Quantitative Evaluation of the Performance of a New Test Bolus–Based Computed Tomographic Angiography Contrast-Enhancement–Prediction Algorithm:

Artificial Intelligence in CT Angiography for the Detection of Coronary Artery Stenosis and Calcified Plaque: A Systematic Review and Meta-analysis

PURPOSE

We aimed to evaluate the diagnostic performance of artificial intelligence (AI) in detecting coronary artery stenosis and calcified plaque on CT angiography (CTA), comparing its diagnostic performance with that of radiologists.

METHODS

A thorough search of the literature was performed using PubMed, Web of Science, and Embase, focusing on studies published until October 2024. Studies were included if they evaluated AI models in detecting coronary artery stenosis and calcified plaque on CTA. A bivariate random-effects model was employed to determine combined sensitivity and specificity. Study heterogeneity was assessed using I2 statistics. The risk of bias was assessed using the revised quality assessment of diagnostic accuracy studies-2 tool, and the evidence level was graded using the Grading of Recommendations Assessment, Development and Evalutiuon (GRADE) system.

RESULTS

Out of 1071 initially identified studies, 17 studies with 5560 patients and images were ultimately included for the final analysis. For coronary artery stenosis ≥50%, AI showed a sensitivity of 0.92 (95% CI: 0.88-0.95), specificity of 0.87 (95% CI: 0.80-0.92), and AUC of 0.96 (95% CI: 0.94-0.97), outperforming radiologists with sensitivity of 0.85 (95% CI: 0.67-0.94), specificity of 0.84 (95% CI: 0.62-0.94), and AUC of 0.91 (95% CI: 0.89-0.93). For stenosis ≥70%, AI achieved a sensitivity of 0.88 (95% CI: 0.70-0.96), specificity of 0.96 (95% CI: 0.90-0.99), and AUC of 0.98 (95% CI: 0.96-0.99). In calcified plaque detection, AI demonstrated a sensitivity of 0.93 (95% CI: 0.84-0.97), specificity of 0.94 (95% CI: 0.88-0.96), and AUC of 0.98 (95% CI: 0.96-0.99)."

CONCLUSION

AI-based CT demonstrated superior diagnostic performance compared to clinicians in identifying ≥50% stenosis in coronary arteries and showed excellent diagnostic performance in recognizing ≥70% coronary artery stenosis and calcified plaque. However, limitations include retrospective study designs and heterogeneity in CTA technologies. Further external validation through prospective, multicenter trials is required to confirm these findings.

DATA AVAILABILITY STATEMENT

The original findings of this research are included in the article. For additional inquiries, please contact the corresponding authors.

Image quality improvement in head and neck CT angiography: Individualized post-trigger delay versus fixed delay

PURPOSE

To compare the contrast media opacification of head and neck CT angiography (CTA) between conventional fixed trigger delay and individualized post-trigger delay (PTD).

METHODS

In this prospective study (April-October 2022), 196 consecutive participants were randomly divided into two groups to perform head and neck CTA in bolus tracking with either an individualized PTD (Group A) or a fixed 4-second PTD (Group B). All CT and contrast media protocol parameters were consistent between the two groups. One reader evaluated objective image quality, while two readers rated subjective image quality. Objective image quality was compared between groups via two-sample t-test, while the subjective ratings were compared with chi-square analysis.

RESULTS

Participants' clinical information including sex, age, weight, body weight index (BMI), and heart rate were not statistically different between two groups (all p > 0.05). Individualized PTD ranging from 3.5 to 7.9 s (average 5.6 s), which is shorter than fixed delays (p < 0.05). Both readers rated better subjective image quality for the Group A (p < 0.05). The mean vessel enhancement was significantly higher in Group A in all vessels (all p < 0.05).

CONCLUSIONS

Compared to the fixed post-trigger delay in bolus tracking technique, individualized post-trigger delay could achieve reliable scan timing, optimize vessel opacification and obtain better image quality for head and neck CT angiography.

Milestones in CT: Past, Present, and Future

In 1971, the first patient CT examination by Ambrose and Hounsfield paved the way for not only volumetric imaging of the brain but of the entire body. From the initial 5-minute scan for a 180° rotation to today's 0.24-second scan for a 360° rotation, CT technology continues to reinvent itself. This article describes key historical milestones in CT technology from the earliest days of CT to the present, with a look toward the future of this essential imaging modality. After a review of the beginnings of CT and its early adoption, the technical steps taken to decrease scan times-both per image and per examination-are reviewed. Novel geometries such as electron-beam CT and dual-source CT have also been developed in the quest for ever-faster scans and better in-plane temporal resolution. The focus of the past 2 decades on radiation dose optimization and management led to changes in how exposure parameters such as tube current and tube potential are prescribed such that today, examinations are more customized to the specific patient and diagnostic task than ever before. In the mid-2000s, CT expanded its reach from gray-scale to color with the clinical introduction of dual-energy CT. Today's most recent technical innovation-photon-counting CT-offers greater capabilities in multienergy CT as well as spatial resolution as good as 125 μm. Finally, artificial intelligence is poised to impact both the creation and processing of CT images, as well as automating many tasks to provide greater accuracy and reproducibility in quantitative applications.

