Analysis of Hemodynamic Markers in Atrial Fibrillation Using Advanced Imaging Techniques

Abstract

Atrial fibrillation (AF) is a prevalent heart arrhythmia, characterized by an irregularly irregular rhythm and the absence of identifiable P waves on ECG. Given the loss of effective atrial contraction, AF carries a risk of serious complications. If untreated, AF can promote thrombogenesis, leading to stroke, systemic embolism (e.g., limb or organ ischemia), and myocardial infarction. These serious complications highlight the importance of understanding AF and assessing stroke risk to guide optimal management of this chronic arrhythmia. Congruent with recent technological developments, advanced imaging has emerged as a modality to better understand AF. This review highlights advanced imaging techniques and their advantages, with a focus on 4D flow MRI, a novel modality that enables visualization of blood flow patterns in three dimensions and provides unique insights into cardiac hemodynamics. It also synthesizes the current literature on key hemodynamic markers identified by 4D flow MRI, including blood flow stasis, wall shear stress, and vorticity. Quantifying these markers has improved predictive accuracy of future stroke risk in AF patients, allowing clinicians to risk stratifying their patients and optimize management. Finally, the review discusses potential future markers that may further refine our understanding of AF and inform patient care.

1. Introduction

Atrial fibrillation (AF) is a cardiac rhythm characterized by rapid and irregular electrical activity that leads to loss of effective atrial contraction. AF is one of the most prevalent sustained cardiac arrhythmias worldwide, affecting millions of individuals. In 2019, the global prevalence was estimated at 59 million, a figure that continues to rise markedly [1]. The condition is strongly age-dependent, with prevalence increasing from 0.2% among individuals aged 45–54 to 8.0% in those aged 75 years and older [2]. Geographic disparities in prevalence have also been noted, with higher rates in developed countries due to improved diagnostics and longer life expectancy [3]. While AF is rarely fatal, its associated risks (Figure 1), including stroke, heart failure, and overall cardiovascular morbidity, significantly contribute to healthcare burdens worldwide [4]. Specifically, AF accounts for approximately 80% of cardioembolic strokes, a distinction that is clinically important given the differences in treatment and prognosis between cardioembolic and atherosclerotic stroke [5].

AF arises from a combination of initiating triggers and an atrial substrate capable of sustaining the arrhythmia. Common mechanisms include focal ectopic activity originating outside the sinoatrial node, reentrant electrical circuits within the atrial myocardium, and fibrotic remodeling of atrial tissue driven by factors such as aging, hypertension, and diabetes [6]. Over time, these mechanisms drive progressive atrial structural and electrical remodeling, including slowed conduction and altered ion channel function [7]. These changes perpetuate AF and increase susceptibility to recurrent and sustained episodes, a phenomenon commonly described as “AF begets AF” [7]. Given these pathophysiological changes, the early detection and diagnosis of AF are critical to guide effective management and reduce long-term health risks.

The prognosis of AF depends on the timely identification and management of associated risks. While many patients present with characteristic electrocardiographic changes and symptoms, a substantial proportion have silent AF—remaining asymptomatic but carrying the same risks as overt disease [8]. Prognosis is also influenced by AF subtype, classified by duration as paroxysmal, persistent, or permanent, with important implications for treatment. For example, in addition to antiarrhythmic drug therapy, persistent AF may be treated using cardioversion, whereas paroxysmal AF (PAF) is more often treated with catheter ablation [9]. Given the need to understand AF on a deeper level, imaging has emerged as an innovative tool to further understand the pathophysiology of AF. Cardiac imaging, in particular, has provided a more comprehensive appreciation of mechanistic changes in AF, both structural and electrical [10]. Moreover, imaging now informs treatment selection, as interventions such as cardioversion and catheter ablation are frequently guided by imaging findings, underscoring its importance [10].

This review consolidates and evaluates current imaging modalities and explores the role of hemodynamic markers in AF detection and stroke risk stratification, with particular focus on their utility in monitoring disease progression and complications such as stroke and limb ischemia. We examine key imaging techniques—including echocardiography, cardiac computed tomography (CT), and magnetic resonance imaging (MRI) —and assess their utility in detecting structural and functional atrial abnormalities. We also explore the role of hemodynamic markers in AF detection and stroke risk stratification, as seen in Figure 2. A key emphasis is placed on identifying which imaging-derived hemodynamic markers hold the strongest predictive value for thromboembolic events, thereby refining risk assessment models. By synthesizing current evidence, this review aims to advance understanding of AF pathophysiology, improve clinical diagnosis, and inform evidence-based management guidelines to prevent cardioembolic stroke. Ultimately, we highlight existing knowledge gaps and propose future directions for research in AF imaging and hemodynamic profiling.

