Introduction

The application of groundbreaking non-invasive imaging techniques in the field of cardiovascular diseases is steadily increasing. Although transthoracic echocardiography remains the primary and most frequently used imaging modality in cardiovascular medicine, there are additional benefits of alternative non-invasive modalities, including cardiovascular computed tomography (CCT), cardiovascular magnetic resonance imaging (CMR), positron emission tomography (PET), and single-photon emission computed tomography (SPECT). Each imaging modality, characterized by unique advantages, is recommended for specific clinical situations based on these distinctive capabilities.

For CCT, hardware and software developments now allow for high-quality scans with low-dose radiation and very short scan times so that the major previous drawbacks associated with the negative effects of radiation are substantially mitigated [1].

Simultaneously with these technological advancements, there have been significant improvements in imaging quality, enabling the acquisition of intricate details during screening processes, valve repair procedures, evaluation of congenital heart diseases (CHD), and even in the differentiation of tissue composition in cardiomyopathies. Current guidelines now regard CCT as a critical preliminary step before invasive procedures, endorsing the use of CT imaging prior to echocardiography in patients presenting with chest pain [2]. In acute scenarios, CCT is now recommended as a class I recommendation to rule out coronary artery disease (CAD) in patients with out-of-hospital cardiac arrest (OHCA) and no signs of STEMI [3,4,5].

As techniques and applications are rapidly evolving, this review aims to provide an overview on the role of CCT in cardiovascular medicine including recent technical advances, and their potential clinical capabilities for state-of-the-art imaging with a particular focus on the trajectory of future developments (see Central Figure).

Technical Considerations – From Hounsfield units to Multi-Energy Scanning

Prior to delving into the clinical applications and recent developments for specific diagnostic inquiries, it is imperative to consider some fundamental technical aspects concerning the hardware and software components of CCT.

Apart from patient characteristics and compliance (for instance, the ability to breath-hold during examinations), image quality in CCT is technically determined by parameters such as temporal, contrast, and spatial resolution [1].

Temporal Resolution

For cardiac imaging in particular, temporal resolution carries significant importance due to the heart's perpetual motion, which generates motion artifacts. These artifacts are aggravated by respiratory motion. Moreover, temporal resolution assumes crucial significance when conducting functional evaluations of the heart or diagnosing CHD, especially in children who tend to have higher heart rates.

Temporal resolution is primarily determined by the gantry rotation time which reaches up to 250 ms in modern scanners [6, 7]. However, as the rotation speed cannot be increased without limitation, other technical developments are required to make further improvements. One innovation in CCT design was the introduction of Dual Source CT (DSCT), where a second X-ray tube and detector is installed within the CT scanner. By utilizing a second X-ray source, DSCT requires approximately one-quarter gantry rotation to enable image reconstruction [8, 9]. Modern 3rd generation DSCTs are able to acquire images at a temporal resolutions of 66 ms, which allows for imaging at much higher heart rates and even in the presence of arrhythmias at a reasonable spatial resolution [8]. Alternatively, DSCT may be used to provide a high-pitch single-heartbeat acquisition to scan the whole heart within a single cardiac cycle (i.e. flash-mode). Although subject to certain limitations with regard to heart rate, body habitus, and coronary calcification, this scan mode can provide imaging at low radiation doses (see Table 1). [19]

Table 1 Effective radiation dose of invasive and CT-coronary angiography
Full size table

Another important advancement was the introduction of multi-slice CT to increase the number of slices that can be acquired during a single rotation [20]. In turn, the volume of interest can be scanned faster and scans of the whole body can be performed within 1–2 s. [21] Wide volume detectors allow a volumetric imaging in one heartbeat without time delay.

