Clot Composition and Pre-Interventional Radiological Characterization for Better Prognosis and Potential Choice of Treatment in Acute Ischemic Strokes

Abstract

Acute ischemic stroke (AIS) remains a critical concern in clinical practice, with significant implications for patient outcomes and healthcare costs. This review highlights the role of clot composition in AIS, emphasizing the clinical relevance of radiological characterization. Variations in thrombus composition, such as red blood cell (RBC)-rich and white blood cell (WBC)-dominant clots, influence the success of thrombolytic therapies and mechanical thrombectomy. Advanced radiological techniques, including non-contrast CT, CT angiography, and MRI, are essential for pre-interventional clot characterization, guiding optimal treatment decisions. Integrating artificial intelligence (AI) in radiology can enhance the precision of clot composition assessment, facilitating personalized treatment approaches and improving predictive accuracy. By combining histopathological insights with imaging and AI technologies, this review underscores the importance of comprehensive radiological evaluation in the management of AIS, ultimately aiming to enhance clinical outcomes and reduce the burden on healthcare systems.

1. Introduction

Acute ischemic stroke (AIS) stands as a major public health challenge, representing a leading cause of disability and mortality globally. In 2020, the global prevalence of AIS was estimated to be approximately 68 million, with a mortality of 3.48 million.

In the USA, for example, AIS accounted for more than USD 36.7 billion in stroke-related costs in 2015 and is predicted to increase to USD 94.3 billion by 2035. Of the stroke-related expenses, 87% are attributed to AIS [1].

AIS is characterized by the sudden loss of blood circulation to an area of the brain, typically due to a thrombotic occlusion of a cerebral artery, and results in a corresponding loss of neurological function. The initial approach to AIS hinges on prompt diagnosis and rapid recanalization. Time to recanalization is critical in minimizing the expansion of the irreversible necrotic core into the penumbra of the occluded artery as well as patient disability and healthcare costs [2,3].

The escalating prevalence and economic impact of AIS highlight the urgency for effective treatment strategies. Clot composition, although currently not completely applicable to clinical consideration, seems to play a key role in treatment efficacy and patient outcomes [4]. Much of the recent literature highlights how variations in clot makeup, such as fibrin-rich or red-blood-cell-dominant clots, influence the success of treatments like thrombolytic therapy and mechanical thrombectomy. Accurate radiological identification of clot composition can help choose an optimal treatment plan and, therefore, has the potential to decrease recanalization time, ameliorate patient outcomes, and, subsequently, reduce healthcare costs [5].

This review focuses on the role of imaging in characterizing clot composition in acute ischemic stroke (AIS) and its association with therapeutic outcomes. It examines pre-treatment imaging techniques used to evaluate clot properties and potentially guide treatment decisions. Additionally, the review explores various clot compositions, methods of characterization, and emerging technologies that hold the potential to enhance imaging-based analysis and clinical management of AIS.

1.1. Imaging in Acute Ischemic Stroke

In an acute assessment of ischemic stroke, the diagnostic process begins with a comprehensive neurological evaluation using the national institutes of health stroke scale (NIHSS). This scale, through a systematic assessment of neurological function, quantifies stroke severity, informing the neuroradiologist on the imaging approach. Imaging is recommended in the first 25 min of patient presentations to maximize patient outcomes [6]. Currently, the main role of imaging is to distinguish the ischemic stroke from its hemorrhagic counterpart, localize the thrombus, identify the affected vessel and its caliber, quantify the size of the penumbra, and, based on the results, choose a correct treatment plan [7].

1.2. Current Role of CT Imaging in Acute Ischemic Stroke

Non-contrast computed tomography (NECT) is typically the initial imaging modality in the case of neurological deficit due to its accessibility and cost-effectiveness [8]. Its primary role is to differentiate ischemic strokes from hemorrhagic causes [9]. Although the sensitivity in detecting AIS in the first 3 h of NCCT is modest, early onset signs of AIS, such as gray matter hypodensity and effacement of sulci, can potentially be observed [7]. The Alberta stroke program early CT score (ASPECTS) is then used to quantify the severity of ischemic changes, predict functional outcomes, and help decide which treatment modality to pursue [10,11]. NCCT of an AIS can also present with the hyperdense artery sign (HAS), a key radiological finding on NCCT that can suggest the presence of a large vessel occlusion (LVO) [12]. The nature of HAS is also argued to be a sign of a positive prognosis for the patient [13,14].