Patient-specific post-trigger delay in coronary CT angiography: A prospective study comparing with fixed delay

OBJECTIVES

To validate the peak enhancement timing of a patient-specific post-trigger delay (PTD) in Coronary CT angiography (CCTA) and compare its image quality against a fixed PTD.

METHODS

In this prospective study, 204 consecutive participants were randomly divided into two groups to perform CCTA in bolus tracking with either a fixed 5-second PTD (Group A) or a patient-specific PTD (Group B). Test bolus was also performed in Group B to determine the reference peak enhancement timing. One reader evaluated objective image quality, while two readers rated subjective image quality. The predicted PTD was validated through correlation and agreement analysis with the reference measurement. Objective image quality was compared between groups via two-sample t-test and linear regression, while the subjective ratings were compared with chi-square analysis.

RESULTS

The two groups each had 102 participants with comparable characteristics (52.9 ± 11.3 versus 52.1 ± 11.3 years of age, and 53 versus 52 males). The scan timing from patient-specific PTD demonstrated strong correlation (R = 0.77) and consistency (ICC = 0.618) with the reference peak timing. Both readers rated better subjective image quality for the Group B (p < 0.001). The mean vessel enhancement was significantly higher in Group B in all coronary vessels (all p < 0.05). After adjusting for the participant variation, the patient-specific PTD strategy was associated with an average of 33.5 HU higher enhancement compared to the fixed PTD.

CONCLUSIONS

Patient-specific delay could achieve reliable scan timing, optimize vessel opacification and obtain better image quality in CCTA.

Artificial Intelligence in Symptomatic Carotid Plaque Detection: A Narrative Review

Identifying atherosclerotic disease is the mainstay for the correct diagnosis of the large artery atherosclerosis ischemic stroke subtype and for choosing the right therapeutic strategy in acute ischemic stroke. Classification into symptomatic and asymptomatic plaque and estimation of the cardiovascular risk are essential to select patients eligible for pharmacological and/or surgical therapy in order to prevent future cerebral ischemic events. The difficulties in a “vulnerability” definition and the methodical issues concerning its detectability and quantification are still subjects of debate. Non-invasive imaging studies commonly used to detect arterial plaque are computed tomographic angiography, magnetic resonance imaging, and ultrasound. Characterization of a carotid plaque type using the abovementioned imaging modalities represents the basis for carotid atherosclerosis management. Classification into symptomatic and asymptomatic plaque and estimation of the cardiovascular risk are essential to select patients eligible for pharmacological and/or surgical therapy in order to prevent future cerebral ischemic events. In this setting, artificial intelligence (AI) can offer suggestive solutions for tissue characterization and classification concerning carotid artery plaque imaging by analyzing complex data and using automated algorithms to obtain a final output. The aim of this review is to provide overall knowledge about the role of AI models applied to non-invasive imaging studies for the detection of symptomatic and vulnerable carotid plaques.

Artificial intelligence in computed tomography plaque characterization: A review

Cardiovascular disease (CVD) is associated with high mortality around the world. Prevention and early diagnosis are key targets in reducing the socio-economic burden of CVD. Artificial intelligence (AI) has experienced a steady growth due to technological innovations that have to lead to constant development. Several AI algorithms have been applied to various aspects of CVD in order to improve the quality of image acquisition and reconstruction and, at the same time adding information derived from the images to create strong predictive models. In computed tomography angiography (CTA), AI can offer solutions for several parts of plaque analysis, including an automatic assessment of the degree of stenosis and characterization of plaque morphology. A growing body of evidence demonstrates a correlation between some type of plaques, so-called high-risk plaque or vulnerable plaque, and cardiovascular events, independent of the degree of stenosis. The radiologist must apprehend and participate actively in developing and implementing AI in current clinical practice. In this current overview on the existing AI literature, we describe the strengths, limitations, recent applications, and promising developments of employing AI to plaque characterization with CT.

Machine Learning and Deep Neural Networks: Applications in Patient and Scan Preparation, Contrast Medium, and Radiation Dose Optimization

Artificial intelligence (AI) algorithms are dependent on a high amount of robust data and the application of appropriate computational power and software. AI offers the potential for major changes in cardiothoracic imaging. Beyond image processing, machine learning and deep learning have the potential to support the image acquisition process. AI applications may improve patient care through superior image quality and have the potential to lower radiation dose with AI-driven reconstruction algorithms and may help avoid overscanning. This review summarizes recent promising applications of AI in patient and scan preparation as well as contrast medium and radiation dose optimization.

