Cardiac imaging encompasses a range of noninvasive and invasive medical techniques designed to visualize the heart's anatomy, function, and vascular supply, enabling the diagnosis, riskstratification, and management of cardiovascular diseases such as coronary artery disease, valvular disorders, cardiomyopathies, and congenital anomalies.[1]These techniques have evolved significantly since the late 20th century, with echocardiography emerging as the most widely used modality due to its portability, lack of ionizing radiation, and ability to provide real-time assessment of cardiac motion, chamber sizes, and blood flow via Doppler ultrasound.[2] Computed tomography (CT) angiography offers high-resolution three-dimensional images of the coronary arteries and calcium scoring for atherosclerosis detection, while cardiac magnetic resonance imaging (MRI) excels in tissue characterization, perfusion evaluation, and viability assessment without radiation exposure.[2] Nuclear imaging, including positron emission tomography (PET) and single-photon emission computed tomography (SPECT), quantifies myocardial perfusion and metabolism to identify ischemia or infarction, and invasive coronary angiography remains the definitive method for delineating coronary anatomy and guiding interventions like stenting.[1]In recent years, particularly from 2020 onward, innovations such as artificial intelligence (AI) for automated image analysis, photon-counting CT for reduced radiation doses below 1 mSv, and 4D flow MRI for comprehensive hemodynamic mapping[3] have enhanced diagnostic precision, minimized procedural risks, and supported personalized treatment strategies in precision cardiology.[4]Multimodality approaches, integrating data from multiple techniques, further improve outcomes by facilitating early detection and serial monitoring, reducing the need for invasive procedures, and aiding in prognostic assessment for conditions like heart failure and atrial fibrillation.[2]
Introduction
Definition and Scope
Cardiac imaging encompasses a range of non-invasive and invasive techniques designed to visualize the anatomy, function, blood flow, and tissue characteristics of the heart and surrounding vasculature, primarily for diagnostic purposes, treatment planning, and ongoing monitoring of cardiovascular conditions.[1] These methods play a pivotal role in modern cardiology by enabling clinicians to assess structural integrity, myocardial perfusion, and hemodynamic performance without always requiring surgical intervention.[5] The scope extends to both static and dynamic evaluations, supporting everything from initial screening to post-procedural follow-up, while balancing diagnostic yield against procedural risks such as radiation exposure or contrast-related complications.[6]At its core, cardiac imaging relies on physical principles involving sound waves, magnetic fields, X-rays, or radioactive tracers to generate interpretable images of cardiac structures.[1] Techniques like ultrasound provide real-time imaging at frame rates up to 100 per second, ideal for capturing motion and function, whereas modalities such as computed tomography (CT) or magnetic resonance imaging (MRI) often produce higher-resolution static or three-dimensional reconstructions better suited for detailed anatomy.[1] A key trade-off in these approaches is between spatial resolution, which enhances tissue differentiation, and temporal resolution, which is crucial for assessing rapid cardiac cycles; invasive methods may improve resolution but introduce additional risks.[5]Broadly, cardiac imaging modalities are categorized into ultrasound-based echocardiography, which uses sound waves for non-invasive assessment; MRI, leveraging magnetic fields for superior soft-tissue contrast; CT, employing X-rays for rapid vascular imaging; nuclearscintigraphy, involving radioactive tracers to evaluate perfusion; and invasive cardiac catheterization, which provides direct angiographic views via fluoroscopy.[1]Radiation exposure is a consideration primarily with CT and nuclear techniques, while contrast agents—used across CT, MRI, and catheterization—facilitate enhanced visualization but carry risks like nephrotoxicity in vulnerable patients.[1] These categories allow for tailored applications, with non-invasive options preferred for routine use and invasive ones reserved for confirmatory or therapeutic roles.[6]In cardiology, cardiac imaging is indispensable for detecting congenital heart defects, ischemic conditions, and cardiomyopathies, offering quantitative insights into myocardial viability and blood flow that inform risk stratification and therapeutic decisions.[5] For instance, it guides interventions such as valve repairs or revascularization by providing precise anatomical and functional data, ultimately improving patient outcomes through evidence-based management.[6] This foundational role underscores its integration into preventive and acute care strategies across diverse cardiovascular pathologies.[1]
Historical Overview
The early foundations of cardiac imaging were laid in the 1920s through 1950s with plain film radiography and fluoroscopy, which provided basic assessments of heart size, shape, and gross abnormalities such as cardiomegaly or pericardial effusions.[7] These techniques, building on Wilhelm Röntgen's 1895 discovery of X-rays, allowed for static and dynamic visualization of the cardiac silhouette but were limited by overlapping structures and lack of internal detail.[8] By the 1970s, the introduction of ECG-gated imaging marked a significant advancement, synchronizing image acquisition with the cardiac cycle to reduce motion artifacts and enable more precise evaluation of ventricular function, initially in nuclear and early CT applications.[9]The ultrasound era began in the 1950s with the inception of echocardiography, pioneered by Inge Edler and Hellmuth Hertz in 1953, who used an industrial ultrasonic flaw detector to record the first M-mode echoes from the heart, demonstrating mitral valve motion.[10] This noninvasive method revolutionized cardiac assessment by providing real-time motion data. By the 1970s, M-mode evolved into routine clinical use, and two-dimensional (2D) echocardiography emerged, with key developments including real-time linear array imaging by Nico Bom and colleagues in Rotterdam, allowing cross-sectional views of cardiac anatomy.[11] The 1980s saw the advent of Doppler echocardiography, integrating pulsed and color flow mapping to quantify blood velocities and detect valvular regurgitation or stenosis, further enhancing hemodynamic evaluation.[12]Advanced modalities expanded in the late 20th century: nuclear imaging originated in the 1970s with early perfusion tracers like thallium-201 for planar myocardial imaging, progressing to SPECT in the 1980s for myocardial viability assessment.[13] Computed tomography (CT) prototypes in the 1970s laid groundwork for cardiac applications, with electron beam CT in the 1980s and helical multislice CT in the 1990s enabling coronary calcium scoring via the Agatston method introduced in 1990.[14] Cardiac MRI debuted in the 1970s with initial spectroscopy for metabolic assessment, achieving clinical imaging by the 1980s and advancing to tissue characterization in the 1990s through contrast-enhanced techniques for fibrosis and viability.[15] Standardization of coronary CT angiography occurred in the 2000s, driven by 64-slice scanners and Society of Cardiovascular Computed Tomography guidelines in 2009, improving noninvasive coronary evaluation.[16]Recent milestones include the 2010s integration of 3D and 4D imaging across modalities, such as real-time 3D echocardiography for volumetric assessment and hybrid PET/CT systems introduced commercially in 2001 for combined perfusion and anatomic evaluation.[12] Post-2020 developments feature AI-assisted interpretation to automate segmentation and risk stratification in echocardiography and MRI, alongside low-radiation CT protocols using deep learning reconstruction to achieve doses below 2 mSv while preserving diagnostic quality.[17][18]
