The Evolving Landscape of Non-invasive Cardiac Autonomic Imaging

The cardiac autonomic nervous system (CANS) exerts moment-to-moment control over heart rate, contractility, conduction velocity, and coronary vascular tone through its sympathetic and parasympathetic branches. Dysregulation of this delicate balance is a hallmark of numerous cardiovascular pathologies, including heart failure, atrial fibrillation, hypertension, and sudden cardiac death. Historically, clinicians have relied on indirect, global metrics like heart rate variability (HRV) or baroreflex sensitivity to gauge CANS function. While valuable, these tools lack spatial resolution, cannot differentiate between neuronal subtypes, and offer only a blurred functional picture of the cardiac neural network. The recent revolution in non-invasive imaging technologies now permits direct visualisation and quantification of CANS structure and activity, opening a new era of precision autonomic assessment. This article provides a comprehensive, state-of-the-art overview of these advances, their underlying principles, clinical applications, and the trajectory of future innovation.

Foundations of the Cardiac Autonomic Nervous System

The CANS comprises an intricate mesh of intrinsic cardiac ganglia (located in epicardial fat pads) and extrinsic projections from the stellate ganglia, vagus nerve, and paravertebral chain. Sympathetic activity accelerates heart rate, enhances contractility, and promotes ventricular arrhythmias; parasympathetic (vagal) input slows heart rate, reduces atrioventricular conduction, and exerts protective antifibrillatory effects. Assessment of both branches is essential for understanding disease mechanisms and guiding therapy.

Traditional non-invasive methods—such as HRV power spectral analysis, heart rate turbulence, or exercise recovery indices—measure net autonomic output but cannot localise abnormalities to specific neural pathways or cardiac regions. Moreover, these metrics are confounded by age, medications, respiration, and emotional state. The direct imaging approaches described below overcome these limitations by visualising innervation density, neurotransmitter turnover, or neural metabolism with high spatial and temporal resolution.

Modern Non-invasive Imaging Modalities for CANS Assessment

A growing arsenal of imaging techniques now enables detailed anatomical and functional characterisation of cardiac autonomic nerves. Each modality offers unique strengths and trade-offs in sensitivity, specificity, radiation exposure, and clinical accessibility.

Positron Emission Tomography (PET) with Specific Radiotracers

PET imaging of the cardiac sympathetic nervous system has become the gold standard for regional denervation assessment. The most widely used radiotracer is 11C-hydroxyephedrine (HED), a norepinephrine analogue that accumulates in presynaptic sympathetic nerve terminals via the uptake-1 transporter. Another tracer, 18F-fluorodopamine, is taken up by dopaminergic neurons and converted to norepinephrine, allowing visualisation of sympathetic stores. Reduced tracer uptake indicates nerve damage or dysfunction, often preceding functional mechanical impairment.

PET imaging has demonstrated that sympathetic denervation in patients with ischemic cardiomyopathy is heterogeneous—with infarcted zones showing absent uptake and peri-infarct regions exhibiting variable denervation that correlates with arrhythmia vulnerability. In heart failure with preserved ejection fraction (HFpEF), global sympathetic hyperactivity can be quantified by increased tracer washout rates. Quantification of tracer retention can also guide prognostication after myocardial infarction; a study by Fallavollita et al. (2014) found that patients with substantial denervation (>37.6% of left ventricle) had a significantly elevated risk of sudden cardiac arrest (reference: Fallavollita et al., J Am Coll Cardiol, 2014).

Limitations include radiation exposure, relatively high cost, limited availability of cyclotron-produced tracers, and the need for dedicated cardiac PET imaging protocols. Nonetheless, PET remains unmatched for its sensitivity and capacity to image the entire myocardial sympathetic neural network.

Cardiac Magnetic Resonance Imaging (MRI) for Autonomic Neurography

While MRI traditionally excels at anatomical imaging of myocardial scar and fibrosis, recent advances now permit visualisation of cardiac nerves. Diffusion tensor imaging (DTI) with cardiac gating allows tractography of epicardial and intramyocardial nerve fibre orientation. Because autonomic nerves are highly anisotropic (preferentially aligned along the long axis of the heart and coronary vessels), DTI can map their trajectory and detect disruption caused by infarction or inflammation.

Preliminary studies in humans have demonstrated that DTI-derived fractional anisotropy (FA) and mean diffusivity (MD) differ between healthy myocardium and regions affected by autonomic neuropathy in diabetes. T2-weighted MRI with inversion recovery sequences can also identify oedematous changes in the epicardial ganglia during acute myocarditis, providing a novel marker of autonomic involvement.