Spatial Resolution

The second major influencing factor of CCT is spatial resolution. As previously discussed, temporal resolution is vital for dose reduction, shorter scanning times, and functional analyses. Conversely, spatial resolution has a significant role in the primary application of CCT, which is morphological coronary imaging. This facet is essential in visualizing and quantifying intracoronary structures. Regular clinical CCT acquisition provides a resolution of approximately 0.25–0.5 mm while maintaining a satisfactory signal-to-noise ratio [18]. Initial utilisation of ultra-high resolution (UHR) CCT, with down to 150-200um resolution where made in a research environment [22]. UHR CT demonstrated higher positive and negative predictive values within calcified lesions and a more detailed quantification of in-stent dimension compared to regular CCT [23]. However, UHR CCT was associated with increased mean radiation doses (compare Table 1) [23, 24].

One method of increasing CT resolution is reducing the size of the detector elements (or increasing the slice number while maintaining detector width). However, increasing detector row numbers may be subject to a progressive deformation at the edge of the field of view, and increased scatter and loss of efficiency due to the relative increase in the area of the anti-scatter grid [21]. Additionally, higher gantry rotation speeds place multiple increased technical demands on CT detector design [1], [25].

Other methods to increase the resolution include rapid alternation of the X-ray tube electron beam focal point (‘flying focal spot’). This allows to obtain overlapping slices to increase the number of samples for a single projection and subsequently achieve sub-millimeter resolution [9, 26]. With the addition of advanced reconstruction methods it is possible to reconstruct thinner slices than acquired using standard retrospective helical acquisition techniques [27, 28]. As the increase in spatial resolution is generally traded for increased image noise, iterative reconstruction algorithms, and deep learning reconstruction methods have been developed to improve image quality and simultaneously reduce radiation [29,30,31].

Contrast Resolution

While the identification of small lesions and detailed structures is a significant feature of CCT, the differentiation of plaque subtypes and specific plaque features is also crucial. These distinctions are significant because they are associated with cardiovascular events [32]. The identification of tissue subtypes within those structures is determined by contrast resolution, which is the last technical aspect with high importance to be discussed.

Spectral CT is able to identify and quantify different materials by calculating x-ray photon energy information by subtraction of images at different energy fields as originally proposed by Hounsfield [33].

A more recent development are photon counting detectors (PCD) which were clinically approved by the FDA in Sep. 2021 [34]. The major difference to conventional energy-integrating detectors (EID) is that with EID, the total energy from a large number of photons is registered within one detector element. In comparison, a PCD registers the interaction of individual photons and approximates the energy spectrum of an individual pixel [35]. To subtract different energy levels in EID, either a physically resolved (e.g. DSCT) or temporally resolved scan is performed to measure different energy levels [36]. For PCD, the detector is able to sort incoming photons in predefined energy “bins” by individually set thresholds within one acquisition [35]. This increases spatial resolution and material contrast within single pixels during a single scan and reduces artifacts [37]. Consequently, PCD reduce radiation dose as well as the required dose of contrast agents [38].

Applications – Present and Future

After providing insights into technical advances of CT imaging, the next chapter is dedicated to a brief introduction of CCT applications in cardiovascular medicine including ongoing developments (Fig. 1).

Fig. 1
figure 1

© CC-BY 4.0 https://creativecommons.org/licenses/by/4.0/.)

Central Figure: Cardiac CT applications in guidelines and current research. Cardiac CT is widely recognized and recommended in current guidelines. Recent developments potentially enable a broader application in various diseases offering a fast and effective alternative to established diagnostic pathways. (Parts of the image have been adapted from Yamada et al. [39]

Full size image

Coronary Stenosis, Anomaly, and Ischemia – from CT-Angiography to CT-FFR and Perfusion

CCTA and Calcium Scoring

Isotropic resolution allows for reconstruction in any orientation to depict coronary artery anatomy, anomalies and stenosis with high detail in 3D visualizations (see Fig. 2). Furthermore, CCT can provide sophisticated plaque characterization based on levels of calcification and adverse geometric features of plaques [40]. The latest scanner generations, especially using PCD technology, even allow for an accurate description of coronary artery in-stent re-stenosis [41].