CT angiography (CTA) enhances vascular imaging, offering precise thrombus localization and insights into the at-risk tissue and infarct core. CTA facilitates the assessment of clot mechanical properties, including perviousness, indicated by thrombus attenuation increase (TAI), which correlates with improved functional outcomes, increased recanalization rates, and reduced necrotic tissue volume [15].

Computed tomography perfusion (CTP) complements these modalities, proving especially useful in the extended thrombolysis window (6–24 h) by identifying candidates for treatment despite late presentation [16].

1.3. Current Role of MRI in Acute Ischemic Stroke

MRI, despite being less accessible and cost-effective than NCCT, excels in AIS diagnosis due to its high sensitivity [17]. It employs various sequences, particularly DWI, renowned for its rapid and specific detection of acute infarction through signal intensity changes and appearance diffusion coefficient (ADC) value alteration [18]. The latter can be combined with MR perfusion (MRP) to assess collateral blood flow and enable the differentiation between potentially salvageable hypoperfused tissue and regions likely suffering irreversible damage. Complementarily, MRP evaluates cerebral blood flow and tissue viability, whereas MRA visualizes occlusions in the cerebral vasculature. FLAIR and GRE further assist in determining the age of infarctions and detecting hemorrhages, respectively, enhancing the comprehensive diagnostic capabilities of MRI in AIS management [18,19].

Although there are many sequences used for the diagnosis and management of AIS, the T2 or susceptibility weighted imaging enhances AIS diagnosis by directly visualizing clots through the susceptibility vessel sign (SVS) and distinguishing the clot from brain parenchyma and vascular blood flow [20,21]. SVS can also be a prognostic sign; for example, Bourcier et al. [22] showed that the presence of SVS favors a better clinical outcome when treated with a thrombectomy [22]. This association has been confirmed by Belachew et al., who performed a retrospective study on 577 patients and found the SVS to have a higher rate of successful reperfusion (TICI score of >2b), increased functional independence, and a lower mortality rate [23].

1.4. Current AIS Management

Once the AIS patient has been clinically evaluated by the NIHSS and has undergone a form of imagery to confirm the AIS diagnosis, the patient can begin treatment. The initial step, in the absence of contraindications, such as the presence of active bleeding, severe hypertension, and coagulopathies, is to begin a thrombolytic treatment using intravenous alteplase [24]. A CTA can then be performed to assess the efficiency of the thrombolytics and the residual clot. In the patients in which recanalization is not achieved with the thrombolytic treatment, a thrombectomy is performed. There are two main categories of thrombectomy devices: direct aspiration and stent retrievers. These two can be used in combination, if necessary, along with a balloon-guided catheter (BGC) to prevent thrombus fragments from traveling with the blood flow and embolizing distal arteries. Currently, the choice of the device is decided based on the artery size, location, and tortuosity in which the clot is located and is aimed at achieving a 2b/3 angiographic result on the thrombolysis in cerebral infarction (TICI) scale [25,26].

Although current literature is still somewhat contradictory, multiple in vivo and ex vivo studies have further shown that clot composition also has a direct influence on the interactions between the thrombectomy device and the clot [27,28]. Thus, the characterization of the clot before intervention has the potential to have a significant influence on the choice of thrombectomy technique.

Once the procedure is performed, the patient is once again evaluated based on the NIHSS scale after about 24 h and at 90 days using the modified Rankin score [3].

1.5. Clot Compositions in Acute Ischemic Stroke

Acute ischemic stroke (AIS) thrombi can be characterized by a variety of features, including shape, size, consistency, composition, and morphology [29]. One approach to classifying clots involves analyzing their cellular and proteomic components, leading to the categorization into three subtypes: red clots, white clots, and aged or calcified clots. Red clots, also recognized as red blood cell (RBC)-rich or soft clots, predominantly consist of erythrocytes and are characterized by a loose fibrin network. Conversely, white clots, or hard clots, are defined by a high concentration of platelets, neutrophils, neutrophil extracellular traps (NETs), and a dense fibrin matrix. Aged or calcified clots, though less common, are distinguished by the presence of calcifications, often in conjunction with atherosclerotic components such as cholesterol crystals [3].

