgeon's intraoperative decision making? Pilot data from a case report.

BACKGROUND: Three-dimensional (3D) visualization is thought to improve the anatomical understanding of clinicians, thus improving patient safety. CASE PRESENTATION: A 58-year-old male was admitted to our hospital in April 2007 with a suspected metastasis of a sigmoid cancer in the Couinaud segment (CS) 7. The anatomical situation of this patient was analyzed using both a CT scan and 3D images. The initial CT scan revealed that the proximal part of the middle hepatic vein was completely missing and the metastasis in the CS 7 was closely attached to the right hepatic vein. After analyzing additional 3D images, it became clear that due to the close proximity of metastasis and right hepatic vein, the resection of the right hepatic vein was inevitable. Based on this 3D analysis, it was decided to perform a right-sided hemihepatectomy. In this case report, 3D visualization resulted in a faster and clearer understanding of the unique anatomical situation in a patient with complicated liver anatomy than transversal CT images. CONCLUSION: The here presented data shows for the first time 3D visualization of intravenous anastomoses in the human liver. The information offered by 3D visualization is not redundant but rather serves as a true source of additional information, indicating the potential benefit of 3D visualization in surgical operation planning.

Imaging of extralobar pulmonary sequestration in a patient with tetralogy of Fallot with pulmonary atresia, ventricular septal defect and multiple aorto-pulmonary collateral arteries with multidetector computed tomography.

Complex pulmonary vascular blood supply is common in patients with tetralogy of Fallot with pulmonary atresia, major systemic to pulmonary collateral arteries and hypoplastic or deficient central pulmonary arteries. An extralobar lung sequestration, which has not been described previously in these patients, was imaged in a 6-week-old infant with multidetector computed tomography with sub-millimeter resolution. Arterial and venous vessels were analyzed using three-dimensional vascular exploration tools and results were confirmed with cardiac catheterization.

Visualization of intrarenal vessels by 3.0-T MR angiography in comparison with digital subtraction angiography using renal specimens.

BACKGROUND: MRI at 3.0 T enables high-spatial resolution for renal MR angiography. OBJECTIVE: Evaluation of an arterial tree model in animal kidney specimens with comparison of digital subtraction angiography (DSA) and high-field MRI to find out the maximum spatial resolution of intrarenal vessels. It was considered that objective quantitation of angiogram quality could be achieved. MATERIALS AND METHODS: A total of 27 pig kidney specimens were examined by MR angiography (flash 3D) using a 3.0-T scanner (TRIO; Siemens, Erlangen, Germany) with an eight-channel head coil and a voxel size of 0.9x0.9x1.1 mm in the early arterial phase after implantation of a 4F catheter in the renal artery. DSA (Integris, Philips, Best, The Netherlands) was performed immediately after the MRI. With the help of semiautomated segmentation, all vessels were marked for comparison of the vessel trees. The Wilcoxon rank test was used for statistical evaluation of vessel numbers and branching depths. RESULTS: Objective comparison between DSA and MR angiography was achieved. High-field MR angiography had the ability to depict vessels up to the seventh branching on average. Significant differences in vessel delineation and counts were found from the fifth level of intrarenal branching onwards with DSA showing an advantage. CONCLUSION: High-field MRI has great potential in the detection of intrarenal arteries and is comparable to DSA in visualization of the central intrarenal vessel tree.

Quantification of lung tumor volume and rotation at 3D dynamic parallel MR imaging with view sharing: preliminary results.

The purpose of this study was to prospectively evaluate the volumes and rotations of pulmonary nodules during respiration by using three-dimensional fast low-angle shot dynamic magnetic resonance (MR) imaging (1.5/0.6 [repetition time msec/echo time msec], 3.8 x 3.8 x 3.8-mm voxel size, imaging time per three-dimensional data set of 1 second). The feasibility of the technique was verified by using 130-, 40-, and 12-cm3 phantoms made of meatballs and in five patients with solitary intrapulmonary tumors (four men, one woman; median age, 60 years) at computed tomography and histologic analysis. All patients provided written informed consent, and the study was institutional review board approved. It was proved that there were no substantial differences among the 21 algorithms used to correct partial volume effects. The most precise algorithm (r > 0.9, P < .01) used to correct partial volume effects--with which mean phantom volumes of 120.8 cm3 +/- 4.1, 36.1 cm3 +/- 3.98, and 13.1 cm3 +/- 1.5 were calculated--yielded a root mean square error of 14%. The MR imaging-derived nodule volume and rotation during respiration could be quantified by using oriented bounding box techniques.

Monitoring of lung motion in patients with malignant pleural mesothelioma using two-dimensional and three-dimensional dynamic magnetic resonance imaging: comparison with spirometry.

