Time-dependent effects of formalin fixation on magnetic resonance imaging relaxation times in ex vivo human thrombus tissue

Introduction

Magnetic resonance imaging (MRI) is a highly sensitive modality capable of detecting subtle structural and biochemical characteristics within different tissue types (1). However, the accuracy and reliability of MRI-derived parameters are highly dependent on uniform sample handling. In research settings, MRI scan is commonly performed on chemically fixed tissue samples, typically in formalin. Formalin fixation preserves the histological integrity of tissues, including blood clots over extended periods, making it a suitable method for storing thrombus samples for further histopathological analysis (2,3). While there are articles that have addressed how this fixation affects MRI relaxation times in various types of tissues—brain, heart, and liver, there is a notable lack of literature investigating the same effects on human thrombus. A number of recent studies have applied MRI to evaluate thrombus composition, primarily using in vivo imaging modalities [e.g., gradient recalled echo (GRE), susceptibility weighted imaging (SWI), or non-contrast computed tomography (CT)/magnetic resonance (MR) vessel signs] to infer thrombus structure and correlate with clinical outcomes (4,5). A smaller subset of studies have performed ex vivo MRI on human thrombi, notably including venous/pulmonary emboli and stroke clots, showing promising results, for example, showing that T1-weighted and susceptibility-based imaging can distinguish red blood cell (RBC)—from platelet-rich regions and relate imaging features to thrombolysis susceptibility (6-9). These studies show us the importance of ex vivo scanning: it enables motion-free, high-resolution imaging with direct histologic correlation, but on the other hand, they often do not report fixation protocols, or if they do, they do not consider that fixation and tissue handling may substantially alter the observed MRI parameters. This methodological oversight may lead to variability or misinterpretation of relaxation values, as fixation-induced changes can mask or mimic biological features of the thrombus (8,9). Given that fixation and dehydration can dramatically alter relaxation times, this omission may introduce variability or lead to misinterpretation of imaging findings. This represents a critical gap, especially given that MRI-based thrombus characterization is of growing concern in stroke and vascular disease studies.

With rising incidence of venous thromboembolism (VTE) [deep vein thrombosis/pulmonary embolism (DVT/PE)] and a growing number of thrombectomy devices entering clinical use, there is a pressing need for imaging biomarkers that can non-invasively characterize thrombus composition and guide treatment strategy (8,10).

Thrombi are fundamentally different from other human tissues. Unlike solid organs with well-defined cellular architecture, clots are semi-solid, highly hydrated structures with variable proportions of RBCs, fibrin, platelets, and inflammatory cells (11-13). They often contain high concentrations of hemoglobin and its degradation products, including methemoglobin and hemosiderin, which are paramagnetic and have a pronounced impact on susceptibility-based MRI measurements such as T2* (1). In addition, clot consistency and structure may be highly variable with the etiology and age of the thrombus (3,14,15). Such conditions render blood clots particularly vulnerable to the structural and chemical changes induced by formalin fixation, with the potential to affect MRI signal properties in ways inapplicable to other tissues. This is particularly relevant for ex vivo imaging studies that aim to correlate MRI characteristics with histopathology to create imaging biomarkers for thrombus composition and response to treatment. Without a clear understanding of how fixation alters the MRI parameters of clot tissue, there is a significant risk of misinterpreting signal changes that are fixation-induced rather than biologically meaningful, which would limit the reliability and translational value of the examination.

This study aims to bridge the literature gap about the effect of formalin fixation on the MRI relaxation properties of human blood clots by comparing fresh and fixed clot samples using ultra-high-resolution 9.4 T MRI scanner. By assessing how the different relaxation time values are influenced by fixation, our study can stand as a reference for future studies involving ex vivo clot imaging, ensuring that fixation-related changes are properly accounted for in the interpretation of the imaging data.