However, the categorization of clots into merely three types—red, white, and calcified—somewhat oversimplifies the complexity and structural heterogeneity observed in AIS thrombi. This is further underscored by instances of thrombi that feature a blend of compositions, manifesting the limitations of such a classification system. Notably, the phenomenon of layered thrombi, showcasing alternating stratifications of RBCs and fibrin, along with serpentine clots characterized by their distinctive winding patterns of these components, exemplifies the structural variations that exist [30]. These structural and compositional differences are believed by some to result from hemodynamic changes that affect the organization of RBCs and fibrin within the thrombus during its formation [29,31] For instance, when wall shear rates and blood flow velocity are elevated, fibrin aligns with the direction of blood flow, consequently increasing its stiffness [32].

2. Methods of Clot Characterization

2.1. Histological Methods

The classical method for determining clot composition is through hematoxylin and eosin (H&E) staining. This technique facilitates the visualization of various cell types and some extracellular components, allowing for the classification of the clot as red, white, or calcified. It is important to note, however, that even within each class of clots, there is a large variability in composition [4,33].

Additionally, other staining techniques, such as Martius scarlet blue (MSB), are employed to enhance the contrast of fibrin, enabling a more reliable visualization of its distribution [34]. Immunohistochemical staining is another pivotal tool in identifying the cellular, proteomic, and molecular composition of clots. For example, staining against CD42b, a glycoprotein present on platelets, is utilized to highlight platelet-dense regions [35]. A study using CD42B staining, for example, visualized platelets at the periphery of RBC-rich clots delineated by fibrin layers [30].

Similarly, other staining methods have been effective in delineating the von Willebrand factor (vWf), collagen, elastic fibers, and calcifications, thereby enriching the understanding of clot composition and its implications for AIS [30,36,37].

Two-dimensional histological analysis, while valuable, has inherent limitations due to its reliance on thin tissue sections from heterogeneous clots, failing to capture their full compositional complexity. This approach provides limited insights into the clot’s volumetric and structural properties, potentially leading to misclassification and inaccurate correlations with therapeutic strategies. Addressing the limitations of 2D histological analysis, Somayeh Sghamanesh et al. have utilized microtomography (micro-CT) to delve into the 3D structures of clots removed via mechanical thrombectomy. The team was able to render 3D images, revealing the cellular and fibrillary arrangements within clots and visualizing the porosity distribution and density across different clot regions [38]. Furthermore, signals obtained from microCT are also hypothesized to correlate with signals perceived in clinical CT imaging, especially clots with a high RBC content [39,40]. With further investigations, the latter may have crucial implications for correlating clinical radiomic signatures to different clot compositions.

2.2. Clot Characterization Through Scanning Electron Microscopy

Another modality used to characterize AIS thrombi is through scanning electron microscopy (SEM). SEM enables a detailed visualization of the three-dimensional ultrastructure of clots, revealing intricate details about the cellular and fibrillary components [41]. High-resolution SEM analysis has demonstrated the ability to distinguish various structural features within AIS clots, such as specific organizations of RBCs, fibrin, and platelets, allowing the characterization of compact cores and porous peripheries [42]. SEM further plays a crucial role in identifying specific markers within the clots, such as polyhedrocites, deformed red blood cells that can indicate the age of a clot, as well as the precise arrangement of fibrin fibers and the distribution of platelets and white blood cells [40,41]. One study, for example, by using SEM, successfully visualized the presence of a compressed polyhedrocytes along with a dense network of fibrin. The research found a significant association between the latter and thrombolysis-resistant clots [43].

2.3. Limitations of Clot Characterization

Both histological and imaging methods for clot characterization face limitations. Fixation techniques, such as formalin or ethanol, can alter clot properties, with over- or underfixation affecting the preservation of delicate structures like NETs. Thrombi are often retrieved in fragments or damaged during thrombectomies, compromising structural integrity and making it difficult to analyze the original compositions [29,30,33,44]. Histological sectioning provides only two-dimensional slices, failing to capture three-dimensional clot heterogeneity, while imaging methods like micro-CT, despite offering 3D visualization, may miss fine details due to voxel size. SEM is further constrained by its two-dimensional nature and preparation requirements, such as dehydration and coating, which can distort clot morphology [40]. Handling artifacts, storage conditions, and variability in protocols across studies further exacerbate the challenges of finding a reliable consensus.