Clinical Applications
Indications for Use
Cardiac imaging is indicated for the diagnostic evaluation of patients presenting with chest pain suggestive of acute coronary syndrome, where modalities such as coronary CT angiography (CCTA) are recommended as a first-line test in stable patients without known coronary artery disease to assess for obstructive lesions.[19] For dyspnea, imaging is employed to investigate potential heart failure or valvular disease, with transthoracic echocardiography (TTE) serving as the initial modality to evaluate left ventricular function, chamber sizes, and valve abnormalities, as per Class I recommendations in heart failure guidelines. In cases of syncope, cardiac imaging helps identify structural issues or arrhythmias, such as aortic stenosis or hypertrophic cardiomyopathy, with echocardiography recommended for initial assessment in patients with exertional syncope or associated cardiac symptoms.Therapeutic planning often involves cardiac imaging for pre-procedural assessment in interventions like transcatheter aortic valve replacement (TAVR), percutaneous coronary intervention (PCI), or device implantation, where multimodality imaging including CT and MRI provides detailed anatomical and functional data to guide device selection and placement.[20] Post-myocardial infarction (MI), viability testing using cardiac magnetic resonance (CMR) or positron emission tomography (PET) is indicated to determine the presence of hibernating myocardium, informing decisions on revascularization in patients with reduced ejection fraction.[21]Prognostic applications of cardiac imaging include risk stratification in asymptomatic patients at intermediate risk for atherosclerosis, such as through coronary artery calcium (CAC) scoring via non-contrast CT, which helps predict future cardiovascular events and guide preventive therapy.[22] Serial imaging is recommended for monitoring progression in cardiomyopathies or congenital heart disease, with CMR providing reproducible measures of ventricular function and fibrosis to assess disease trajectory and response to therapy.[23]In population-specific contexts, cardiac imaging is crucial for pediatrics in evaluating congenital anomalies, where echocardiography is the primary modality for initial diagnosis and follow-up of structural defects like ventricular septal defects, with advanced imaging such as CMR reserved for complex cases.[24] During pregnancy, safe modalities like echocardiography are preferred for assessing suspected cardiac conditions such as peripartum cardiomyopathy or valvular disease, avoiding ionizing radiation while providing essential diagnostic information.[25] For athletes, screening with echocardiography is indicated in those with symptoms or family history suggestive of hypertrophic cardiomyopathy to detect left ventricular hypertrophy and guide participation eligibility.[26] These indications align with ACC/AHA guidelines emphasizing appropriate use criteria to optimize diagnostic yield and patient outcomes across diverse clinical scenarios.[19]
Contraindications and Safety Considerations
General contraindications for cardiac imaging include pregnancy, particularly for modalities involving ionizing radiation, to avoid potential fetal harm from radiation exposure. Renal impairment is a key concern due to the heightened risk of contrast-induced nephropathy when using iodinated or gadolinium-based contrast agents. Claustrophobia represents a relative contraindication for magnetic resonance imaging (MRI), as it may prevent patients from tolerating the enclosed scanner environment.[1][1][27]Safety profiles across cardiac imaging emphasize minimizing risks associated with radiation, contrast agents, and procedural factors. The ALARA (as low as reasonably achievable) principle guides radiation use in ionizing modalities like computed tomography (CT) and nuclear imaging, aiming to limit cumulative effective doses in line with the ALARA principle and ICRP public exposure guidelines of 1 mSv per year on average, though medical patient exposures are justified case-by-case without strict lifetime limits, to reduce stochastic effects like cancer induction. Gadolinium-based contrast agents carry a risk of nephrogenic systemic fibrosis (NSF) in patients with severe chronic kidney disease (stage 4 or 5, eGFR <30 mL/min/1.73 m²), though this risk is minimal (<0.07%) with group II agents and has decreased with modern formulations and screening. Allergic-like reactions to iodinated contrast occur in less than 1% of administrations, with severe anaphylaxis in approximately 0.04%, necessitating premedication protocols like corticosteroids and antihistamines for at-risk patients.[28][29][30]Patient preparation strategies are essential to enhance safety and image quality. Oral or intravenous beta-blockers may be administered to achieve a target heart rate below 60-65 bpm for optimal imaging in select cases, with continuation of chronic therapy unless contraindicated. Informed consent is mandatory for all procedures, particularly invasive or contrast-based ones, ensuring patients understand potential risks such as allergic reactions, renal complications, or rare procedural adverse events.[31][32][33]Special populations require tailored approaches to mitigate unique risks. In patients with cardiac implantable electronic devices like pacemakers or defibrillators, MRI is feasible only with MR-conditional devices implanted after the early 2010s, under strict protocols including device interrogation and reprogramming to avoid heating, torque, or malfunction. Pediatric patients often necessitate sedation or general anesthesia for non-cooperative imaging, with continuous monitoring of vital signs to ensure safety, as complications like respiratory depression are rare but possible in this group.[34][35][36]Risk mitigation involves proactive techniques to reduce exposure and monitor for complications. Dose optimization in radiation-based imaging employs methods like prospective electrocardiogram (ECG) gating, which limits X-ray exposure to specific cardiac phases, potentially reducing effective doses by 70-80% compared to retrospective gating while maintaining diagnostic quality. During stress testing integrated with imaging, continuous ECG monitoring is critical to detect and manage arrhythmias, with immediate access to resuscitation equipment to address rare events like ventricular tachycardia. These strategies, supported by multidisciplinary protocols, prioritize patient safety without compromising clinical utility.[37][38][39]
Echocardiography
Transthoracic Echocardiography (TTE)