Methods

Patients and sample handling

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Houston Methodist Hospital (PRO00027258) and informed consent was obtained from all individual participants. Between November 2024 and April 2025, 13 patients with DVT of lower and upper extremities (DVT) or PE or arterio-venous graft (AVG) thrombosis underwent mechanical thrombectomy. The removed blood clots underwent 9.4T ex vivo MRI scanning. Seven clots were split into multiple tubes due to their large volume, resulting in 19 sample specimens in total. Mechanical thrombectomy procedures were performed in each case, either through an open surgical approach using a Fogarty embolectomy catheter or via an endovascular technique utilizing the Indigo Aspiration System (Penumbra, Alameda, CA). A total of 15 samples underwent fresh scanning, after which, they were placed in 10% formalin. In order to evaluate the effect of time of formalin fixation 11 of the 15 samples were scanned after being immersed in formalin within 6 hours post-procedure, and 12 samples were scanned following formalin fixation after more than 24 hours of fixation. Four samples did not undergo any post-formalin imaging and were therefore excluded from further statistical analysis.

MRI imaging

Images were acquired using a Bruker 9.4T Advance 400 scanner, running ParaVision 7.0.0 software (Bruker BioSpin, Billerica, MA) using the wide bore MicWB40 RF probe (MicWB40 Body W1 In Vivo). Table 1 shows the acquired parameters of each sequence: T1 maps were acquired using a rapid acquisition with relaxation enhancement (RARE) sequence, and a parametric fit was performed on the scanner using a saturation recovery technique. T2 maps were calculated from the multi-slice multi-echo (MSME), and T2* maps were obtained from a multi-gradient echo (MGE) sequence, both by using a mono-exponential fit. Figure 1 shows T1, T2 and T2* maps of a sample imaged without formalin immersion.

Figure 1 T1, T2, and T2* maps of the thrombus samples (this scan was performed without formalin fixation).

Image segmentation

Image segmentation was performed using OsiriX MD (v14.0.1) program to isolate regions of interest (ROIs) based on intensity thresholds: the Digital Imaging and Communications In Medicine (DICOM) images were imported into the two-dimensional (2D) viewer, where initial ROIs were identified by estimating minimum and maximum intensity values for the clot material. The “Grow Region (Threshold)” tool was then used to segment tissue regions by adjusting these thresholds to highlight the desired structures.

Where automated threshold masking was not feasible, due to wide variation in signal intensities or poor contrast between the clot and surrounding environment (typically formalin), the ROI was manually drawn using the Polygon ROI tool to accurately delineate the tissue.

This process ensured the distinction of the clot components from the surrounding tissue.

Additional refinement was accomplished with the Eraser tool to delete unwanted regions, and the “Invert Selection” option was used when background region exclusion was necessary. Segmented ROIs were transformed into binary masks with the “Convert ROI to Mask” option. This threshold-based masking procedure was implemented throughout each image slice to generate composite volumetric maps for each MRI sequence. The different options or ROI-selection is presented on Figure 2. Finally, the overall mean and pooled standard deviation (SD) values of signal intensity measurements were calculated over the whole three-dimensional (3D) clot volume.

Figure 2 ROI-selection: the first picture shows a thrombus sample in formalin, the second picture shows the selection by threshold masking, in the third picture, the ROI-selection was made by manual drawing. ROI, region of interest.

Further statistical analysis

Statistical analyses were performed to evaluate the effects of formalin fixation and patient-specific covariates on MRI relaxation times (T1, T2, and T2*) in the ex vivo blood clot samples. Since the different samples were collected from the same patient and the same collecting site in each case, the three fixation stages—fresh, formalin <6 hours, and formalin >24 hours—were treated as repeated measures. Because not all samples had complete measurements across all three timepoints, a linear mixed-effects model was employed to perform a repeated-measures analysis of variance (ANOVA). This approach allowed the inclusion of all available data without excluding partially scanned samples, improving statistical power and maintaining internal consistency.

We reported the overall mean and pooled SD values from each clot sample of each relaxation time and fixation stage and assessed statistical significance of differences across fixation stages using a significance threshold of P<0.05. When the overall fixation effect was significant, post-hoc pairwise comparisons were conducted to identify specific differences between timepoints.