2. Standard of Care

The current standard of care analysis, for the purpose of this review article, can be consolidated into diagnosis and management. While there are many nuances to the standard of care, dependent on AF type and other risk factors, the current guidelines formulated by Andrade et al. (2020) and Joglar et al. (2023) comprehensively cover these details [11,12]. In clinical practice, a 12–lead electrocardiogram (ECG) is often used to confirm the diagnosis of AF due to its high sensitivity and specificity compared to other diagnostic techniques [13]—specifically characterized by an absence of a clear P wave, as seen in Figure 3 [11]. In non-diagnosed patients with PAF, the ECG may capture sinus rhythm at the time of testing, which makes ECG-based detection of AF unreliable [11]. In these patients, non-invasive imaging techniques can be leveraged in routine practice, to detect hemodynamic markers that inform on a patient’s paroxysmal AF status, which remain evident even in sinus rhythm [14]. This gap in coverage presents a unique opportunity for non-invasive imaging to integrate into specialized practice to detect remodeling and hemodynamic markers associated with AF to identify the underlying mechanistic causes of AF, ultimately fostering a preventative approach.

Following diagnosis, treatment of AF is categorized into two major strategies: rhythm control and rate control, often nuanced and guided by presentation of symptoms and hemodynamic stability [11,12]. Abnormal rhythm is managed by pharmacological or electrical cardioversion, again dependent on hemodynamic stability, amongst other factors [11,12]. Additionally, abnormal rate control is acutely managed with beta blockers or non-dihydropyridine calcium channel blockers, dependent on potential contraindications and side effect profiles [12]. Given the multitude of factors involved in deciding between rate and rhythm control, shared decision making with patients is essential [12]. Interestingly, treatment strategies for both rate and rhythm control are directly modulated by the percentage of left ventricular ejection fraction which is quantified by transesophageal echocardiography or cardiac MRI—underscoring the crucial role of imaging in AF treatment [11]. In addition to directing pharmacotherapy, imaging is quintessential in advancing long-term health outcomes through its role in guiding cardiac catheter ablation procedures. Specifically, when AF is attributed to an underlying structural abnormality, which can only be definitively identified through cardiac imaging, cardiac catheter ablation offers a therapeutic intervention by directly targeting and correcting the structural pathology [11,12]. This approach has the potential to prevent subsequent complications associated with AF, namely stroke, and to reduce the overall burden of the arrhythmia [11]. Given the growing evidence on the fundamental role of imaging in AF treatment, both pharmacologic and interventional, it is important to explore what other information can be isolated by cardiac imaging to facilitate early detection, ultimately preventing complications of AF.

3. Imaging Modalities

Given the importance of early detection for prognosis and treatment of AF, imaging has emerged as an integral aspect of the treatment protocol—optimizing management and reducing mortality. While an ECG can provide great diagnostic utility in standard clinical practice, understanding the underlying mechanism and pathophysiology of AF often requires advanced imaging techniques to mitigate the risk or prevent AF altogether. Specifically, cardiac imaging can elucidate risk factors for AF such as structural heart disease, left ventricular dysfunction, and changes in left atrial size and function [15,16]. Adequate identification and treatment of the risk factors for AF can greatly improve treatment outcomes for AF patients and reduce future stroke risk for example. Therefore, this section will explore the various imaging modalities used in AF management—ranging from echocardiography to cardiac MRI and specifically focusing on the strengths of each imaging modality in detecting the various risk factors for AF.