Transthoracic echocardiography (TTE) is a non-invasive ultrasound-based imaging modality that serves as the primary initial diagnostic tool for evaluating cardiac structure, function, and hemodynamics through the chest wall.[40] It employs a phased-array transducer typically operating at frequencies of 2.0–5.0 MHz to generate real-time images, making it suitable for bedside use in various clinical settings.[41]The technique involves strategic probe placement to obtain standard acoustic windows for comprehensive visualization. In the parasternal window, the probe is positioned along the left sternal border with the index marker oriented toward the right shoulder for the long-axis view (PLAX) or left shoulder for the short-axis view (PSAX), allowing assessment of left ventricular (LV) dimensions, mitral valve, and aortic root.[42] The apical window is accessed by placing the probe at the apex, below the left breast, with the index at 4–5 o'clock for the four-chamber view (A4C), which can be rotated to yield two-chamber (A2C) or three-chamber views for evaluating chamber volumes and septal motion.[42] The subcostal window requires the patient to be supine with the probe below the xiphoid, index to the left for a four-chamber view, ideal for right ventricular and inferior vena cava assessment.[42] The suprasternal window positions the probe above the sternal notch, index to the left shoulder, for imaging the aortic arch and pulmonary veins.[42]TTE utilizes multiple imaging modes to assess both anatomy and physiology. Two-dimensional (2D) echocardiography provides grayscale cross-sectional images for structural evaluation, often enhanced by harmonic imaging to improve endocardial border definition.[42] M-mode echocardiography offers high temporal resolution (up to 1000 frames per second) for precise motion analysis of walls and valves, with sweep speeds of 100–200 mm/sec.[42] Color Doppler overlays velocity information on 2D images to detect turbulent flow, regurgitation, or shunts, with region-of-interest optimization and velocity scales of 50–70 cm/sec to minimize aliasing.[42]Quantitative assessments in TTE include measurements of chamber dimensions and function using standardized protocols. LV dimensions, such as end-diastolic internal diameter (LVIDd), are measured in PLAX perpendicular to the long axis below the mitral valve leaflets at end-diastole.[42] Left atrial dimensions are obtained via anterior-posterior diameter at end-systole using the leading-edge method.[42] Ejection fraction (EF), a key indicator of LV systolic function, is commonly calculated using the biplane Simpson's disk summation method from A4C and A2C views by tracing endocardial borders at end-diastole and end-systole to derive volumes:\text{EF} = \frac{\text{EDV} - \text{ESV}}{\text{EDV}} \times 100where EDV is end-diastolic volume and ESV is end-systolic volume.[42] Wall motion is scored segmentally in a 16- or 17-segment model across PSAX and apical views to identify regional abnormalities, such as hypokinesis in ischemia.[41]TTE's strengths include its portability for point-of-care use, absence of ionizing radiation, and ability to provide real-time dynamic imaging without contrast in most cases.[41] However, limitations arise from suboptimal acoustic windows in patients with obesity, emphysema, or chest wall deformities, which can obscure images and reduce diagnostic accuracy.[41]Clinically, TTE is essential for initial screening in suspected cardiac pathology, such as evaluating heart murmurs for valvular disease, detecting pericardial effusions in acute chest pain, and assessing LV function in heart failure or post-myocardial infarction.[43][44][40]
Transesophageal Echocardiography (TEE)
Transesophageal echocardiography (TEE) is a semi-invasive imaging modality that utilizes ultrasound waves transmitted from a probe positioned in the esophagus to produce detailed, high-resolution images of the heart, with particular emphasis on posterior structures such as the left atrium, mitral and tricuspid valves, and thoracic aorta.[45][46] The technique leverages the esophagus's proximity to the heart to minimize acoustic interference from the chest wall, lung tissue, or ribs, enabling clearer visualization than surface-based methods.[47]The TEE procedure involves inserting a flexible, transducer-equipped probe through the mouth and into the esophagus under moderate sedation and topical anesthesia to ensure patient comfort and minimize gag reflex.[45][46] The probe's multiplane capability allows the imaging plane to rotate electronically from 0° to 180°, facilitating comprehensive cross-sectional views of targeted cardiac regions, including the atrial septum for septal defects, valvular apparatus for regurgitation or stenosis, and proximal aorta for dissection assessment.[48][49] The examination typically lasts 15-90 minutes, during which 2D, 3D, and Doppler modes are employed to evaluate structure and function in real time.[45]TEE plays a critical role in intraoperative monitoring during cardiac surgeries, where it provides dynamic guidance for procedures such as valve repairs or closures by assessing anatomy pre- and post-intervention.[46] It is particularly valuable for detecting left atrial thrombi prior to interventions like cardioversion and for identifying vegetations in infective endocarditis, offering superior sensitivity for these pathologies compared to noninvasive alternatives.[45][46] Relative to transthoracic echocardiography (TTE), which often serves as the initial screening tool, TEE delivers enhanced resolution for subtle abnormalities like mitral valve prolapse or aortic dissection due to its direct esophageal access.[47][50]For quantitative evaluation, especially of prosthetic valves, TEE incorporates Doppler interrogation to derive transvalvular pressure gradients using the simplified Bernoulli equation:\Delta P = 4v^2where \Delta P represents the mean pressure gradient in mmHg and v is the peak blood flow velocity in m/s across the valve.[51] This approach aids in diagnosing obstruction or regurgitation by providing precise hemodynamic data.[51] While TEE is generally safe, rare complications include esophageal perforation, with an incidence of 0.03%-0.09%.[52]
Contrast and Stress Echocardiography
Contrast echocardiography enhances the diagnostic capabilities of standard transthoracic or transesophageal echocardiography by utilizing microbubble contrast agents, such as Definity (perflutren lipid microspheres), to improve visualization of cardiac structures. These gas-filled microbubbles, typically 1-4 micrometers in diameter, oscillate in response to ultrasound waves, producing strong harmonic signals that opacify the left ventricular cavity and delineate endocardial borders, particularly in patients with poor acoustic windows due to obesity or lung disease.[53] This technique is also applied for myocardial perfusion imaging, where low mechanical index ultrasound allows assessment of blood flow in the myocardium by observing microbubble replenishment after high-intensity destruction pulses.[54]The use of contrast agents significantly improves the accuracy and reproducibility of left ventricular ejection fraction (LVEF) measurements compared to non-contrast echocardiography, reducing inter-observer variability and enhancing correlation with reference standards like cardiac MRI.[55] In clinical practice, contrast opacification is recommended when two or more contiguous segments are not well visualized at baseline, leading to more reliable volumetric assessments and better detection of regional wall motion abnormalities.[56]Stress echocardiography builds on these foundational platforms by inducing physiological or pharmacological stress to unmask inducible ischemia through changes in myocardial contractility. Exercise stress, typically via treadmill or bicycle protocols, or pharmacological agents like dobutamine (up to 40 mcg/kg/min) combined with atropine, increase myocardial oxygen demand to provoke wall motion abnormalities in territories supplied by stenotic coronary arteries.[57] Images are acquired at baseline and peak stress, with analysis focused on new or worsening hypokinesis, akinesis, or dyskinesis in a 17-segment model of the left ventricle.[58]Wall motion is scored using a 0-4 index per segment—0 for normal, 1 for mild hypokinesis, 2 for severe hypokinesis, 3 for akinesis, and 4 for dyskinesis—with the wall motion score index (WMSI) calculated as the average score across all segments to quantify the extent and severity of ischemia (e.g., WMSI >1.0 indicates abnormality).[59] This semi-quantitative approach has high sensitivity (80-90%) and specificity (80-85%) for detecting coronary artery disease (CAD) when compared to invasive angiography.[60]In applications for CAD, contrast-enhanced stress echocardiography excels in viability assessment by evaluating contractile reserve during low-dose dobutamine infusion, where improvement in wall motion suggests hibernating myocardium amenable to revascularization.[61] It also detects exercise-induced mitral regurgitation, a dynamic valvular dysfunction that may only manifest under stress and contribute to symptoms in patients with ischemic or functional mitral disease.[62]Limitations of contrast echocardiography include artifacts from microbubble destruction, which can cause transient apical swirling or shadowing if high mechanical index imaging is used inappropriately, potentially obscuring subendocardial perfusion defects.[63] Stress echocardiography is contraindicated in unstable angina or recent myocardial infarction due to the risk of provoking arrhythmias or infarction, and contrast agents carry rare risks of hypersensitivity or pulmonary effects in patients with severe lung disease.[64] According to appropriate use criteria, stress echocardiography is rated as rarely appropriate for asymptomatic low-risk patients, reserving it for intermediate-risk individuals with symptoms or prior equivocal tests to optimize diagnostic yield and avoid unnecessary radiation exposure compared to nuclear alternatives like SPECT.[65]