2.4. Influence of Clot Composition on Thrombolysis Outcomes

With the current AIS management protocol, RBC-rich clots are associated with a higher recanalization rate and a lower NIHSS score, while clots rich in WBCs are associated with the opposite [4].

Although further investigation is required, some studies show that RBC-rich clots have a more favorable response to tPA treatments and are often attributed to the loose fibrin networks accessible to plasminogen [4]. In contrast, clots rich in platelets, vWF, NETs, and WBCs are much less prone to tPA degradation due to their dense fibrin network and the presence of adhesive vWF and extracellular DNA typically derived from NETs [44]. These extracellular traps are key components of the thrombus formation and a major factor contributing to thrombolytic treatment [45]. Deoxyribonuclease 1 (DNase 1), an enzyme that degrades DNA, could be a potentially significant contributor to thrombolytic therapy. Recent studies on ex-vivo models have found a significant acceleration of tPA thrombolysis when used in combination with DNase 1 [46].

Research has also demonstrated that the von Willebrand factor (vWF) is crucial for platelet adhesion to damaged endothelial cells. It directly interacts with extracellular DNA released by leukocytes, facilitating the attachment of leukocytes to endothelial cells making them more resistant to thrombolysis [47]. Thus, another potential contributor to thrombolytic therapy is ADAMTS13, a metalloprotease that cleaves large vWF multimers into less coagulable, smaller proteins [48]. Prochazka et al. [47] performed a study on 131 AIS patients and found that the vWF:ADAMTS13 ratio within the thrombi was significantly higher in patients with worse outcomes. Perhaps, restoring the balance between vWF and ADAMTS13 may play a beneficial role in the outcome.

Nevertheless, other studies differ on the impact of clot composition on therapy outcomes: while some point to the protective role of platelet-rich shell structures against thrombolysis, others note that tPA reduces clot size across all components proportionately [27,49,50,51]. For instance, Di Meglio’s study [50] found that AIS thrombi possess a dense outer shell composed of fibrin, von Willebrand factor, and platelets. In contrast, the core is characterized by clearly identifiable red blood cells (RBCs), fibrin fibers, and aggregated platelets. The in vitro experiment on 199 thrombi showed that the outer shell is less susceptible to thrombolysis than the inner core [50].

Rossi et al. [51], on the other hand, conducted a study on more than 1000 samples in which they compared the size and composition of clots in AIS victims that underwent either a tPA treatment along with mechanical thrombectomy (MT) or MT alone. The study, while confirming the core–shell model, found that the extracted clots that underwent tPA were significantly smaller, yet their composition was not significantly altered. Another similar retrospective study on 1430 patients came to the same conclusion [49]. Although the clot’s overall susceptibility to a thrombolytic therapy is promising, there is a potentially greater risk for distal embolization, making it more difficult to attain a successful extraction by MT [52]. Indeed, Mohammaden et al. [53] found that distal embolization was significantly correlated with the tPA dose administrated pre-thrombectomy. The core–shell morphology model was further confirmed in a study by Khismatullin et al. [54], in which 41 cerebral thrombi were extracted and analyzed with SEM. Khismatullin et al. [54], however, also noted a direct correlation between increased stroke severity and platelet aggregates and polyhedrocyte content within the core. In contrast, thrombi with alternating layers of RBCs, platelets, and fibrin had more favorable outcomes. These studies may indicate that the susceptibility of a thrombus to thrombolysis is determined not merely by its composition but, more significantly, by the structural arrangements of its components.

2.5. Influence of Clot Composition on Thrombectomy Outcomes

In the context of a thrombectomy treatment, physical and mechanical properties dictated by the clot composition are shown to have a significant influence on success rates and complications [5].