3.1. Echocardiography

Echocardiography plays a central role in the detection and management of AF. Transthoracic echocardiography (TTE) is typically the first line imaging modality, providing essential information on left atrial size, left ventricular function, and valvular pathology that can predispose patients to AF. When detailed images are required, particularly during pre-cardioversion or for thrombus detection, transesophageal echocardiography (TEE) offers superior visualization of the posterior cardiac structures, especially the LAA [17]. As such, TEE has become standard practice for guiding cardioversion and ruling out thrombi, reducing the risk of stroke for patients. Beyond structural assessment, advanced imaging modalities like speckle-tracking echocardiography (STE) provide information about atrial mechanics and function [18]. STE quantifies atrial deformation and function through strain parameters, which allows clinicians to detect early atrial remodeling before overt arrhythmia develops [18,19]. Eren et al. (2021) demonstrated that reductions in left atrial reservoir strain and strain rate during early diastole were independently associated with the development of AF in patients with atrioventricular nodal reentrant tachycardia, highlighting STE’s potential in identifying at-risk patients even in the absence of established AF [19]. The clinical value offered by echocardiography extends beyond diagnosis to treatment planning. Echocardiography informs anticoagulation decisions, candidacy for cardioversion or catheter ablation, and helps to characterize the areas of the heart that cause abnormal rhythms [15,16]. It also helps visualize pulmonary vein anatomy, screen for procedural contraindications, and assess recurrence following intervention, making it an essential part of both pre- and post-treatment [15,16].

Compared to other imaging modalities such as CT or MRI, echocardiography offers several advantages: it is non-invasive (for TTE), widely available, cost-effective, and free of ionizing radiation or contrast agents [16,20]. However, echocardiography can be limited by acoustic window quality, operator dependency—particularly for TTE—and, in the case of STE, dependence on accurate tracking, frame rate sensitivity, and inter-vendor variability, which can affect reproducibility and hinder the establishment of universal cut-offs [20,21,22,23]. Despite these limitations, the combined use of TTE, TEE, and STE provides a versatile toolkit to guide clinical decision making across all stages of AF care.

3.2. Cardiac CT

While echocardiography techniques have focused on increasing understanding of cardiac function to prevent future AF, cardiac CT has played a pivotal role in the management of AF [24]. Cardiac CT is particularly valuable for real-time guidance during catheter ablation and pre-cardioversion planning [24,25,26] as therapeutic procedures for AF. Additionally, cardiac CT facilitates understanding of the heart, specifically by noninvasively identifying thrombi in the LAA, as compared to invasive imaging performed by TEE, further highlighting the utility of CT [24,27]. Although cardiac CT is non-invasive and offers much utility, TEE still remains the gold standard for thrombi identification as cardiac CT can detect pseudo-thrombi based on the timing and delays of contrast [28].

Cardiac CT angiography has also been effective in predicting the risk of recurrent AF by measuring the fat radiomic profile around cardiac structures such as the LA, LAA, and pulmonary veins [29]. Specifically, a higher fat profile around the aforementioned cardiac structures, as seen on cardiac CT, is associated with AF and recurrent AF, underscoring its imperative role in AF risk management [29]. Therefore, it is evident that cardiac CT plays a pivotal role in the treatment of AF, while also minimizing complications due to the noninvasive nature of the imaging modality—notwithstanding the implied radiation exposure.

3.3. Cardiac MRI

Although echocardiography and cardiac CT possess significant advantages and unique use-cases in the management of AF, cardiac MRI excels in facilitating the understanding of subtleties in cardiac structure and cardiac function in patients with AF [30]. Specifically, MRI can generate multiple contrast weighted images, allowing clinicians to further understand pathologic disease processes. Within the framework of AF, there are three fundamental contrast mechanisms to understand—T1 weighting to isolate underlying anatomy and soft tissue structure, T2 weighting to isolate fluid and inflammation, and Late Gadolinium Enhancement to elucidate tissue scarring [31,32]. Coupled with its unparalleled ability to visualize cardiac tissue, MRI is noninvasive and without radiation exposure—further highlighting the advantageous nature of cardiac MRI. With this foundational information about the inner workings of cardiac MRI, it is important to illustrate how this imaging modality better allows us to manage AF.

Through a prognostic perspective, cardiac MRI could significantly improve the predictive ability to stratify and determine stroke risk in patients with AF—the most devastating complication [30,32,33,34,35]. As previously mentioned, left atrial fibrosis—a risk factor for atrial fibrillation—also has implications in predicting future stroke risk, as a remodeled, fibrotic LA is more susceptible to pathologic thrombogenesis leading to thrombotic stroke [30]. Specifically, the degree and severity of left atrial fibrosis—based on the Utah scale of enhancement on MRI—directly correlates with a greater risk of pathologic thrombosis [30,32]. Further, the fibrotic state of the LA can enhance the predictive value of stroke risk once added to the CHA₂DS₂-VASc score [30]. These findings directly highlight how cardiac MRI can determine the amount of atrial fibrosis, which ultimately allows clinicians to better identify patients with increased stroke risk and begin early management.