Cardiac Magnetic Resonance Imaging (MRI)
Structural and Functional Assessment
Structural and functional assessment in cardiac magnetic resonance imaging (MRI) primarily relies on non-contrast sequences to evaluate heart anatomy and motion, providing detailed insights into myocardial morphology and ventricular performance without the need for gadolinium-based agents. These techniques are essential for diagnosing and monitoring various cardiac conditions, offering superior visualization compared to other modalities in certain scenarios.[66]Anatomical imaging begins with black-blood sequences, such as double inversion recovery, which suppress the signal from flowing blood to highlight myocardial walls and surrounding structures. This approach excels in measuring wall thickness, identifying congenital defects like septal abnormalities, and delineating tissue boundaries with high resolution. In parallel, bright-blood sequences, often integrated into cine acquisitions, enhance visualization of vascular structures and intracardiac anatomy by maintaining high blood signal intensity, aiding in the assessment of chamber sizes and great vessel patency.[67][68][69]Functional assessment utilizes cine MRI sequences based on balanced steady-state free precession (bSSFP), which provide dynamic imaging of cardiac motion across the cardiac cycle with excellent blood-myocardium contrast. These sequences enable precise quantification of ventricular volumes, including end-diastolic and end-systolic volumes, through short-axis stack acquisitions. Ejection fraction (EF) is calculated using an adaptation of Simpson's method, summing epicardial and endocardial contours across 3D slices to derive stroke volume and EF, offering reproducibility superior to echocardiography, particularly for the right ventricle.[70][71][72]In clinical applications, these non-contrast techniques are particularly valuable for evaluating non-ischemic cardiomyopathies, where patterns of hypertrophy—such as asymmetric septal thickening in hypertrophic cardiomyopathy—can be characterized in three dimensions to guide diagnosis and risk stratification. Similarly, assessment of right ventricular function is critical in pulmonary hypertension, where cine bSSFP sequences quantify RV volumes, EF, and remodeling, correlating with disease severity and prognosis. Transthoracic echocardiography often serves as an initial screening tool, but cardiac MRI provides more accurate volumetric data when acoustic windows are suboptimal.[73][74]The advantages of these cardiac MRI approaches include the absence of ionizing radiation, making them suitable for serial imaging in younger patients or those requiring repeated evaluations, and exceptional soft-tissue contrast that delineates subtle anatomical variations. Typical scan times range from 30 to 45 minutes, incorporating breath-hold acquisitions to minimize motion artifacts and ensure high-quality images. Perfusion imaging can be added post-structural assessment if ischemia is suspected, but it requires contrast.[75][66][76]
Perfusion and Viability Imaging
Perfusion imaging in cardiac magnetic resonance (CMR) evaluates myocardial blood flow (MBF) by tracking the first-pass transit of an intravenous gadolinium-based contrast agent through the myocardium, typically using saturation recovery pulse sequences to achieve T1-weighted imaging.[77] These sequences, such as saturation recovery turboFLASH or TrueFISP, saturate myocardial spins prior to imaging to enhance contrast during the bolus arrival, enabling semiquantitative assessment via signal intensity curves or fully quantitative MBF measurement in ml/min/g through deconvolution of arterial input and tissue enhancement curves.[78] Quantitative approaches model the contrast uptake kinetics to derive rest and stress MBF values, with stress induced by vasodilators like adenosine to detect ischemia as regional hypoperfusion.[79]Viability assessment in CMR relies on late gadolinium enhancement (LGE) imaging, performed 10-20 minutes after contrast administration, where hyperenhancement on inversion-recovery sequences identifies scarred or fibrotic myocardium due to gadolinium accumulation in expanded extracellular spaces.[80] A transmural extent of hyperenhancement exceeding 50% in a myocardial segment strongly predicts non-viability and lack of functional recovery, while lesser extents suggest hibernating myocardium amenable to revascularization.[81] Complementary techniques include native T1 and T2 mapping, which quantify relaxation times to detect edema or inflammation; elevated T2 values indicate acute injury like myocardial edema, whereas T1 mapping highlights chronic changes such as diffuse fibrosis.[82]In clinical applications, perfusion and viability imaging guide post-myocardial infarction (MI) prognosis, where preserved viable myocardium on LGE correlates with improved left ventricular functional recovery and reduced adverse remodeling after reperfusion therapy.[80] These methods also detect microvascular disease, manifesting as microvascular obstruction on first-pass perfusion or persistent hypoenhancement within infarct zones on LGE, which portends worse outcomes including heart failure progression.[83] A study presented at the American Heart Association’s Scientific Sessions in November 2025 found that using stress CMR to guidediagnosis and treatment in patients with chest pain and no obstructive coronary artery disease resulted in more accurate identification of microvascular angina (affecting about 50% of cases, mostly women) and significant improvements in quality-of-life scores compared to angiography alone.[84]Limitations include susceptibility to motion artifacts from irregular heart rhythms or poor breath-holding, which can degrade image quality and MBF quantification, as well as contraindications in patients with non-MRI-conditional implants due to magnetic field interactions.[77] Emerging quantitative models, such as the Fermi deconvolution function for arterial input correction, aim to improve MBF accuracy by accounting for bolus dispersion but remain conceptual in routine practice pending broader validation.[85]
Computed Tomography (CT)
Coronary CT Angiography (CCTA)