Fig. 2
figure 2

CT coronary angiography. The image on the left shows a 3D model of the coronary arteries. The right image displays a radial series of the left anterior descending artery of the same patient with calcified and non-calcified plaque in the proximal segment

Full size image

Next to coronary computed tomography angiography (cCTA) coronary artery calcium scoring has emerged as an easy-to-use non-invasive parameter for the atherosclerotic burden and adds incremental prognostic value and safety to regular cCTA [42]. Novel applications include virtual non-contrast spectral CT imaging techniques, that allow calcium scoring with very low radiation doses (see Table 1) [43] and the estimation of the calcium score from contrast cCTA without the need for an additional scan [44].

CT-FFR

The reference standard for quantifying the hemodynamic relevance of coronary narrowings are invasive measurements of either fractional flow reserve (FFR) [45], to calculate the hyperemic reserve of a vessel during pharmacologically induced vasodilation or alternatively during the wave-free period using the instantaneous wave-free ratio (IFR) [46, 47]. However, using CCT imaging, an equivalent to the invasive approach can be offered with high per-vessel and per-patient agreement in published head-to-head comparisons [48, 49].

CT-FFR measurements allow for calculation of blood flow over the stenosis using computational fluid dynamics or machine learning approaches (see Fig. 3) [50]. However, this technical approach is highly dependent on the image quality and segmentation of the coronary arteries.

Fig. 3
figure 3

Functional testing using CT-FFR. The image on the left shows an intermediate stenosis of the left anterior descending artery by a mixed plaque. Using CT-FFR as shown on the right, it was possible to exclude haemodynamic relevance of the stenosis

Full size image

CT-Perfusion

Guidelines recommend functional imaging options like stress-echocardiography, PET/SPECT, or CMR to detect wall-motion abnormalities or reduced perfusion in viable tissue [47]. CCT multidetector perfusion imaging during adenosine stress has emerged as a potential alternative for functional perfusion assessments [51,52,53]. However, in comparison to invasive FFR quantification, CT-Perfusion imaging revealed only a moderate accuracy to identify perfusion defects associated with ischemia. Notwithstanding, in combination with cCTA, CCT-Perfusion may allow to identify or exclude relevant ischemia within the myocardium subtended by a stenosed artery [54, 55, 56].

Cardiomyopathies – Volumes and Tissue from CMR to CT

Cardiac Volumes

Initial studies of volume quantification using CCT revealed overall good feasibility with varying but reasonable agreement with CMR as the reference standard. Mostly, a trend of an overestimation of volumes and underestimation of the ejection fraction (EF) was observed [57, 58]. However, with the development of multi-slice CT later studies could prove an increased accuracy of measurements of left ventricular EF compared to CMR and transthoracic echocardiography [59] and a concomitant assessment of right ventricular function [60]. An example of CT-based volumetric and functional measurements of the left ventricle is shown in Fig. 4.

Fig. 4
figure 4

CT-derived volumetry and functional analysis. The images on the left describe the process of CT-based volumetry using endo- and epicardial contouring. Functional assessments allow the quantification of global cardiac function as well as in-depth visualization of wall-motion abnormalities as shown by the bullet plot on the right

Full size image

More recently, even CT-derived strain measurements have been shown to be possible [61] and provided comparable measurements to CMR [62].

Tissue Characterization

In CCT, current developments address the assessment of myocardial edema and the extra cellular volume as an index of diffuse myocardial fibrosis.

Focal myocardial edema can be sufficiently assessed, as low Hounsfield units indicate water as the main component of the edema. Comparisons to T2-weighted CMR imaging using unenhanced DSCT imaging demonstrated good agreement between both methods [63].

For quantification of the extra cellular volume (ECV) in CCT, measurements are performed by the creation of a subtraction image from pre- and post-contrast acquisitions. Using newer generations of dual-energy PCTs, ECV can be derived from a single delayed acquisition with similar accuracy and lower radiation doses (see Fig. 5) [64]. Even though the determination of diffuse myocardial abnormalities is very challenging considering the low contrast resolution of CCT, validation of current methods against CMR showed overall good agreement [65].

Fig. 5
figure 5

© CC-BY 4.0 https://creativecommons.org/licenses/by/4.0/.)