RBC-rich clots have been strongly associated with higher recanalization success rates and lower severity scores. Clots with a lower RBC content have been reported to be more difficult to retrieve and require a longer procedure time [5,55]. WBC content is also correlated with the NIHSS score at the time of discharge and the modified Rankin score at 90 days [4]. Some studies associate this increased severity and unfavorable outcomes of WBC-rich clots, in part, to the NET concentration. Kaesmacher J. et al. investigated the correlation between NET content and the intervention outcome and found that NET-containing clots had a lower rate of recanalization and an increased risk of distal embolization [56]. These results can be attributed to NETs’ role in modifying the physical and mechanical proprieties of the clot and the effect it has on the interactions between the clot and the thrombectomy device [57]. Staessens et al. [57] showed that NETs can modify the fibrin structure and play a key role in resisting mechanical force.

Nevertheless, other physical and mechanical properties such as elasticity, viscosity, and the coefficient of friction also play a major role in thrombectomy success.

The elasticity of the clot is argued to be greatly influenced by RBC percentage. Gersh et al. [58] measured the elasticity of clots using a strain-controlled dynamic time sweep and found that clots containing more than 20% RBCs were more deformable compared to thrombi with lower RBC concentrations. The viscosity was also found to be higher in RBC-rich clots [58].

These properties in RBC-rich thrombi have been shown to influence their interaction with the thrombectomy device. Machi et al. [59], using clot analogs in a silicon vascular model, found that permeating the clot and centralizing it in the lumen of the stent retriever was more achievable among red clots. In contrast, white clots were found to remain in between the stent retriever and the vessel wall, making it harder to capture and extract as the device would slide past it. The authors further hypothesized that a constant radial force from the stent on the vessel could potentially improve the results.

In white clots, fibrin density is known to significantly increase the coefficient of friction compared to RBC-rich clots [60]. In turn, a higher coefficient of friction means that more force is needed to retrieve the thrombi, making it more challenging. This increases the risk of fragmentation of the thrombi which can lead to distal embolization and incomplete recanalization [61]. Dense fibrin networks also increase the stiffness and compressibility of the clot, making it harder to integrate it within the stent retriever and, thus, retrieve it [25,60].

The age of the clot is also shown to influence its mechanical properties. A study characterizing the mechanical behavior of various thrombi concluded that more mature clots are less deformable and have strong adhesion to the vascular bed [62]. Such clots are harder to fully retrieve, whether with the stent retriever or through aspiration, and may increase the risk of damaging the blood vessel [63].

Currently, there is no clear approach to optimizing recanalization when faced with the heterogeneous nature of AIS clots. Van der Marel et al. [64], however, argue that increasing the integration time to 5 min can significantly increase the clot volume captured in the lumen of the stent. The result showed that this technique had little effect on the integration of analogs of red or soft clots, while a significant integration increase in analogs of cholesterol-rich clots was observed. K.J. Wagner et al. [65], on the other hand, suggest using a longer stent structure with larger gaps to facilitate the centralization of the clot within the device (

Table 1. Clot composition.

Clot Subtype	Cellular Components	Reperfusion	Correlation with Prognosis
RBC-rich	High erythrocyte (RBC) content, loose fibrin network	Higher rates with both thrombolysis and mechanical thrombectomy	Favorable prognosis: higher recanalization rates, lower NIHSS scores
Platelet/WBC-rich	High platelet, neutrophil, and neutrophil extracellular traps (NETs) count and dense fibrin matrix	Lower rates, more challenging to retrieve; dense fibrin resists mechanical force	Poorer prognosis: associated with severe outcomes, higher NIHSS, and mRS scores
Calcified/aged clots	Calcifications, cholesterol crystals, atherosclerotic components	Lower rates due to strong adhesion to vascular walls and stiffness	Poorer prognosis: difficult to treat and increases the risk of vessel damage

).