In addition to fibrotic identification in the LA, cardiac MRI also has the ability to identify various morphologies for the LA and LAA, due to its exceptional ability to visualize cardiac tissue [30]. Specifically, a more spherical LA may correlate with more frequent occurrences of pathologic thrombosis and including the degree of left atrial sphericity improves the positive predictive value of the CHA₂DS₂-VASc score for future thrombotic events [33]. Mechanistically, left atrial sphericity has been proposed to increase atrial wall tension, and decrease left atrial contractile efficacy—ultimately proving to be a risk factor for AF [36].

While LA morphology has demonstrated imperative clinical significance in the context of AF and future stroke risk, the morphology of the LAA is markedly more consequential for future thrombotic events. Specifically, it has been demonstrated that most pathologic thrombogenesis, resulting in sinister events such as cerebral infarct, originate in the LAA [34]. Given this finding, it is also crucial to elucidate LAA morphologies and their relation to future pathologic thrombogenesis. Although there is conflicting evidence on which LAA morphology is most predictive of future stroke risk, there is a predictive and pathologic hypothesis on the role of LAA morphology in stroke risk [34,35].

Ultimately, this review is more focused on the usage of cardiac MRI to predict future thromboembolic risk in AF patients; however, the review article by Zhao et al. (2020) offers a broad-spectrum analysis of many use-cases of cardiac MRI—from risk stratification to direct medical therapy [30]. Considering the overwhelming evidence on the utility of cardiac MRI, its role in clinical settings has proven to be instrumental in advancing the prognostic outlook of thrombotic events related to AF. Recently, technological innovations, such as 4D flow MRI, have been integrated into cardiac MRI to facilitate an even deeper understanding of the hemodynamics of the LA and LAA in patients with AF.

Although conventional cardiac MRI can isolate the underlying anatomy, 4D flow MRI furthers our physiologic understanding of pathologic conditions by demonstrating blood flow patterns over time within the heart in three dimensions—revealing a new focus for clinical investigation [37,38,39,40,41]. Through the identification of specific hemodynamic markers used to stratify future stroke risk, these investigations have yielded numerous findings that have ultimately enhanced our understanding of AF [42]. In addition to stratification of stroke risk, 4D flow MRI may provide clinical utility for post-treatment monitoring of patients with AF. In particular, atrial remodeling is known to continue following ablation therapy [43]. Certain types of atrial remodeling contribute to AF pathophysiology, and thus may contribute to atrial arrhythmia recurrence after ablation [44]. Since anticoagulants may only be continued in patients with arrhythmia recurrence, as seen in the recent ALONE-AF trial [45], cardiac MRI’s ability to detect atrial remodeling, and the ability of 4D flow MRI to detect hemodynamic perturbations indicative of AF, may play a role in guiding anticoagulant administration following ablation.

Given that the analysis of 4D flow MRI, in the context of AF and future stroke risk, is our primary focus of this article, our next section will take a deeper dive into the nuances of the primary hemodynamic markers identified in the literature—highlighting the potential gaps while identifying areas of future research.

4. Hemodynamic Markers of Atrial Fibrillation

With the advent of new imaging technologies, namely 4D flow MRI, various hemodynamic markers implicated in the prognosis of future stroke risk for AF have been isolated [14,38,39,42,46,47,48,49,50,51,52,53,54,55,56,57,58,59]. These hemodynamic markers will be explored thoroughly in the present section, with a particular emphasis on studies that employed 4D flow MRI, given its novelty as an imaging modality (

Table 1. Selected studies examining hemodynamic markers in the context of atrial fibrillation (AF), with a particular emphasis on studies employing 4D flow magnetic resonance imaging (MRI).