Calculation of ECV using CT. Delayed enhancement images and polar ECV maps of a patient with normal ECV (a and c) and a patient with increased ECV (b and d). ECV can be calculated by a subtraction image of a pre-contrast and a delayed enhancement post-contrast image. (By Yamada et al. [39]

Full size image

Structural Heart Disease – from Valvular Pathologies to 4D Flow

CCT allows for spatially highly resolved imaging of the valves including morphological details such as the leaflets, and for the mitral valve, chordae, or papillary muscles. In contrast to echocardiography, valve assessment is not based on hemodynamical pressure and velocity gradients, but on the direct planimetry of the aortic orifice during mid systole.

The highly detailed visualization of the valves is an increasing desire for the calculation of hemodynamics and cardiovascular flow. CCT requires the use of a contrast bolus for calculations of cardiovascular bloodflow [66] or to obtain transvalvular pressure gradients to complement the detailed morphological imaging with functional and physiological data to enhance valvular assessments [67].

First implementations of algorithms using computational fluid dynamics were even able to analyze 4D intracardiac flow patterns [68, 69]. Even though those techniques are still in their infancy, calculation of 4D data in cardiovascular assessments has received considerable attention in general and is also likely to grow in CCT.

Clinical Perspective

Novel imaging capabilities of CCT have enforced a widespread use across various cardiac diseases. Considering the high performance of CCT regarding diagnosis, prognostication, and screening particularly of CAD and increasing guideline recommendations, we have detailed current and future clinical value in the following section.

Clinical Value in Prevention

Non-contrast coronary artery calcium scoring at the start of the scan offers accurate coronary screening and crucial predictive information [70, 71] with high therapeutic impact [72].

CCT is able to detect coronary artery plaque at an early stage [70]. Even though plaques might be non-obstructive, their prognostic importance has been well demonstrated [73]. In particular, in mild nonobstructive plaques with 0–49% stenosis, coronary calcification predicted mortality significantly better than conventional risk assessment by the Framingham risk score [74].

The Multi-Ethnic Study of Atherosclerosis has established percentiles for four major racial and ethnic groups, discovering that a doubling of coronary artery calcification elevates the risk of major coronary events from 15 to 35% percent [75]. Moreover, the total plaque burden was found to be a major risk predictor of cardiovascular events and death independently of the presence of stenosis. This led to an amendment in the revised CAD-RADS 2.0 criteria, now including the total plaque burden as a new category for risk assessment [76, 77].

From a therapeutical perspective, an early identification of coronary plaques can lead to an early start of statin therapy to stabilize plaques and reduce atherosclerotic risk [78]. Next to statins non-invasive coronary artery calcium scoring also governs the use of aspirin for primary prevention as patients with more severe coronary artery calcifications are more likely to benefit from statins and aspirin [79]. Patients with a risk for CAD greater than  10% as indicated by the Framingham risk score and a calcium score between 1–99 were shown to benefit from medication already, since the number needed to treat falls below the number needed to harm. In patients with a calcium score greater than 100,  this additional benefit of medical prevention was observed for all patients, irrespective of their Framingham risk score [80]. This impact of CT-based risk screening on primary preventive measures  has not been shown for any other modality, underlining the importance of CCT in preventive cardiovascular care.

Counterintuitively to the findings on medical therapy, a recent study found that lifelong endurance training was associated with higher plaque burden as detected by CCT despite the presumed healthier lifestyle of athletes [81]. Notwithstanding, the study found that the plaques athletes, may arise from wall stress during enduring exercise and are as stable as  plaques during statin therapy. A longitudinal follow-up of this study cohort would be required to further enhance our understanding of the importance of different forms of exercise in cardiovascular prevention after plaque detection by CCT.

Clinical Value in Coronary Artery Disease

CTA has experienced a major boost after consideration and valuation as a class I recommendation within the 2016 NICE guidelines, 2019 ESC guidelines, 2021 AHA/ACC guidelines for chest pain and chronic coronary syndrome [2, 47, 82], and the 2022 guidelines for ventricular arrhythmias and sudden cardiac death [5].