Coronary angiography using second-generation dual source computed tomography: Feasibility of low dose and low flow rate to achieve appropriate individual contrast enhancement using a test bolus-based contrast medium protocol-A CONSORT compliant article

Improved contrast enhancement consistency can be achieved using an individualized contrast media (CM) protocol. This study aimed to assess the feasibility of a low-dose, low-flow rate CM protocol to achieve appropriate individual contrast enhancement using a newly advocated individualized test bolus-based protocol for second-generation dual-source computed tomography angiography.CM containing iodine (370 mg I/mL) was used in this study. A CM flow rate of 3.5 mL/s for patients with a body mass index (BMI) <25.0 kg/m, and 4.5 mL/s for those with BMI ≥25.0 kg/m was used in group 1 (n = 189). An individualized test-bolus based contrast injection protocol was then derived from the information gained from the test bolus and coronary enhancements in group 1. The proposed individualized test-bolus based CM injection protocol was applied in group 2 (n = 219). Ascending aortic attenuations (AAo) were measured and compared with both groups.The contrast enhancement consistency of AAo in group 2 improved significantly (31.8 vs 56.3 Hounsfield units [HU]; P < .001). The number of patients in group 2 with a contrast flow rate ≤3 mL/s was 63 (28.8%), with 77 (35.2%) using a contrast dose ≤40 mL. In group 2, no significant differences in mean AAo were found among subgroups with contrast flow rates ≤3.0, 3.1 to 4.0, 4.1 to 5.0 and >5.0 mL/s (351, 344, 346, and 348 HU, respectively), nor among subgroups with contrast doses ≤40, 41 to 50, 51 to 60, and >60 mL (349, 345, 344, and 350 HU, respectively).Improved individual contrast enhancement uniformity can be achieved using an individualized CM protocol tailored to a test bolus. Approximately, one-third of patients received CM at a flow rate of no more than 3 mL/s and a total dose of no more than 40 mL.

Relationship between low tube voltage (70 kV) and the iodine delivery rate (IDR) in CT angiography: An experimental in-vivo study

Objective

Very short acquisition times and the use of low-kV protocols in CTA demand modifications in the contrast media (CM) injection regimen. The aim of this study was to optimize the use of CM delivery parameters in thoraco-abdominal CTA in a porcine model.

Materials and methods

Six pigs (55–68 kg) were examined with a dynamic CTA protocol (454 mm scan length, 2.5 s temporal resolution, 70 s total acquisition time). Four CM injection protocols were applied in a randomized order. 120 kV CTA protocol: (A) 300 mg iodine/kg bodyweight (bw), IDR = 1.5 g/s (flow = 5 mL/s), injection time (t_i) 12 s (60 kg bw). 70 kV CTA protocols: 150 mg iodine/kg bw: (B) IDR = 0.75 g/s (flow = 2.5 mL/s), t_i = 12 s (60 kg bw); (C) IDR = 1.5 g/s (flow = 5 mL/s), t_i = 12 s (60 kg bw); (D) IDR = 3.0 g/s (flow = 10 mL/s), t_i = 3 s (60 kg bw). The complete CM bolus shape was monitored by creating time attenuation curves (TAC) in different vascular territories. Based on the TAC, the time to peak (TTP) and the peak enhancement were determined. The diagnostic window (relative enhancement > 300 HU), was calculated and compared to visual inspection of the corresponding CTA data sets.

Results

The average relative arterial peak enhancements after baseline correction were 358.6 HU (A), 356.6 HU (B), 464.0 HU (C), and 477.6 HU (D). The TTP decreased with increasing IDR and decreasing t_i, protocols A and B did not differ significantly (systemic arteries, p = 0.843; pulmonary arteries, p = 0.183). The delay time for bolus tracking (trigger level 100 HU; target enhancement 300 HU) for single-phase CTA was comparable for protocol A and B (3.9, 4.3 s) and C and D (2.4, 2.0 s). The scan window time frame was comparable for the different protocols by visual inspection of the different CTA data sets and by analyzing the TAC.

Conclusions

All protocols provided sufficient arterial enhancement. The use of a 70 kV CTA protocol is recommended because of a 50% reduction of total CM volume and a 50% reduced flow rate while maintaining the bolus profile. In contrast to pulmonary arterial enhancement, the systemic arterial enhancement improved only slightly increasing the IDR from 1.5 g/s to 3 g/s because of bolus dispersion of the very short bolus (3s) in the lungs.