Author	Sample Size and Controls (N)	Methods	Hemodynamic Markers	Major Findings
[53]	29 age-matched participants; 19 healthy controls, 10 with AF (4 with persistent AF, 6 post intervention)	- 4D flow MRI	- Mean velocity - Flow coherence - Percentage of velocities > 0.2 m/s	Persistent AF patients exhibited reduced blood flow velocities and impaired flow coherence compared to post-intervention AF patients and healthy controls.
[38]	70 subjects; 62 AF patients (33 in sinus rhythm, 29 with persistent AF), 8 healthy controls	- 4D flow MRI	- Mean velocity - Median velocity - Stasis (% of atrial velocities < 0.2 m/s)	AF patients showed no left-right atrial flow velocity differences but had significantly lower velocities and higher stasis than controls.
[39]	75 subjects; 60 AF patients (30 in sinus rhythm, 30 in AF), 15 healthy controls)	- 4D flow MRI	- Peak velocity - Mean velocity - Stasis (% of velocities < 0.1 m/s)	AF patients had reduced LAA velocities, higher stasis, and more disorganized flow compared to controls, with worsening flow dynamics linked to higher CHA2DS2-VASc scores.
[54]	70 subjects; 40 AF patients in sinus rhythm, 20 age-appropriate controls, 10 young healthy volunteers	- 4D flow MRI	- Mean LA velocity - Median LA velocity - Peak LA velocity	AF patients exhibited significantly lower LA velocities than controls, with greater reductions linked to higher CHA2DS2-VASc scores.
[14]	60 subjects; 45 PAF patients, 15 healthy controls	- 4D flow MRI	- Vortex size (cm3) - Peak pulmonary vein inflow velocities - LA velocity distribution (mean, median, peak velocities) - Flow stasis (% of velocities < 0.1 m/s)	PAF patients had larger LA vortices than controls, which correlated with slower pulmonary vein flow, enlarged LA volume, and higher CHA2DS2-VASc scores.
[46]	109 subjects; 91 PAF patients and 18 healthy controls	- 4D flow MRI	- RTD Time Constant - Atrial velocity (percentage of particles with velocities < 0.2 m/s and <0.1 m/s)	PAF patients showed greater atrial stasis (higher RTD Time Constant) than controls, with even more pronounced stasis in those with elevated CHA2DS2-VASc scores.
[47]	15 subjects; 10 PAF patients and 5 age/gender-matched controls	- 4D flow MRI	- Flow velocity (mean, peak) - Stasis Fraction (volume with velocities < 10 cm/s) - Kinetic Energy - LA strain - Strain rates	PAF patients exhibited significantly reduced LA flow velocities, higher stasis, and lower kinetic energy compared to controls, indicating impaired atrial hemodynamics.
[48]	86 subjects; 64 in sinus rhythm and 22 in AF	- 4D flow MRI	- LA Peak velocity - LA Mean velocity - LA Vorticity - LA vortex volume - LA stasis	AF patients had greater LA stasis, lower peak velocity, and altered vortical flow, while LA peak velocity and vorticity were stable across heart rate, blood pressure, and rhythm changes.
[49]	25 patients with a history of AF	- A self-gated, 3D radial, free-running 5D flow MRI sequence	- Peak velocity - Mean velocity - Stasis	High AF burden correlated with increased LA stasis and reduced peak velocity and mean velocity.
[59]	80 subjects; 50 PAF patients and 30 healthy controls	- 4D flow MRI	- Direct flow - Retained inflow - Delayed ejection - Residual volume - All measures percentages of total LV flow	PAF patients had reduced direct flow and increased delayed ejection with occult LV hemodynamic inefficiencies despite normal systolic function.
[55]	95 participants: Group 1 (37 patients with persistent AF), Group 2 (35 individuals with no AF but similar stroke risk), Group 3 (23 low-risk individuals).	- 4D flow MRI	- LA Peak velocity - LA Mean velocity - LA Vorticity - LA vortex volume ratio	Patients with persistent AF (Group 1) had impaired LA flow velocities and vorticity, while Groups 1 and 2 (moderate-to-high stroke risk) showed altered LA flow in sinus rhythm, linked to LA and LV diastolic dysfunction.
[50]	N/A as this was a literature review	- Literature review assessing physiological and pathological vortex flow.	- Stasis: % LA velocities < 10 cm/s. - Peak/Mean Velocity - Vorticity/Vortex Volume: λ2-based vortex detection. - Residence Time: Particle tracking from pulmonary veins to mitral valve.	AF patients exhibit reduced vortex flow and increased stasis, with vortex flow preservation correlating to lower thrombotic risk.
[52]	3 3D-printed LA phantoms from 86 year old male patient with AF	- 3D printed phantoms - 4D flow MRI	- Stasis volume - ECAP - PRT - WSS	Correct occlusion reduced stasis volume and had the lowest ECAP and highest WSS, while the incorrect occlusion model showed high stasis and longer PRT compared to the corrected occlusion rate.
[51]	45 subjects; 35 AF patients, 10 healthy controls	- 4D flow MRI	- Peak velocity - Mean velocity - Stasis - HRV - Beat-to-beat variability	High HRV patients showed greater variability in flow metrics, lower mean velocity, higher stasis, and a correlation between longer RR intervals and increased stasis was observed.
[58]	1 AF patient with 4 wall motion control models: rigid, generic, semi-generic, patient-specific	- 4D CT-based LA segmentation - CFD simulations with OasisMove	- WSS - OSI - RRT - ECAP - KE	There were minimal LA hemodynamic differences between the models; the rigid model underestimated WSS and overestimated RRT/ECAP in the LAA, while generic/semi-generic models matched patient-specific motion.
[42]	12 studies with mixed cohorts (AF patients vs. healthy controls)	- Literature review analyzing 4D flow MRI and CFD simulations in select studies	- Velocity - Vorticity - Blood stasis - ECAP - KE - RTD	AF impacts velocity, stasis, ECAP, and vortices, with flow changes linked to thrombosis risk, CHA2DS2-VASc scores, LAA closure, ablation, and remodeling.
[57]	5 AF patients in sinus rhythm	- High-resolution CT - 4D flow MRI - Dynamic CFD (morphing vs. rigid walls)	- TAWSS - OSI - ECAP - Mitral valve outflow waveforms.	Morphing model improved TAWSS, OSI, and mitral flow accuracy; LAA had lower TAWSS and higher OSI than LA.
[56]	15 AF patients (10 paroxysmal AF, 5 persistent AF, 3 atrial flutter)	- MRI to detect LA fibrosis - Patient-specific CFD simulations	- TAWSS - Blood Age (indicator of stagnation) - Bipolar Voltage - Image Intensity Ratio (MRI-based fibrosis marker)	Fibrosis and electrical scarring were more prevalent in high-TAWSS regions, while low-TAWSS areas were associated with blood stagnation but not fibrosis, with left pulmonary veins exhibiting higher TAWSS than right pulmonary veins.