Due to high sensitivity, it is being recommended for patients with low- to intermediate-risk for chronic CAD to rule-out CAD due to its high negative predictive value [47, 83,84,85]. In turn, the lower specificity may require functional testing in patients with higher-risk constellations [47]. Nevertheless, when assessing its utility in routine clinical practice, CCT was validated as an important alternative to functional testing in the context of the PROMISE trial [86] and as an effective tool for additional clinical decision-making in the SCOT-HEART trial [87]. Non-inferiority of CCT as an initial diagnostic step compared to invasive coronary angiography has been shown by multiple studies [87,88,89].

Even in the very early stages of atherosclerosis, CCT yields capabilities for in-depth risk prediction using the fat attenuation index by quantifying perivascular adipose tissue. This may point to anti-inflammatory therapy interventions and allow for disease activity monitoring [90]. As such, it offers a cost-effective alternative in patients with stable chest pain symptoms [91] while helping to avoid unnecessary coronary angiographies [87, 92].

For clinical reporting the use of the Coronary Artery Disease – Reporting and Data System (CAD-RADS) is recommended by international societies [93]. In addition to a standardized framework for a comprehensive reporting on plaque burden, stenosis and specific modifiers, CAD-RADS provides dedicated recommendations on further diagnostic and therapeutic steps.

Other studies investigated CCT in acute coronary syndromes and found that safe discharge of patients with low- to intermediate-risk for acute coronary syndrome can be provided following CCT assessment. Strategies involving CCT in a triage system at the emergency department for patients with suspected acute coronary syndrome resulted in a more rapid and safe discharge with similar or lower total costs of treatment [94,95,96]. However, potential downstream testing and total radiation exposure increased, indicating the need for precise clinical assessments when using CCT for decision-making at the emergency department [96].

To address the lower specificity of CCT, further functional testing of hemodynamic relevance of stenoses have been introduced by CT perfusion or CT FFR within the same modality [97]. Testing for ischemia using CCT has now been included in the revised CAD-RADS II criteria as an additional modifier “I” and is suggested in patients with moderate to high (50–90%) stenosis i.e. CAD-RADS 3 and 4A when technically available [93].

Compared to non-invasive CMR-Perfusion, isolated CCT-Perfusion is inferior, even though its combination with anatomical cCTA increases accuracy and results in similar global performance [55]. While CT-FFR offers an alternative functional assessment without the need of additional radiation, it is important to note that high standardization is recommended according to SCCT and RSNA guidelines. [97, 98] This is mainly due to the rate of undiagnostic results that are still described to occur between 7–12% of scans, which may explain why this technique has not yet fully been adopted clinically [99, 100]. In addition, CT-FFR suffers from a “grey area” between 0.76–0.8 potentially requiring referral for invasive diagnostics [93]. Therefore, guidelines consider CT-FFR currently only as an amendment to regular cCTA [97, 98]. Bearing these limitations in mind, CT-FFR has been demonstrated to increase the specificity and positive predictive value of CCT compared to regular cCTA [101]. Furthermore, it has been suggested that CT-FFR could potentially reduce costs compared to invasive procedures, taking into account that pressure wires are single-use items and must be disposed of post-measurements. Interestingly, no total cost reduction could be found compared to standard clinical care, but CT-FFR was able to significantly reduce the number of invasive coronary angiography [102].

Clinical Value in Cardiomyopathies

For assessments of the impact of plaque burden and degree of coronary stenosis on the myocardium, the quantification of myocardial function and tissue viability is of equal significance. In CAD with a subsequent risk of heart failure and development of scar tissue, these assessments can guide medical and device therapy. Even though transthoracic echocardiography is the most widely used diagnostic tool for quantification of function and CMR remains the mainstay for tissue characterization, it would be of interest to obtain additional data on cardiac function and remodeling within the same scan and modality.

Even though volumetric assessments can be easily obtained from CCT scans comprising the whole cardiac cycle and volume, they still require a dedicated protocol, as a post hoc analysis from a scan solely focusing on coronary arteries is not possible yet. Furthermore, volumetric scans come with the trade-off of increased doses of radiation and contrast agents [103] which currently makes them less applicable in clinical routine. Several approaches have been made to overcome the need for contrast agent application and to retrospectively analyze data from non-contrast images, however, those techniques require further validation for broader applicability in the future [104,105,106].