Coronary CT angiography (CCTA) is a non-invasive imaging technique that utilizes electrocardiogram (ECG)-gated multidetector computed tomography (MDCT) to visualize the coronary arteries and detect coronary artery disease (CAD). The procedure involves intravenous administration of 50-120 mL of iodinated contrast material, injected at a rate of 5-7 mL/s, to achieve adequate opacification of the coronary lumen, typically reaching approximately 300 Hounsfield units. ECG gating, either prospective (triggered at 75% of the R-R interval in end-diastole) or retrospective (continuous acquisition with 10% phase binning), synchronizes image acquisition with the cardiac cycle to minimize motion artifacts from heart beating, enabling high-fidelity depiction of the coronaries. Modern scanners, such as 64-slice or higher MDCT systems, provide isotropic spatial resolution of approximately 0.5 mm and temporal resolution up to 75 ms, allowing comprehensive evaluation of the entire coronary tree in a single breath-hold of 5-10 seconds.[86][87][88]Interpretation of CCTA focuses on assessing luminal narrowing and plaque morphology to guide clinical decision-making. Stenosis is graded quantitatively as a percentage of diameter reduction compared to a reference segment, with >50% luminal narrowing indicating moderate stenosis that may warrant further evaluation, >70% signifying severe stenosis, and 100% denoting occlusion; recent meta-analyses report sensitivity of 91-95% and specificity of 88-90% for detecting >50% stenosis.[88][86][87][89] Plaque characterization distinguishes between calcified (high-attenuation >130 HU), non-calcified (soft, low-attenuation 30-130 HU), and mixed plaques, which provide insights into plaque vulnerability—such as low-attenuation plaques associated with acute coronary syndromes. These features are evaluated segmentally using standardized models like the 17-segment American Heart Association classification, often with multiplanar reformations and volume-rendered images for optimal visualization.[88][86][87]CCTA is particularly valuable in evaluating patients with low-to-intermediate pretest probability of CAD presenting with chest pain, serving as a first-line diagnostic tool to rule out obstructive disease in stable settings. In emergency departments, it efficiently excludes significant CAD in low-risk acute chest pain patients, reducing unnecessary admissions with a negative predictive value exceeding 95% for major adverse cardiac events over 1-2 years. Its high negative predictive value (up to 99%) makes it ideal for confirming the absence of stenosis, thereby avoiding invasive angiography in appropriate candidates.[90][87][86]Recent advances in CCTA include dual-energy CT, which differentiates plaque composition by exploiting material-specific attenuation differences, improving detection of non-calcified plaques and reducing blooming artifacts from stents. Radiation exposure has been substantially lowered with iterative reconstruction algorithms and prospective gating, achieving effective doses of 3-5 mSv on modern scanners—comparable to or lower than many nuclear stress tests—while maintaining diagnostic accuracy. Emerging photon-counting detector CT (PCCT), with clinical implementation as of 2021 and studies through 2025, further enhances performance by providing ultrahigh spatial resolution (down to 0.2 mm isotropic), spectral imaging for advanced plaque characterization, reduced metal artifacts, and potential doses below 1 mSv, improving diagnostic precision in coronary assessment.[87][86][91][92] These developments enhance CCTA's role as a safer, more precise modality for CAD assessment.
Calcium Scoring and Gated Functional CT
Calcium scoring is a non-contrast, electrocardiogram (ECG)-gated computed tomography (CT) technique used to quantify coronary artery calcification as a marker of atherosclerosis burden and cardiovascular risk.[93] The scan is performed prospectively at end-systole or mid-diastole to minimize motion artifacts, typically covering the heart in a single breath-hold with low radiation dose.[94]The Agatston score, the most widely adopted method, is calculated semi-automatically by identifying calcified lesions above a threshold of 130 Hounsfield units (HU), multiplying the lesion area by a density factor based on peak attenuation, and summing scores across coronary arteries.[95]Density factors are assigned as 1 for 130-199 HU, 2 for 200-299 HU, 3 for 300-399 HU, and 4 for ≥400 HU, providing a weighted measure that emphasizes denser plaques.[96] Risk stratification categorizes scores as low (0-100), moderate (101-400), and high (>400), with scores >100 indicating elevated 10-year coronary event risk independent of traditional factors.[97]Gated functional CT extends beyond plaque quantification to assess cardiac performance using the same ECG-synchronized acquisition, enabling retrospective reconstruction of cine images across the cardiac cycle for volumetric analysis. This yields left ventricular end-diastolic and end-systolic volumes, stroke volume, and ejection fraction, comparable to echocardiography in accuracy for routine evaluation.[98] Dual-source CT scanners improve temporal resolution to ~66 ms, allowing reliable functional assessment even at heart rates up to 100 bpm by reducing motion artifacts.[99]These techniques are applied in asymptomatic risk stratification, particularly for intermediate-risk adults aged 40-75 years per ACC/AHA guidelines, and selectively in those over 75 to guide statintherapy initiation.[100] In arrhythmia evaluation, gated CT supports ventricular function assessment despite irregular rhythms, offering advantages over non-gated methods through high temporal resolution.[101]Limitations include potential overestimation of stenosis severity in heavily calcified vessels due to blooming artifacts, which can complicate integrated lumen evaluation.[102] Radiation dose, typically 1-3 mSv, is optimized using prospective triggering and iterative reconstruction algorithms like adaptive statistical iterative reconstruction, which maintain score accuracy while reducing exposure by up to 80%.[103]
Single-photon emission computed tomography (SPECT) is a cornerstone nuclear medicine modality for myocardial perfusion imaging (MPI) in evaluating coronary artery disease (CAD), enabling the detection of regional blood flow abnormalities through the uptake of radiotracers proportional to myocardial perfusion. By comparing images acquired during stress and rest conditions, SPECT distinguishes between inducible ischemia and fixed perfusion defects due to infarction or scar. While traditional SPECT provides semi-quantitative assessment, newer cadmium-zinc-telluride (CZT)-based systems enable quantitative myocardial blood flow (MBF) measurement, approaching positron emission tomography (PET)'s capabilities, though PET generally offers higher spatial resolution. This technique's accessibility makes it a first-line tool for noninvasive CAD assessment, though it is limited by lower spatial resolution compared to PET (12–15 mm for conventional SPECT vs. 4–6 mm for PET), which provides quantitative flow but requires shorter-half-life tracers less widely available.[31][104]The most commonly used radiotracers in SPECT MPI are technetium-99m (99mTc)-sestamibi and 99mTc-tetrofosmin, which are lipophilic cationic complexes injected separately for stress and rest studies to evaluate perfusion differences. These 99mTc agents offer optimal dosimetry, allowing for higher administered activities and better image quality due to their 6-hour physical half-life and myocardial retention tied to mitochondrial integrity. For viability assessment, thallium-201 (201Tl) is preferred, as its redistribution over 2-4 hours post-injection highlights viable but underperfused myocardium, distinguishing it from nonviable scar tissue.[105]Standard protocols emphasize stress testing to provoke flow heterogeneity, with exercise on a treadmill (e.g., Bruce protocol) preferred for patients able to achieve ≥85% of maximum predicted heart rate; alternatively, pharmacological vasodilation using adenosine (140 μg/kg/min intravenous infusion over 6 minutes) simulates stress in those with contraindications to exercise. Tracer injection occurs at peak stress, followed by imaging 15-60 minutes later for 99mTc agents, with rest imaging performed similarly or on a separate day to minimize radiation exposure. Acquisition employs rotating gamma cameras for 3D tomographic reconstruction via iterative algorithms, incorporating CT-based attenuation correction to account for photon absorption by soft tissues like the diaphragm or breast, thereby reducing artifacts and enhancing specificity.