Abbreviations: CFD, computational fluid dynamics; CT, computed tomography; ECAP, endothelial cell activation potential; HRV, heart rate variability; KE, kinetic energy; LA, left atrium; LAA, left atrial appendage; OSI, oscillatory shear index; PAF, paroxysmal atrial fibrillation; PRT, particle residence time; RRT, relative residence time; RTD, residence time distribution; TAWSS, time-averaged wall shear stress.

).

4.1. Blood Flow Stasis

As previously mentioned, AF is defined as a chaotic and irregular heart rhythm, where the atria are ineffectively contracting. Coupled with the loss of atrial kick, AF patients with disorganized atrial contraction often experience an increase in stasis or stagnant blood in the LA and LAA, as the atrium cannot effectively pump out all the blood into the ventricle within a cardiac cycle [60]. Physiologically, in states of increased blood flow stasis, there is a high risk of stagnant blood pooling to form a clot—as outlined by Virchow’s Triad of thrombosis, where stasis is one of the three main pillars. This clot can then travel through systemic circulation to ultimately cause more worrisome complications such as cerebral infarct or acute ischemia. Considering these devastating effects, 4D flow MRI has facilitated the quantification of atrial blood flow stasis in the context of AF, as seen in Figure 4 [14,38,39,40,42,46,47,48,49,50,51,52]. In the literature, blood flow stasis is often defined as the percentage of atrial blood flow velocities less than 0.1 m/s–0.2 m/s, varying by study type and dependent on study parameters [14,38,39,40,42,46,47,48,49,50,51]. From a prognostic perspective, blood flow stasis is a notable indicator of future stroke risk [14,38,39,40,42,46,47,48,49,50,51]. Specifically, many studies have found that patients with AF have significantly higher blood flow stasis as compared to healthy controls [14,38,39,40,42,46,47,48,49,50,51]. While percentage of blood flow velocities is commonly used to quantify stasis, the residence time distribution time constant (RTDtc) can also quantify stasis, as this marker measures how long blood particles reside in the LA during a cardiac cycle. Similar findings have been elucidated using the RTDtc value, as patients with AF had greater stasis in the form of higher RTDtc compared to healthy controls [46].

Congruent with blood flow stasis, studies have also demonstrated that blood flow velocities in the LA—mean, median, and peak—are all significantly decreased in patients with AF compared to healthy controls, which could further contribute to increased pathologic blood flow stasis [14,38,39,40,47,48,49,51,53,54,55]. This relationship has also been associated with future stroke risk, as AF patients with higher blood flow stasis also had higher CHA₂DS₂-VASc scores compared to healthy controls, which is crucial to guide risk stratification and pharmacological therapy [14,39,46,54].