PURPOSE: To monitor lung motion in patients with malignant pleural mesothelioma (MPM) before and after chemotherapy (CHT) using 2-dimensional (2D) and 3-dimensional (3D) dynamic MRI (dMRI) in comparison with spirometry. METHODS AND MATERIALS: Twenty-two patients with MPM were examined before CHT, as well as after 3 and 6 CHT cycles (3 months and 6 months) using 2D dMRI (trueFISP; 3 images/s) and 3D dMRI (FLASH 3D, 1 slab (52 slices)/s) using parallel imaging in combination with view-sharing technique. Maximum craniocaudal lung dimensions (2D) and lung volumes (3D) were monitored, separated into the tumor-bearing and nontumor-bearing hemithorax. Vital capacity (VC) was measured for comparison using spirometry. RESULTS: Using 2D technique, there was a significant difference between the tumor-bearing and the nontumor-bearing hemithorax before CHT (P < 0.01) and after 3 CHT cycles (P < 0.05), whereas difference was not significant in the second control. In the tumor-bearing hemithorax, mobility increased significantly from the status before versus after 3 CHT cycles (4.1 +/- 1.1 cm vs. 4.8 +/- 1.4 cm, P < 0.05). Using 3D technique, at maximum inspiration, the volume of the tumor-bearing hemithorax was 0.6 +/- 0.4 L and of the nontumor-bearing hemithorax 1.25 +/- 0.4 L before CHT. In the follow-up exams, these volumes changed to 1.05 +/- 0.4 L (P < 0.05) and 1.4 +/- 0.5 L, respectively. Using spirometry, there was no significant change in VC (1.9 +/- 0.4 L vs. 2.2 +/- 0.7 L vs. 2.2 +/- 0.9 L). CONCLUSION: dMRI is capable of monitoring changes in lung motion and volumetry in patients with MPM not detected by global spirometry. Thus, dMRI is proposed for use as a further measure of therapy response.

Imaging of aortopulmonary collateral arteries with high-resolution multidetector CT.

BACKGROUND: Precise visualization of the pulmonary vasculature is mandatory for adequate treatment of patients with pulmonary atresia and ventricular septal defect (PA-VSD). Aortopulmonary collateral arteries (APCs) can be visualized by selective injections of contrast agent in the catheterization laboratory. OBJECTIVE: To evaluate multidetector CT (MDCT) and different image postprocessing methods for analysis of complex pulmonary blood supply in patients with PA-VSD. MATERIALS AND METHODS: Eight patients (6 weeks to 27.8 years of age) with PA-VSD and APCs underwent MDCT and cardiac catheterization. Using multiplanar reformatting, volume rendering and semiautomatic segmentation algorithms, the aorta, pulmonary arteries and APCs were displayed. MDCT and cardiac catheterization were analyzed by two independent observers. RESULTS: MDCT accurately imaged central pulmonary arteries (n=8), aortopulmonary shunts (n=2), right ventricular to pulmonary artery conduits (n=2) and origin, course and intrapulmonary connections of APCs (n=25), compared to X-ray angiography. A high correlation was found between the MDCT vessel diameter measurements by two independent observers (n=70, r=0.96, P<0.01) and between MDCT and angiographic vessel diameter measurements (n=68, r=0.96, P<0.01). CONCLUSIONS: Using three-dimensional imaging software, a complex pulmonary blood supply can be non-invasively and accurately imaged with high-resolution MDCT. This technique may help to reduce the number of cardiac catheterizations or guide interventional or surgical therapy.

Evaluation of lung volumetry using dynamic three-dimensional magnetic resonance imaging.

RATIONALE AND OBJECTIVES: We sought to investigate lung volume and surface measurements during the breathing cycle using dynamic three-dimensional magnetic resonance imaging (3D MRI). MATERIALS AND METHODS: Breathing cycles of 20 healthy volunteers were examined using a 2D trueFISP sequence (3 images/second) in combination with a model and segmented 3D FLASH sequence (1 image/second) MR images using view sharing. Segmentation was performed semiautomatically using an interactive region growing technique. Vital capacity (VC) was calculated from MRI using the model (2D) and counting the voxels (3D) and was compared with spirometry. RESULTS: VC from spirometry was 4.9+/-0.9 L, 4.4+/-1.2 L from 2D MRI measurement, and 4.7+/-0.9 L for 3D MRI. Using the 3D technique, correlation to spirometry was higher than using the 2D technique (r>0.95 vs. r>0.83). Using the 3D technique, split lung volumes and lung surface could be calculated. There was a significant difference between the left and right lung volume in expiration (P<0.05). CONCLUSIONS: Dynamic 3D MRI is a noninvasive tool to evaluate split lung volumes and lung surfaces during the breathing cycle with a high correlation to spirometry.