[106][31]Interpretation relies on visual and semi-quantitative analysis of short-axis and polar map displays, where perfusion defects are scored using a 17-segment model of the left ventricle, assigning values from 0 (normal) to 4 (absent uptake) per segment to yield the summed stress score (SSS, range 0-68). A high SSS (>13) indicates extensive ischemia or scar, while the summed difference score (SDS = SSS minus summed rest score) quantifies reversibility; reversible defects (SDS ≥4) signify inducible ischemia from stenotic coronaries, whereas fixed defects (no change from stress to rest) typically represent myocardial scar, though mild fixed defects may reflect hibernating viable tissue. Gated SPECT further assesses ejection fraction and wall motion to corroborate findings.[107]In clinical applications, SPECT MPI demonstrates approximately 85% sensitivity for detecting obstructive CAD when defining it as ≥50% stenosis on angiography, outperforming stress echocardiography in perfusion specificity while providing prognostic insights. Post-myocardial infarction, it stratifies risk by defect extent, with SSS ≥8 predicting adverse events like reinfarction or death, guiding revascularization decisions. Hybrid SPECT/CT systems facilitate coregistration of perfusion data with coronary anatomy from low-dose CT, improving defect localization and reducing false positives in obese patients or those with attenuation artifacts.[31]
Positron Emission Tomography (PET)
Positron emission tomography (PET) plays a pivotal role in cardiac imaging by enabling the quantitative assessment of myocardial perfusion and metabolism, offering insights into coronary artery disease, viability, and microvascular function that surpass the relative perfusion evaluations typical of single-photon emission computed tomography (SPECT). Unlike SPECT, PET utilizes positron-emitting radiotracers to provide absolute measurements of myocardial blood flow (MBF) and metabolic activity, facilitating precise diagnosis and risk stratification in patients with suspected ischemia. This modality's ability to quantify flow at rest and during stress, along with metabolic tracers, supports its clinical superiority for detecting subtle abnormalities in myocardial perfusion and viability.[108][109][110]Key radiotracers in cardiac PET include rubidium-82 (Rb-82) and nitrogen-13 ammonia (N-13 ammonia) for perfusion imaging, which allow dynamic acquisition to measure MBF in ml/g/min. Rb-82, with its ultra-short half-life of 76 seconds, is generator-produced and enables rapid sequential rest-stress studies, while N-13 ammonia, with a 10-minute half-life, requires on-site cyclotron production but offers excellent image quality for quantitative perfusion. For metabolic assessment, particularly in viability studies, fluorine-18 fluorodeoxyglucose (F-18 FDG) is employed to evaluate glucose utilization in the myocardium, highlighting viable but dysfunctional tissue in ischemic regions. Additionally, 18F-flurpiridaz (approved by the FDA in September 2024) is a newer perfusion tracer with high myocardial extraction (94%), enabling accurate quantitative MBF and available as unit doses due to its 110-minute half-life. These tracers' short half-lives (except FDG's and flurpiridaz's 110 minutes) permit low radiation doses and multiple scans in a single session.[108][111][112][113]Quantitative metrics derived from PET include absolute MBF at rest (typically 0.4–1.2 ml/g/min) and stress (often >2.0 ml/g/min in healthy myocardium), with coronary flow reserve (CFR) calculated as the ratio of stress MBF to rest MBF; a normal CFR exceeds 2.0, indicating preserved microvascular function. These measurements provide objective evaluation of endothelial and microvascular integrity, outperforming SPECT's relative uptake assessments. In clinical applications, PET excels in diagnosing microvascular angina through reduced CFR despite normal epicardial arteries and identifying hibernating myocardium—dysfunctional but viable tissue—via preserved FDG uptake in perfusion-metabolism mismatch patterns, guiding revascularization decisions. Additionally, FDG-PET aids in the overlap with oncology by detecting and characterizing cardiac tumors, distinguishing malignant from benign masses based on metabolic activity.[114][115][112]PET's advantages include higher spatial resolution of approximately 4–6 mm compared to SPECT's 12–15 mm, enabling better delineation of small perfusion defects and microcirculatory changes. However, limitations such as the need for a cyclotron for N-13 and FDG production (though Rb-82 uses a generator), higher operational costs (scans ~$900–$1400), and limited availability restrict its widespread adoption relative to more accessible alternatives like SPECT.[116][117][118]
Invasive Imaging Techniques
Coronary Catheterization and Angiography
Coronary catheterization and angiography is the invasive reference standard for evaluating coronary artery luminal narrowing and overall vascular anatomy, providing high-resolution two-dimensional projections essential for guiding therapeutic decisions. The procedure typically begins with vascular access via the radial or femoral artery, where a sheath is inserted to facilitate guidewire and catheter navigation to the aortic root and coronary ostia. A guide catheter is advanced over the guidewire, and iodinated contrast is selectively injected into the left and right coronary arteries under fluoroscopic guidance to perform cine angiography, capturing dynamic images of blood flow and lesions in multiple projections. Radial access is increasingly preferred over femoral due to reduced procedural time and complications in most patients, though femoral access may be selected for complex cases requiring larger sheaths.Interpretation of angiographic images focuses on flow dynamics and lesion characteristics to stratify risk and plan interventions. The Thrombolysis in Myocardial Infarction (TIMI) flow grade assesses epicardial perfusion, with grade 0 indicating no antegrade flow beyond an occlusion, grade 1 minimal penetration without distal perfusion, grade 2 partial flow with delayed distal filling, and grade 3 normal flow with complete and timely distal opacification. For patients with multivessel disease, the SYNTAX score quantifies anatomical complexity by evaluating the number, location, and features of lesions—such as occlusions, bifurcations, and tortuosity—yielding a total score that predicts procedural outcomes and helps decide between percutaneous coronary intervention (PCI) and coronary artery bypass grafting. While non-invasive coronary CT angiography serves as a precursor for initial screening, invasive angiography provides the definitive lumen outline necessary for intervention.This technique is primarily applied in acute myocardial infarction for immediate PCI to restore flow in the culprit vessel, significantly reducing mortality compared to fibrinolysis. It also aids pre-surgical planning by delineating anatomy for bypass grafting or valve procedures, often integrated with Heart Team discussions using scores like SYNTAX to optimize revascularization strategy. Adjunctive right heart catheterization can be performed concurrently via venous access to measure pulmonary pressures and assess pulmonary hypertension, informing comprehensive hemodynamic evaluation.Despite its utility, the procedure carries risks, including vascular complications such as hematoma or pseudoaneurysm, occurring in approximately 1-5% of cases depending on access site, with radial approaches conferring lower rates than femoral. Contrast-induced acute kidney injury affects 2-13% of patients, particularly those with preexisting renal impairment, though hydration and minimized contrast volumes mitigate this. Effective radiation exposure typically ranges from 5-10 mSv for diagnostic procedures, equivalent to 2-3 years of background radiation.