Furthermore, hyperpolarized 13C MRI and phosphorus-31 (31P) MR spectroscopy are emerging as tools to assess cardiac sympathetic metabolism by imaging substrates like 13C-labelled pyruvate or phosphocreatine/ATP ratios. The major advantages of MRI are the absence of ionising radiation, excellent soft-tissue contrast, and the ability to combine autonomic imaging with standard cardiac function, fibrosis, and perfusion assessment in a single session. Remaining challenges include long acquisition times, respiratory and cardiac motion artifacts, and the need for advanced post-processing algorithms to resolve thin nerve structures.

Near-Infrared Spectroscopy (NIRS) and Functional NIRS (fNIRS)

Near-infrared spectroscopy detects changes in the concentration of oxyhaemoglobin and deoxyhaemoglobin in superficial tissues, allowing real-time monitoring of local blood flow that correlates with neural activation. Although primarily used for cortical brain imaging, cardiac-gated NIRS can now be applied to the anterior chest wall to assess autonomic nerve activity indirectly through changes in coronary perfusion and microvascular reactivity.

Recent engineering advances have miniaturised NIRS sensors and improved depth penetration (up to 4-5 cm), making it feasible to interrogate epicardial fat pads containing intrinsic cardiac ganglia. In a proof-of-concept study, synchronous electrocardiography and NIRS signals showed detectable light-absorbance changes coincident with sympathetic nerve bursts during tilt-table testing. While NIRS cannot match PET or MRI for spatial resolution, its portability, low cost, and ability to provide continuous monitoring make it attractive for bedside or ambulatory autonomic assessment. Near-infrared spectroscopy is also being explored as a biomarker for autonomic storm in conditions such as takotsubo cardiomyopathy and autonomic dysreflexia after spinal cord injury.

Single-Photon Emission Computed Tomography (SPECT) with 123I-MIBG

123I-metaiodobenzylguanidine (MIBG) is an analogue of guanethidine that shares the same uptake and storage mechanisms as norepinephrine. 123I-MIBG SPECT has been widely used for cardiac sympathetic imaging, particularly in heart failure. The heart-to-mediastinum ratio (HMR) and washout rate provide indices of global sympathetic innervation density and tone. Reduced HMR and accelerated washout predict adverse outcomes in systolic heart failure, independent of left ventricular ejection fraction (see ADMIRE-HF study, Circ Cardiovasc Imaging, 2010).

Compared to PET, SPECT is more widely available, less expensive, and does not require on-site cyclotron production. However, its spatial resolution is inferior, and quantification of regional denervation is more challenging. Newest-generation cadmium-zinc-telluride (CZT) cameras offer improved count sensitivity and resolution, partly mitigating these limitations. Despite competition from PET, 123I-MIBG remains a mainstay for clinical autonomic imaging in many centres.

Emerging Hybrid Systems: PET/MRI and SPECT/CT

The combination of modalities addresses individual weaknesses. PET/MRI simultaneously acquires PET tracer kinetics and high-resolution MRI anatomical, functional, and metabolic data, enabling direct correlation of denervation with scar, oedema, or perfusion defects. This hybrid approach is particularly valuable for arrhythmogenic cardiomyopathy, where the autonomic substrate can be precisely overlapped with fibrotic tissue. Similarly, SPECT/CT provides attenuation correction and anatomic localisation, improving diagnostic confidence in patients with complex cardiac anatomy or prior sternotomy. As these integrated systems become more refined, they promise a holistic, single-examination evaluation of the cardiac autonomic landscape.

Clinical Applications and Translational Impact

Non-invasive CANS imaging is moving from research tool to clinical necessity in several well-defined scenarios.

Heart Failure and Left Ventricular Assist Devices

Sympathetic hyperactivation is a hallmark of chronic heart failure and correlates with mortality, arrhythmia, and readmission rates. Serial PET or 123I-MIBG imaging can track the progression of cardiac denervation and its reversal with medical therapy (e.g., beta-blockers, sacubitril/valsartan) or resynchronisation. In patients receiving left ventricular assist devices (LVADs), the autonomic imaging can identify persistent sympathetic activation that may predict right ventricular failure or arrhythmia development after implant, guiding pharmacologic modulation of the “autonomic storm.” Current guidelines from the European Association of Nuclear Medicine and the American College of Cardiology recognise cardiac sympathetic imaging as a useful prognostic marker in systolic heart failure (Class IIa recommendation).

Diabetes and Diabetic Cardiovascular Autonomic Neuropathy (CAN)

Cardiovascular autonomic neuropathy is a severe complication of diabetes that increases the risk of silent ischaemia, arrhythmia, and sudden death. NIRS and DTI-MRI can detect early degeneration of cardiac nerves before clinical HRV abnormalities emerge, enabling earlier intervention with intensive glycaemic control and neuroprotective agents. A 2022 study using PET with 11C-HED showed that nearly half of asymptomatic type 2 diabetes patients had regional sympathetic denervation confined to the distal posterolateral wall, a pattern distinct from ischaemic heart disease (Bhatt et al., J Nucl Cardiol, 2022).