4.2. LA Strain

LA strain is a measure of atrial deformation and contractile function. It has emerged as a sensitive marker of atrial myopathy and may provide insight into stroke risk independent of traditional structural remodeling. Impaired LA strain is reflective of atrial dysfunction, which can promote stasis and thrombogenesis even in the absence of overt AF. In a study by Demirkiran et al. (2021), LA strain did not differ significantly between paroxysmal AF patients and controls, despite notable impairments in LA flow dynamics (e.g., reduced velocities, increased stasis) [47]. This suggests that flow abnormalities may precede detectable LA mechanical dysfunction in early AF, potentially contributing to thromboembolic risk before structural remodeling becomes apparent. Further research is needed to clarify whether LA strain, in conjunction with other markers such as flow velocity, can refine stroke risk stratification in AF.

4.3. Vorticity

Vorticity, specifically in the LA, is another important quantifiable hemodynamic marker in patients with AF. The vortex is a type of blood flow pattern, initially originating in the pulmonary vein, that optimizes energy transfer between atrium and ventricle, and vorticity can be defined as the strength of this vortex [50]. Given that blood flow in the vortex pattern provides optimal energy transfer during the cardiac cycle, vorticity and overall vortex volume, as detected by 4D flow MRI, can provide insights into blood flow characteristics. Multiple studies have found that vortex volume or overall vorticity decreases in patients with AF compared to healthy controls, indicating impaired left atrial flow dynamics [48,55]. Decreased vorticity and overall impaired flow dynamics in the LA could also contribute to increased thrombogenesis due to greater stasis in the LA—further providing mechanistic evidence for the potential of pathologic thrombosis events in patients with AF.

Interestingly, however; Garcia et al. (2020) found that vortex volume was increased in patients with PAF compared to healthy controls [14]. Early-stage LA remodeling, commonly occurring during PAF [61], could potentially explain this contrasting finding as patients with PAF have not developed chronic AF yet, and the increase in vortex volume is instead due to remodeling of the LA. While vorticity could provide insight into the hemodynamic status of the LA, currently there are various methods used to measure vortex volume [14,48,55], resulting in a lack of direct comparability between studies. Future studies can look to employ a consistent vorticity metric to facilitate understanding of this hemodynamic marker and enable comparison of vorticity values across studies.

4.4. Wall Shear Stress

Wall shear stress (WSS) is defined as the frictional force exerted by blood upon the endocardial surfaces, and aberrant WSS has been associated with AF pathophysiology and its complications. Namely, excessive WSS may contribute to fibrosis of the heart, by triggering endothelial cell signaling cascades [56]. As discussed previously, fibrotic remodeling of the LA contributes to AF pathophysiology [6], as well as increased stroke risk thereafter [30,62]. Therefore, identifying excessive WSS using 4D flow MRI may serve as a valuable screening tool to prevent the development of fibrosis and subsequently AF.

However, low WSS has also been implicated in AF. Namely, low WSS in the LAA in particular may indicate stagnant blood flow therein, resulting in clot formation and subsequently contributing to increased stroke risk [57,63]. This suggests that WSS may have a spatially specific role in the context of AF, wherein excessive WSS in the LA contributes to fibrosis [56] and subsequently AF [6], while low WSS in the LAA may be diagnostic for thrombosis [63]. The precise temporospatial resolution of 4D flow MRI renders it suitable for detection of WSS patterns across areas of the heart. However, care must be taken when estimating WSS, as computational fluid dynamics (CFD) analyses have demonstrated that WSS values vary considerably depending on whether rigid-wall or moving-wall boundary assumptions are applied [58]. Given the multifaceted role of WSS in AF, it is important for future studies to further elucidate this relationship, to better leverage this hemodynamic marker for the prevention and treatment of AF and stroke complications.