The domain of tissue characterization in cardiac CT has been established more recently and offers promising prospects for future applicability. Even though the range of normal values of ECV is still a major impediment to clinical application [107], ECV plays an important role in patients with aortic stenosis [108]. Cardiac amyloidosis is characterized by a significantly increased ECV and co-present in 15–30% of aortic stenosis cases, which can theoretically be diagnosed from regular planning CTs for TAVR. In this matter, studies already found that the calculation of CT-derived ECV during routine TAVR evaluation is able to depict cardiac amyloidosis with high accuracy and provides information on the degree of infiltration [109].

While CMR will most likely remain superior to CCT for tissue characterization, CCT may be considered as an alternative in the presence of contraindications to CMR [110, 111]. Especially in patients with pacemakers, artifact compensation in CT is more robust compared to CMR, even though CMR imaging quality in these patient groups has improved in the past years.

Clinical Value in Valvular and Congenital Heart Disease

In clinical routine, CCT is mostly applied for sizing and selection of the optimal prosthesis as part of the preparation of invasive procedures, including the visualization of the access route for the aortic valve along the femoral arteries and descending aorta within the same examination (see Fig. 6) [112, 113, 114].

Fig. 6
figure 6

Transcatheter aortic valve replacement (TAVR) planning using CT. The images on the left delineate the measurement of the aortic annulus prior to TAVR in 3 different orientations of the aortic valve. The right image shows the access path via the right (A. fem. r) and left femoral artery (A. fem. l)

Full size image

In addition, CCT was recommended by current ESC guidelines as flow-independent method of choice for advanced decision-making in patients with borderline findings in aortic stenosis using the assessment of aortic valve calcification [112, 115]. The distribution of calcifications along the aortic annulus can be visualized using CCT which helps to identify adverse features indicating a more complicated intervention, periprocedural risks [116, 117] or even an increased risk for postprocedural LVOT obstruction [118, 119]. Those prospects for an enhanced risk prediction extended the value of CCT beyond conventional procedural planning [120].

Furthermore, CCT can be used to determine reversible causes of valvular disease such as thrombosis or endocarditis if transthoracic and transesophageal echocardiography is inconclusive [112] and was now endorsed by the 2023 ESC guidelines more extensively [121].

The utility of dynamic CCT in the setting of infective endocarditis has been shown to be particularly important if cardiac devices are present such as in the setting of right ventricle to pulmonary artery conduits, which are in an anterior location that is not easily imaged on trans-esophageal echocardiography [122].

Similarly to valvular diseases in adults, CCT has become the modality of choice for planning valvular device suitability in CHD for the new generation of percutaneously implanted large RVOT pulmonary valve systems [123]. Fig. 7 shows the ability of CCT not only to predict device choice and to define its landing zone but also the ability to provide the estimated angiographic projection.

Fig. 7
figure 7

Shows the estimated optimal angiographic projection using the CT images on the right. The estimated distal landing zone shown in green. On the left panel, the actual angiographic deployment is shown taking benefit from CT-based predictions

Full size image

But also considering other CHD such as coronary anomalies, it has also been shown that cCCT is able to reproducibly measure the proximal dimension of anomalous aortic origin of a coronary artery and so, may be helpful in risk prediction [124]. Another important advantage of CCT in CHD is the consistent threshold characteristics which are particularly advantageous for volume rendering and 3D printing of CHD. This approach has been shown to be able to depict complex anatomy for improved interventional and surgical planning [125].

Conclusion

CCT has become an irreplaceable instrument in contemporary medicine. Its capabilities in anatomical imaging supply crucial information in an array of heart diseases, while additional functional assessments significantly broaden the range of CCT applications. Some of these techniques still need to establish their additional value in clinical contexts and require further validation in large, multicentric trials. If the hardware and software refinements keep up to speed, CCT possesses significant potential to become a "one-stop-shop" non-invasive imaging method, not only for CAD but also for other diagnostic applications.