Intracoronary Imaging
Intracoronary imaging encompasses catheter-based techniques that provide high-resolution, cross-sectional views of the coronary artery lumen, vessel wall, and plaque composition from within the vessel, offering insights beyond the luminal silhouette provided by angiography. The two primary modalities are intravascular ultrasound (IVUS) and optical coherence tomography (OCT), which are used to assess plaque burden, characterize lesion morphology, and optimize percutaneous coronary interventions (PCI). These techniques enable real-time evaluation during catheterization, guiding decisions in ambiguous lesions and improving procedural outcomes by detecting issues like stent underexpansion.IVUS employs a catheter-mounted transducer operating at frequencies of 20 to 45 MHz to generate grayscaleultrasound images, providing a 360-degree view of the vessel with axial resolutions of 150 to 250 μm and penetration depths up to 5 to 10 mm. It accurately measures lumen dimensions, plaque area, and vessel diameter, with a minimal lumen area (MLA) less than 4 mm² often indicating functionally significant stenosis in non-left main coronary arteries.[119] IVUS distinguishes plaque types based on echogenicity: fibrous plaques appear bright and homogeneous, lipid-rich plaques show low echogenicity with shadowing, and calcified plaques produce acoustic shadowing with high reflectivity. This modality excels in quantifying overall plaque burden and external elastic lamina measurements, which are critical for assessing positive remodeling in early atherosclerosis.[120]OCT utilizes near-infrared light interferometry via a fiber-optic catheter to produce images with axial resolutions of 10 to 20 μm and lateral resolutions approaching 20 to 30 μm, enabling detailed visualization of superficial structures up to 1 to 2 mm in depth. It excels at identifying thin-cap fibroatheromas (TCFA), defined by a fibrous cap thickness less than 65 μm overlying a necrotic core, which are associated with plaque vulnerability and rupture risk. OCT also assesses stent apposition and expansion post-implantation, detecting malapposition (strut-to-wall distance >0.2 mm) and neointimal coverage with greater precision than grayscale imaging. However, OCT requires transient blood clearance for optimal imaging, limiting its penetration compared to ultrasound-based methods.[121]Clinical applications of intracoronary imaging include evaluating intermediate or ambiguous angiographic lesions, where IVUS or OCT can quantify stenosis severity and plaque composition to inform revascularization decisions. The 2024 European Society of Cardiology (ESC) Guidelines recommend (Class I, Level A) the use of IVUS or OCT to guide PCI in complex cases such as left main disease, bifurcations, long lesions, or calcified stenoses to reduce adverse events.[122] In stent optimization, both modalities detect underexpansion (e.g., stent area <5.5 mm² in non-left main vessels) and edge dissections, reducing the risk of target lesion failure; randomized trials show IVUS-guided PCI lowers major adverse cardiac events compared to angiography alone. OCT is particularly valuable for guiding bioresorbable scaffolds, assessing strut degradation and vessel healing over time. For ambiguous left main lesions, IVUS-derived MLA thresholds (e.g., <4.5 mm²) correlate with fractional flow reserve ≤0.80, aiding deferral or intervention strategies.[123]Compared to IVUS, OCT offers superior resolution for superficial details like fibrous cap thickness and intimal hyperplasia but has shallower tissue penetration, making it less effective for deep plaque assessment or large vessel evaluation. IVUS provides better visualization of adventitial structures and overall plaque volume, though it may underestimate neointimal thickness due to lower resolution. Emerging hybrid IVUS-OCT systems integrate both modalities for comprehensive imaging, combining deep penetration with high superficial resolution to enhance plaque characterization and PCI outcomes.[124] Both techniques are susceptible to artifacts, such as guidewire shadowing or nonuniform rotational distortion, which can bias measurements; however, OCT's faster pullback speeds (18 to 36 mm/s) reduce motion artifacts relative to IVUS (0.5 to 1 mm/s). Meta-analyses indicate equivalent clinical benefits for PCI guidance in complex lesions, with OCT preferred for tissue characterization and IVUS for procedural safety in calcified vessels.[125][126]
Physiological Flow Assessments
Physiological flow assessments in cardiac imaging evaluate the functional significance of coronary stenoses by measuring pressure-derived indices during invasive catheterization, providing insights into lesion-induced ischemia beyond anatomical visualization. These methods, primarily fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR), guide decisions on revascularization, particularly for intermediate lesions where angiography alone is inconclusive.[127]Fractional flow reserve (FFR) quantifies the pressure drop across a stenosis under conditions of maximal hyperemia, induced typically by intravenous or intracoronary adenosine, to simulate stress-induced flow demand. It is calculated using a coronary pressure wire positioned distal to the lesion, with the formula FFR = P_d / P_a, where P_d is the mean distal coronary pressure and P_a is the mean proximal aortic pressure during hyperemia; this ratio assumes negligible microvascular resistance under maximal vasodilation, isolating the epicardial stenosis's impact. An FFR value ≤ 0.80 indicates hemodynamically significant ischemia warranting intervention, while values > 0.80 suggest deferral is safe, though a gray zone of 0.75–0.80 may require clinical judgment. Microvascular dysfunction, such as in post-infarct settings or left ventricular hypertrophy, can elevate FFR readings by impairing hyperemia, potentially underestimating lesion severity.[127]Instantaneous wave-free ratio (iFR) offers a non-hyperemic alternative, measuring the resting pressure ratio during a specific diastolic "wave-free" period when microvascular resistance is naturally low and stable, calculated as iFR = P_d / P_a without adenosine. A cutoff of ≤ 0.89 denotes significant stenosis, aligning closely with FFR's diagnostic accuracy while avoiding pharmacological induction. This approach reduces procedure time and patient discomfort, as adenosine-related side effects like chest pain, dyspnea, or bradycardia occur in up to 68% of cases with FFR but are eliminated with iFR.[128]These indices are particularly applied to intermediate stenoses (40–70% diameter reduction), where visual assessment is unreliable, to determine ischemia risk and avoid unnecessary stenting. In multivessel disease, FFR or iFR guides selective percutaneous coronary intervention (PCI), reducing stent use by about 30% compared to angiography alone. Post-PCI optimization involves re-measuring FFR or iFR to confirm functional success, with values < 0.86–0.90 prompting additional interventions like post-dilation, improving outcomes in up to 20% of cases where initial results are suboptimal.[129][130][131]The Fractional Flow Reserve versus Angiography for Multivessel Evaluation (FAME) trial demonstrated that FFR-guided PCI in multivessel disease reduced the 1-year rate of death, myocardial infarction, or repeat revascularization to 13.2% versus 18.3% with angiography-guided PCI (P=0.02), with fewer stents deployed (1.9 versus 2.7 per patient). Similarly, the DEFINE-FLAIR trial established iFR's noninferiority to FFR, with 12-month adverse event rates of 6.7% for iFR-guided versus 6.1% for FFR-guided strategies (P=0.007 for noninferiority), alongside procedural efficiencies. Despite these benefits, FFR's reliance on adenosine limits its use in patients with contraindications like asthma or atrioventricular block, though iFR mitigates this issue.[130][128][132]