4.5. Oscillatory Shear Index

Oscillatory shear index (OSI) quantifies the directional instability of endothelial shear stress, with high values (>0.3–0.4) indicating bidirectional or reversing flow, a condition known to promote endothelial dysfunction, inflammation, and thrombogenesis [58]. In AF, abnormal OSI patterns, particularly within the LAA, are hypothesized to be critical hemodynamic markers of stroke risk that may occur independently of structural remodeling. Parker et al. (2025) utilized a patient-specific moving-wall CFD model and demonstrated pathologically high OSI and low time-averaged WSS in the LAA, which are hemodynamic patterns strongly associated with thrombogenesis [57]. Their analysis further identified a correlation between LAA volume and OSI (R² = 0.70), suggesting structural changes can amplify oscillatory flow [57]. Further, Kjeldsberg et al. (2024) demonstrated that rigid-wall CFD simulations produced unphysiologically low OSI values (0.1–0.34) in the LAA, whereas dynamic moving-wall models generated values (0.3–0.4) consistent with clinical ranges previously linked to stroke/transient ischemic attack (TIA) risk [58]. Collectively, these studies suggest that accurate OSI quantification requires dynamic wall modeling and that pathologically elevated OSI in the LAA may be a plausible hemodynamic mechanism underlying thrombus formation in AF.

5. Discussion

5.1. Gaps in the Literature and Future Directions

As mentioned above, WSS and OSI are hemodynamic markers with a promising diagnostic role in the context of AF. However, there is a paucity of literature comprehensively characterizing WSS and OSI in AF using 4D flow MRI. Existing studies using this methodology present with small sample sizes [56,57], or examine different imaging modalities than 4D flow MRI, namely 4D cardiac CT [58]. Therefore, future studies should build upon these preliminary characterizations of WSS and OSI as diagnostic hemodynamic markers in AF with larger, more representative patient samples.

Another potential marker to be investigated in the future is the endothelial cell activation potential (ECAP). ECAP is an experimental marker calculated as the ratio of WSS to OSI [62]. Combining these two measurements, WSS and OSI, allows clinicians and researchers to determine overall endothelial dysfunction risk, as the ratio provides insights into the magnitude and directionality of shear stress into one singular value. Similarly to WSS and OSI, there is limited literature on ECAP in the context of patients with AF and in relation to other hemodynamic markers specifically using 4D flow MRI. Future studies can analyze ECAP on a larger sample size of patients while also examining the different subtypes of AF to facilitate a more comprehensive understanding of the utility of ECAP, especially in non-traditional patient settings.

In addition to the aforementioned hemodynamic markers, the shape of the LAA and the fat volume of the LA could also be investigated. Currently, LAA morphology is hypothesized to contribute to increased stasis and eventual pathologic thrombosis [34,35]; however, the exact morphology subtype and its mechanistic role remain unclear. Therefore, future studies could utilize 4D flow MRI to isolate which LAA morphology contributes the most to stasis and pathologic thrombosis. Moreover, the fat volume of the LA is significantly associated with AF, as patients with greater LA fat volume had a significant increase in the prevalence of AF [64]. Given this finding, future studies should investigate LA fat volume in relation to AF, hemodynamic markers, and downstream stroke risk. Establishing a causal and predictive relationship between LA fat volume and stroke risk in AF could directly inform prognostic and therapeutic strategies.

5.2. Limitations

Although this review investigated articles pertinent to using 4D flow MRI to investigate the hemodynamics of AF, there are a few limitations. While much of the review was focused on MRI in the context of AF, there are limitations regarding its availability and utility. Namely, MRI can be difficult to acquire, especially for conditions such as AF where it is not directly indicated. Similarly, 4D flow MRI can be even more challenging to routinely implement in standard care practices given its relative novelty. Moreover, with the irregularly irregular rhythm seen in AF, performing cardiac MRI can be challenging, often requiring additional gating techniques and longer acquisition times.

Given that this was a narrative review, there was no assessment of the quality of literature and no formal systematic process for screening literature, potentially resulting in selection bias. To overcome study design limitations and conduct a more comprehensive review of all imaging modalities for investigating AF, and their pertinent nuances, future studies could conduct a systematic review with a broad analysis scope.

6. Conclusions

Overall, this review summarizes current knowledge of advanced imaging in the context of AF. Given that AF has numerous complications—namely ischemia and infarct—it is crucial to gain a better understanding of the disease and risk stratify patients to optimize management. Thus, this paper highlights the importance of integrating advanced imaging into AF management, leveraging hemodynamic markers from current practice to emerging techniques such as 4D flow MRI. Through advanced imaging, our current understanding of LA hemodynamic markers such as blood flow stasis, vorticity, strain, and fibrosis, may improve predictive stroke risk and ultimately the management of AF. Further clinical research is needed to fully synthesize the importance of WSS, OSI, ECAP, morphology of LAA, and fat volume around the LA, which could further help increase our predictive ability of future thrombogenic events.