Feasibility of a low concentration test bolus in CT angiography

AIM

To investigate the feasibility of using a low-concentration test bolus in abdominal aorta computed tomography (CT) angiography (CTA).

MATERIALS AND METHODS

In 10 patients referred for CTA of the abdominal aorta with a body mass index (BMI) ≤28 kg/m², a standard test bolus of 10 ml contrast medium (CM; 350 mg iodine/ml) was compared with a low-concentration test bolus (5 ml CM; 350 mg iodine/ml; 1:1 diluted with saline) in terms of time to peak enhancement (tPE) and peak enhancement (PE).

RESULTS

No significant differences were found between the standard and low-concentration test bolus in terms of tPE and PE.

CONCLUSIONS

A low-concentration test bolus (5 ml, 1:1 diluted with saline) is feasible in patients with a BMI ≤28 kg/m².

Revival of monophasic contrast injection protocols: superiority of a monophasic injection protocol compared to a biphasic injection protocol in high-pitch CT angiography

BACKGROUND

Biphasic injection protocols are frequently used because they yield homogenous contrast enhancement. We hypothesize that with faster scanners and shorter scan times, biphasic injection protocols are no longer necessary.

PURPOSE

To evaluate whether a monophasic injection protocol is equivalent to a biphasic protocol in terms of contrast enhancement and homogeneity.

MATERIAL AND METHODS

Repeated high-pitch CTA (pitch 3) and conventional standard-pitch computed tomography angiography (CTA) (pitch 1.2) from the cervical region to the symphysis was performed in seven beagles (11.2 ± 2.5 kg) in a cross-over study design. Arterial contrast enhancement was measured along the z-axis in the ascending, descending, and abdominal aorta and the iliac arteries. The z-axis is the longitudinal axis of the human body and at the same time the direction in which the CT table is moving. The data were analyzed using repeated measures ANOVA with a post-hoc t-test and visual assessment of the scans.

RESULTS

In high-pitch CTA, monophasic injection protocols were superior to biphasic injection protocols in enhancement levels (P < 0.05) and enhancement homogeneity along the z-axis (P < 0.05). In conventional CTA, enhancement levels did not differ. Contrast homogeneity was better for biphasic protocols.

CONCLUSION

High-pitch CTA monophasic injection protocols are superior to biphasic injection protocols, due to a higher and more homogeneous contrast enhancement with the same amount of contrast medium used.

Contrast Gradient-Based Blood Velocimetry With Computed Tomography: Theory, Simulations, and Proof of Principle in a Dynamic Flow Phantom

OBJECTIVES

The aim of this study was to introduce a new theoretical framework describing the relationship between the blood velocity, computed tomography (CT) acquisition velocity, and iodine contrast enhancement in CT images, and give a proof of principle of contrast gradient-based blood velocimetry with CT.

MATERIALS AND METHODS

The time-averaged blood velocity (v(blood)) inside an artery along the axis of rotation (z axis) is described as the mathematical division of a temporal (Hounsfield unit/second) and spatial (Hounsfield unit/centimeter) iodine contrast gradient. From this new theoretical framework, multiple strategies for calculating the time-averaged blood velocity from existing clinical CT scan protocols are derived, and contrast gradient-based blood velocimetry was introduced as a new method that can calculate v(blood) directly from contrast agent gradients and the changes therein. Exemplarily, the behavior of this new method was simulated for image acquisition with an adaptive 4-dimensional spiral mode consisting of repeated spiral acquisitions with alternating scan direction. In a dynamic flow phantom with flow velocities between 5.1 and 21.2 cm/s, the same acquisition mode was used to validate the simulations and give a proof of principle of contrast gradient-based blood velocimetry in a straight cylinder of 2.5 cm diameter, representing the aorta.

RESULTS

In general, scanning with the direction of blood flow results in decreased and scanning against the flow in increased temporal contrast agent gradients. Velocity quantification becomes better for low blood and high acquisition speeds because the deviation of the measured contrast agent gradient from the temporal gradient will increase. In the dynamic flow phantom, a modulation of the enhancement curve, and thus alternation of the contrast agent gradients, can be observed for the adaptive 4-dimensional spiral mode and is in agreement with the simulations. The measured flow velocities in the downslopes of the enhancement curves were in good agreement with the expected values, although the accuracy and precision worsened with increasing flow velocities.

CONCLUSIONS

The new theoretical framework increases the understanding of the relationship between the blood velocity, CT acquisition velocity, and iodine contrast enhancement in CT images, and it interconnects existing blood velocimetry methods with research on transluminary attenuation gradients. With these new insights, novel strategies for CT blood velocimetry, such as the contrast gradient-based method presented in this article, may be developed.