3. Radiological Signs for Clot Characterization

3.1. CT

Most studies analyzing the radiological signs of AIS clot characteristics use CT. One of the most prevalent signs giving insights into clot composition is the hyperdense artery (HAS). It is clinically used for the detection and localization of the thrombi [10]. Current literature also suggests a strong association with RBC-rich clots. On the other hand, an iso-dense artery sign indicates a high fibrin and platelet content and fewer RBCs [34,66]. According to the previously cited literature, HAS is associated with better patient outcomes and lower severity scores, while the absence of HAS is associated with the opposite [13,34]. The higher RBC content, depicted by HAS, is more responsive to thrombolytic therapy, whether administered intravenously or intra-arterially [27]. In theory, standard thrombectomy treatment is also favored due to the soft, deformable, and less adhesive nature of RBC-rich clots [58]. This theory was strengthened by Maekawa et al. [55] by correlating HAS to histological composition and thrombectomy outcome. HAS was indeed associated with higher RBC percentages. The number of maneuvers and time to recanalization were also reduced.

Absolute density in Hounsfield units (HU) is also a useful indicator of clot composition. Hund et al. [67] used thin-slice imaging (≤2.5 mm) to examine the quantitative characteristics of clots in a sub-sample of 94 patients. Using hematoxylin–eosin staining, they quantified RBC% and correlated it to HU values. They concluded that the absolute density of the thrombi was the best predictor of RBC%. Interestingly, Hund et al. found that the relative density of the thrombi to the contralateral artery had little to no association with the composition. They further attributed this finding to a greater interobserver variability.

A growing amount of evidence suggests that CT angiography can be used to characterize the clot itself, notably by measuring the residual blood flow in the occluded area. This residual blood flow is known as the thrombus attenuation increase (TAI) on the CTA and is indicative of the perviousness or permeability of the thrombus. The permeability of the clot has been shown to correlate with the RBC content, while reduced values of TAI generally indicate a low RBC and a high WBC content [68,69]. Consequently, another study found that TAI positively correlates with higher recanalization rates and better functional outcomes. Lower TAI values were correlated with worse functional outcomes and a higher risk of distal embolization during mechanical thrombectomy [15,70,71].

It is worth noting, however, that whatever the composition of the clot may be, the residual blood flow on its own can play a protective role against irreversible damage. Furthermore, other clot characteristics such as length and volume can also influence TAI and, thus, the therapeutic outcome of both thrombolysis and thrombectomy. Thus, the favorable outcomes associated with increased perviousness should not be solely attributed to the clot composition [71,72].

The thrombus enhancement (TE) sign can also be used to provide insights into clot composition. This sign reflects contrast buildup within the clot and is associated with increased fibrin and platelet and reduced RBC proportions [73]. Moreover, Guangchen et al. found that the absence of TE is significantly associated with successful first-pass thrombectomies [74].

3.2. MRI

One of the most clinically relevant MRI signs giving insights into clot composition is the susceptibility vessel sign (SVS), predominantly observed using T2-weighted gradient-recalled echo (GRE) imaging. This sign is characterized by a hypointense signal along the vessel walls which may alter the clarity of their margins. Recent studies have established a correlation between the presence of SVS and an increased red blood cell (RBC) content within clots, whereas its absence has been associated with a predominance of fibrin, as determined through histochemical staining techniques [28,75,76,77].

Bourcier et al. [78] further contributed to this understanding by employing a qualitative methodology to assess the macroscopic characteristics of the clots, determining that SVS is a reliable indicator of clots with a red–black appearance which typically correlate with a higher RBC content [78].

Conversely, research conducted by Horie et al. [79] did not find a significant relationship between SVS presence and RBC content. The authors further cast doubts on the histological characterization of thrombi due to potential damage and fragmentation of the samples post extraction. There was also an absence of details regarding the MRI sequences employed in their analysis [79].

Meglio et al. [80], similarly, found no correlation between SVS and RBC content when analyzing the RBC content with hemoglobin-linked immunosorbent assays [80].

However, the relevance of SVS has been identified as a predictor of improved outcomes following recanalization treatments. This is supported by the findings of Bourcier et al. [22] in a retrospective study linking the presence of SVS to favorable functional outcomes, specifically highlighting its ability to predict a modified Rankin scale (mRS) score of less than or equal to 2 three months post treatment [22].