Multimodality and Emerging Approaches
Hybrid Imaging Techniques
Hybrid imaging techniques in cardiac imaging integrate functional and anatomical data from complementary modalities, enabling more precise diagnosis and management of coronary artery disease (CAD) and other myocardial conditions. These approaches, such as PET/CT, SPECT/CT, and PET/MRI, facilitate attenuation correction, coregistration, and comprehensive assessment in a single session, reducing artifacts and improving clinical workflow.[133]PET/CT combines positron emission tomography for myocardial perfusion imaging—using tracers like ^{13}N-ammonia or ^{82}Rb—with computed tomography for attenuation correction and coronary angiography, all within approximately 45 minutes. This hybrid setup enhances specificity for CAD detection to 93%, compared to 73% for standalone SPECT, by accurately localizing perfusion defects and identifying mismatched findings, such as abnormal PET perfusion with normal CT anatomy indicating microvascular dysfunction.[133] Quantitative myocardial blood flow measurements (e.g., normal stress perfusion at 3.54 ± 1.01 ml·min^{-1}·g^{-1}) further aid in assessing ischemia severity and guiding revascularization.[133] In PET/CT, CT-based attenuation correction minimizes false positives from soft-tissue artifacts, while integrated angiography localizes stenoses to specific coronary territories.[134]SPECT/CT similarly merges single-photon emission computed tomography perfusion data (e.g., with ^{99m}Tc-sestamibi) with low-dose CT for attenuation correction and anatomical coregistration, improving the visualization of reversible defects indicative of ischemia. This coregistration aligns perfusion abnormalities with coronary calcium scores or vessel anatomy, increasing diagnostic specificity and confidence in identifying hemodynamically significant lesions.[135] For instance, hybrid SPECT/CT reveals matched perfusion defects and stenoses, prompting revascularization in 41% of cases versus 11% for unmatched findings, thereby optimizing treatment strategies and reducing unnecessary invasive procedures.[136]PET/MRI provides simultaneous acquisition of metabolic PET data (e.g., ^{18}F-FDG for viability) and detailed MRI tissue characterization, such as late gadolinium enhancement (LGE) for fibrosis detection, with reduced radiation exposure of 1-7 mSv compared to PET/CT.[137] In viability assessment, FDG-LGE fusion identifies hibernating myocardium, supporting decisions on revascularization with up to 80% functional recovery potential.[134] This synergy is particularly valuable in inflammatory conditions like cardiac sarcoidosis, where PET detects active inflammation and MRI quantifies edema or scarring, achieving 94% sensitivity for diagnosis and aiding risk stratification for major adverse cardiac events.[138]Overall, these hybrids minimize misregistration through motion-corrected coregistration and streamline one-session protocols, enhancing specificity over standalone functional imaging. In oncology-cardiology overlaps, such as sarcoidosis, they integrate inflammatory and structural insights for comprehensive evaluation.[137][138]
Artificial intelligence (AI) has revolutionized cardiac imaging through automated segmentation of anatomical structures, enhancing efficiency and reproducibility. Convolutional neural networks (CNNs), such as fully convolutional networks and U-Net variants, enable precise delineation of left ventricular (LV) contours in echocardiography and cardiac magnetic resonance imaging (MRI), achieving Dice similarity coefficients often exceeding 95% on benchmark datasets like ACDC. These models reduce manual annotation time from hours to seconds while maintaining high inter-observer agreement, facilitating quantitative assessments of ventricular function and volumes. In coronary computed tomography angiography (CCTA), deep learning algorithms automate plaque quantification, including non-calcified plaque volume, with intraclass correlation coefficients (ICC) of 0.964 against expert manual readings and 0.949 versus intravascular ultrasound, enabling better characterization of atherosclerotic burden.[139][140][141]Predictive analytics powered by machine learning (ML) integrate multimodality imaging data to forecast clinical outcomes, surpassing traditional risk scores. For instance, ML models combining CCTA-derived plaque features with coronary artery calcium scores predict major adverse cardiac events (MACE) with area under the curve (AUC) values up to 0.86, outperforming single-modality approaches by identifying high-risk patients for targeted interventions. These models leverage radiomic features and anatomical metrics to stratify ischemia risk, supporting personalized management in stable coronary artery disease.[142][143]Emerging technologies are expanding the scope of cardiac imaging beyond static anatomy to dynamic physiology. Four-dimensional (4D) flow MRI captures time-resolved, three-directional blood flow, quantifying vorticity and hemodynamic parameters like wall shear stress and energy loss, which correlate with aortic valve dysfunction and pulmonary hypertension severity. Photon-counting CT detectors achieve ultra-low radiation doses (as low as 1-2 mSv for CCTA) through spectral separation and noise reduction, preserving diagnostic quality for plaque and stenosis evaluation. Wireless intracardiac echocardiography provides tetherless, high-resolution real-time visualization during interventional procedures, such as transcatheter valve repairs, minimizing setup complexity and radiation exposure.[144][145][146][147]Recent developments post-2022 highlight AI's role in virtual physiological assessments, particularly AI-enhanced fractional flow reserve from CT (FFR-CT). Multicenter trials, including analyses from the HeartFlow platform, demonstrate that AI-optimized FFR-CT improves diagnostic accuracy for hemodynamically significant stenoses (sensitivity 89%, specificity 87%) compared to CCTA alone, reducing unnecessary invasive angiographies by up to 65% in intermediate-risk patients. These advancements enable non-invasive simulation of coronary physiology, guiding revascularization decisions with prognostic benefits for MACE reduction.[148][149]Future directions in cardiac imaging emphasize AI integration amid key challenges. Standardization of algorithms remains critical to ensure interoperability across diverse scanners and populations, as variability in training data can lead to inconsistent performance. Regulatory progress includes FDA clearances for over 40 AI tools in cardiac imaging from 2023 to 2025, such as HeartFlow's plaque analysis software and UltraSight's AI-guided ultrasound, accelerating clinical adoption. Ethical issues, including algorithmic bias from underrepresented demographics in training datasets, pose risks of diagnostic disparities; mitigation strategies like diverse data augmentation and bias audits are essential for equitable outcomes.00142-0/fulltext)[150][151][152][17]