Atrial Fibrillation and Autonomic Ganglionated Plexi (GPs)

The intrinsic cardiac autonomic system, particularly the ganglionated plexi located near the pulmonary vein antra and ligament of Marshall, plays a pivotal role in triggering and maintaining atrial fibrillation. Hybrid PET/MRI or 123I-MIBG SPECT can visualise these GPs and quantify their metabolic activity. Pre-ablation imaging may identify high-activity zones that are more likely to sustain re-entry, enabling tailored catheter ablation strategies that target not only pulmonary vein isolation but also active autonomic substrates. Early evidence suggests that patients with intense GP activity on imaging have lower atrial fibrillation recurrence rates when those regions are also ablated.

Arrhythmia Risk Stratification Post-Myocardial Infarction

As noted, sympathetic denervation assessed by PET is a powerful predictor of ventricular arrhythmia and sudden cardiac death after myocardial infarction. Current risk-stratification tools (LVEF, programmed electrical stimulation) have limited positive predictive value. Adding autonomic imaging may improve selection of patients who benefit from implantable cardioverter-defibrillators (ICDs). The ongoing PARADIGM-ICD trial (NCT03627650) is prospectively using 11C-HED PET to guide ICD implantation in patients with intermediate LVEF (35–50%). If positive, this could shift clinical practice toward imaging-guided device therapy.

Technical and Practical Considerations

Widespread clinical adoption of CANS imaging faces several hurdles. Standardisation of acquisition protocols, tracer dosage, image reconstruction, and quantitative metrics is incomplete. For PET, harmonisation across different cameras and tracer kinetics is necessary. For MRI, DTI sequences must be optimised to minimise motion and improve signal-to-noise ratio in the beating heart. The majority of published evidence comes from single-centre cohorts with modest sample sizes; large multicentre trials are needed to validate imaging-derived parameters as therapeutic targets and surrogate endpoints. Moreover, training of readers and integration of autonomic imaging reports into routine clinical workflows present logistical challenges.

Cost-effectiveness also requires scrutiny. A comprehensive CANS imaging workup (e.g., PET/CT with tracer, or multi-sequence MRI) is more expensive than a standard echocardiogram or Holter monitor. However, if it prevents unnecessary ICD implantations or guides successful ablation procedures, it may prove cost-saving in the long term. Reimbursement policies in many countries remain limited or absent.

Future Directions and Emerging Technologies

The next decade will likely witness several transformative developments in non-invasive cardiac autonomic imaging.

Novel radiotracers for parasympathetic imaging: While current tracers target the sympathetic system, no widely available clinical agent selectively visualises cardiac parasympathetic nerves. Molecules targeting the vesicular acetylcholine transporter or muscarinic M2 receptors are being validated in preclinical models; 18F-fluoroethoxybenzovesamicol (FEOBV) has shown preliminary feasibility in human brain imaging and could be adapted for cardiac use.

Artificial intelligence and deep learning: Machine learning algorithms can automatically segment cardiac autonomic structures from DTI tractography or PET parametric maps, extract radiomic features, and predict outcomes with higher accuracy than conventional indices. For instance, a convolutional neural network trained on 123I-MIBG images can predict three-year mortality with an AUC of 0.88, outperforming HMR alone.

Molecular ultrasound: Contrast-enhanced ultrasound using targeted microbubbles conjugated to antibodies against nerve-specific antigens (e.g., tyrosine hydroxylase, neuropeptide Y) could theoretically provide real-time, bedside imaging of cardiac nerve activity. While still in preclinical stages, this approach would combine portability with molecular specificity.

Wearable and implantable NIRS: Miniaturised NIRS patches integrated into chest straps or adhesive devices could enable continuous ambulatory monitoring of autonomic reactivity patterns during daily activities, exercise, or stress. These data would complement imaging-derived snapshots by providing temporal dynamics of CANS behaviour over weeks to months.

Conclusion

Non-invasive imaging has fundamentally changed the assessment of the cardiac autonomic nervous system, moving beyond global HRV metrics toward precise, regional, and molecular characterisation of sympathetic and, potentially, parasympathetic innervation. Positron emission tomography with specific tracers remains the most sensitive tool for denervation imaging; magnetic resonance techniques offer exceptional anatomical detail without radiation; and optical methods such as NIRS promise low-cost, longitudinal monitoring. These technologies have already demonstrated clinical utility for risk stratification in heart failure, myocardial infarction, diabetic neuropathy, and arrhythmias, with emerging roles in guiding ablation and device therapy. Continued innovation in hybrid systems, tracers, and computational analysis will further refine our ability to image the neural circuits that govern cardiac rhythm and contractile function. As accessibility improves and standardisation matures, non-invasive CANS imaging is poised to become an indispensable component of personalised cardiovascular medicine.