Several other studies have explored the relationship between MRI signal intensities and the composition of thrombi. For instance, Fujimoto et al. successfully differentiated between thrombi enriched in red blood cells (RBCs) and fibrin-rich thrombi through signal intensities in controlled laboratory settings and a porcine model [81]. In their study, they utilized a high-field MRI scanner to perform detailed imaging of thrombi created in vitro and subsequently implanted into a porcine cerebral artery model. The results demonstrated distinct differences in MRI signal intensities between the two types of thrombi. RBC-rich thrombi exhibited higher signal intensities on T1-weighted images and lower intensities on T2-weighted images, attributable to the paramagnetic properties of deoxygenated hemoglobin. In contrast, fibrin-rich thrombi showed relatively homogeneous signal intensities on both T1- and T2-weighted images, reflecting the dense fibrin network’s different magnetic susceptibility. Fujimoto et al. also performed histological validation, confirming that the MRI signal variations corresponded accurately to the actual clot compositions observed under microscopic examination [81].

Janot et al. [82] further explored the diagnostic capabilities of MRI imaging by examining the signal intensity ratios derived from susceptibility-weighted imaging and T2-weighted gradient echo imaging, noting a negative correlation with RBC concentration. However, this method proved ineffective in clearly distinguishing between thrombi with RBC concentrations higher than 54% or lower than 23%. To overcome this limitation, Janot et al. suggested combining multiple MRI sequences, including T2-weighted gradient echo, SWI, and fluid-attenuated inversion recovery (FLAIR), to achieve a more accurate quantification of RBC concentrations within thrombi. The study highlighted that, while SWI and T2-weighted imaging alone could provide some insights into thrombus composition, a multimodal approach using various MRI sequences could offer a more comprehensive and accurate assessment [82].

It has been demonstrated that employing R2*, quantitative susceptibility mapping (QSM), and proton density fat fraction (FF) maps from a single gradient echo MRI scan can effectively characterize thrombi based on hematocrit levels, as well as identifying calcified and lipid-rich sections within clots. This method allows for distinguishing acute from chronic clots by analyzing changes in R2* and QSM values over time. Additionally, QSM and FF maps can differentiate between thrombi components, providing a detailed assessment that aids in tailoring treatment strategies for acute ischemic stroke [83].

Bretzner et al. further supported this by using 7T GRE MRI and R2* relaxometry, finding a positive association between R2* values and the concentrations of RBCs and iron within thrombi. Their method involved correlating MRI findings with absorption spectrometry and histological staining, confirming that higher R2* values indicate greater RBC and iron content [84].

Quantitative susceptibility mapping (QSM) is an emerging MRI technique that measures magnetic susceptibility in tissues, allowing for a noninvasive assessment of changes in iron and myelin content in ischemic stroke lesions. A study involving 32 patients found that longitudinal increases in magnetic susceptibility values were significantly associated with worse neurological outcomes, as measured by the national institutes of health stroke scale (coefficient: 0.311, 95% CI: 0.098–0.520, p = 0.017). This suggests that elevated iron concentrations within ischemic lesions during the restorative phase of stroke are linked to a poorer functional recovery. These findings highlight QSM’s potential as a tool to monitor post-stroke tissue changes and guide rehabilitation strategies, although further validation in larger cohorts is necessary (

Table 2. Imaging features.

Imaging Marker	Clot Properties Correlated	Imaging Mechanism
Hyperdense artery sign (HAS)	High RBC content, soft thrombi	Observed on non-contrast CT; increased density due to RBC concentration
Thrombus attenuation increase (TAI)	High RBC content, increased permeability	Observed on CTA; residual blood flow through thrombus highlights RBC-rich clots
Thrombus enhancement (TE)	High fibrin and platelet content, low RBC content	Observed on CTA; contrast buildup within the thrombus highlights dense fibrin and platelets
Susceptibility vessel sign (SVS)	High RBC content, low fibrin content	Observed on MRI (T2-weighted GRE); hypointense signal reflects paramagnetic properties of deoxygenated hemoglobin
T1-weighted and T2-weighted MRI	Differentiates RBC-rich clots (T1: high intensity, T2: low intensity) from fibrin-rich clots (homogeneous signal)	MRI sequences sensitive to hemoglobin and fibrin properties, providing contrast based on thrombus composition
Quantitative susceptibility mapping (QSM) and R2* mapping	Differentiates thrombi by hematocrit, iron, or lipid content; calcified clots	Derived from MRI; identifies magnetic susceptibility differences due to clot composition
Fluid-attenuated inversion recovery (FLAIR)	Distinguishes recent ischemic damage; not directly correlated with clot composition	MRI technique suppressing fluid signal to highlight ischemic damage and older infarctions
Diffusion-weighted imaging (DWI)	Detects acute infarction and hypoperfused tissue; indirect association with thrombi	High sensitivity to ischemia; indirectly linked to clot effects

) [85].

4. Emerging Technologies

Artificial Intelligence

Artificial intelligence (AI) has emerged as a transformative force in medical imaging, offering unprecedented capabilities in the characterization of clot composition in acute ischemic stroke (AIS). By leveraging machine learning and deep learning algorithms, AI can enhance diagnostic precision, predict treatment outcomes, and facilitate personalized therapeutic strategies.

AI-driven imaging techniques are revolutionizing the way clots are detected and analyzed in medical imaging. For instance, automatic detection and segmentation algorithms have significantly reduced the time required for diagnosis and have improved the accuracy of clot identification. Tools such as MethinksLVO and Brainomix rapidly predict large vessel occlusion (LVO) from non-contrast CT (NCCT) scans. MethinksLVO, developed by Olive-Gadea et al., was used on a large historical database of over 24,000 patients for training and 1453 for validation, achieving a sensitivity of 83% and a specificity of 71%. On the other hand, Weyland et al. evaluated the Brainomix software, which demonstrated a sensitivity of 77% and a specificity of 87% in detecting the hyperdense artery sign (HAS) associated with LVO, performing similarly to expert neuroradiologists [12,86].

Moreover, radiomics—the extraction of quantitative features from medical images—combined with AI, provides deeper insights into clot composition. Hanning et al. (2021) developed a machine-learning algorithm that analyzes imaging data from thin-slice NCCT and CTA. They extracted 4844 radiomic features per thrombus and used these features to predict clot composition, specifically differentiating between RBC-rich and fibrin-rich clots. Their study demonstrated that AI could accurately classify clot composition [87].

The future of AI in clot composition characterization is filled with potential. One key area is the integration of data from multiple imaging modalities, such as CT, MRI, and micro-CT. This multimodal approach provides a holistic view of the clot, enabling more precise characterization and personalized treatment planning. LaGrange et al. (2023) emphasize the importance of integrating conventional clot imaging characteristics with AI models to enhance predictive accuracy and clinical utility. Their systematic review highlights that future AI approaches should consider both imaging features and patient-specific vascular characteristics to optimize model performance [14].

Predictive analytics is another area where AI shows great promise. Qiu et al. (2019) developed a machine-learning model using radiomic features from CT scans to predict early recanalization after intravenous alteplase administration. They found that radiomic features were more predictive of treatment outcomes than traditional clot imaging measures [88].

Furthermore, Hofmeister et al. (2020) utilized a support vector machine classifier to predict first-attempt recanalization success with thromboaspiration based on radiomic features. Their model also predicted the number of thrombectomy passes required for successful recanalization [89].

AI represents a significant advancement in the characterization of clot composition in acute ischemic stroke. By improving diagnostic accuracy, enabling personalized treatment, and optimizing clinical workflows, AI has the potential to revolutionize stroke care. As these technologies continue to advance and become more widely adopted, they will contribute to better clinical outcomes and reduced healthcare costs, marking a new era in the management of stroke patients.

5. Conclusions

The comprehensive analysis of clot composition in acute ischemic stroke (AIS) elucidates the critical impact of thrombus characteristics on diagnostic strategies, therapeutic decisions, and patient outcomes. Variations in clot composition, particularly the proportions of red blood cells (RBCs), white blood cells (WBCs), and fibrin, significantly influence the efficacy of thrombolytic treatments and mechanical thrombectomy. Advanced imaging techniques such as CT and MRI play a role in characterizing these clots pre-intervention, potentially enabling tailored treatment approaches that enhance recanalization success and reduce healthcare costs. Emerging technologies like artificial intelligence (AI) further augment these capabilities, offering predictive analytics and personalized therapeutic strategies. This evolving paradigm underscores the necessity for continuous integration of histopathological insights with imaging and AI methodologies to optimize AIS management, ultimately improving prognosis and functional recovery for stroke patients.