To evaluate the influence of clinical and biological factors on relaxation times, continuous variables including clot age (days), patient age (years), and body mass index (BMI) were added as fixed effects in the mixed-effects models. Categorical variables such as sex and clot type were also explored as fixed effects during model development; however, due to the limited sample size and small subgroup representation, they were excluded from final models, as reliable estimation of their independent effects was not statistically feasible.

All models were fitted using the “mixedlm” function from the “statsmodels” Python library (v0.14.0), which is able to perform linear mixed-effects modeling according to the method of restricted maximum likelihood (REML). This framework supports fixed and random effects and can process repeated-measures data with unbalanced observations. A random intercept for each patient was included to model intra-subject correlation due to repeated scanning of clots under different fixation conditions. For each relaxation parameter (T1, T2, T2*), we constructed a base model using fixation stage as a fixed effect and then extended it by including the selected covariates.

Results

Demographic data

All demographic data can be seen in Table 2. Among the 13 patients present in this study, 7 were female (54%), and the mean patient age was 61.3 years (SD ± 13.9 years). In total, comorbid conditions were prevalent and reflective of a high-risk thrombotic population: 12 patients (92%) had hypertension, 6 patients (46%) had diabetes mellitus, while cardiac conditions such as coronary artery disease or valvular heart disease were reported in 4 patients (31%). Hyperlipidemia (HLD) was noted also in 4 patients (31%). The mean BMI was 31.8 kg/m² (SD ± 3.7 kg/m²). Additional risk factors such as prior history of VTE, cancer, recent surgery, and mechanical factors (e.g., stenosis or device-associated thrombosis) were documented in several cases (Table 2). Autoimmune diseases and prolonged immobility were less common but present in isolated patients (Table 2).

Among the 6 patients with thrombosed arteriovenous fistulas (AVF), 3 cases were attributable to mechanical causes, including significant stenosis or occlusion of the anastomosis. Notably, mechanical compression was also implicated in some of the cases of DVT. The high incidence of cardiovascular and metabolic conditions underscores the multifactorial etiology of thrombus formation and management in this patient cohort. Clot samples were derived from a range of thrombotic conditions, with the most common etiologies being AVF in 6 patients, PE in 4 patients, and upper or lower extremity DVT in 3 patients. The median clot age at the time of scanning was 4 days, ranging from fresh (<1 day) to over 100 days.

The effect of fixation and different covariates on the relaxation times

Descriptive statistics for T1, T2, and T2* relaxation times across fixation stages are summarized in Tables 3,4, and on Figure 3. T1 values were highest in fresh samples (1,801.6±236.3 ms), followed by a marked reduction after <6 hours in formalin (1,577.2±298.5 ms). A further significant drop was also seen after >24 hours of fixation (1,205.0±491.6 ms). Similarly, T2 values showed an initial decline from fresh (77.6±16.5 ms) to <6 h formalin (52.5±21.8 ms), and a modest additional decline after >24 h (44.3±13.8 ms). T2* values demonstrated only slight changes across the fixation stages, decreasing from 9.9±2.3 ms (fresh) to 7.7±2.8 ms (<6 h) and 7.5±3.7 ms (>24 h).

Figure 3 The effect of formalin fixation on T1, T2, and T2* parameters.

To further evaluate the influence of fixation and biological covariates on relaxation times, mixed-effects models were constructed for each parameter. For the T1 relaxation time, the fixation stage was the strongest predictor. Compared to <6 h formalin, T1 was significantly higher in fresh samples (+283 ms, P<0.001) and significantly lower in >24 h formalin (–506 ms, P<0.001). Additionally, higher BMI was associated with shorter T1 values (–130 ms per unit BMI, P=0.001), and older clot age showed a strong negative correlation with T1, with samples aged 2 to >100 days showing significantly lower T1 values (–1,250 to –230 ms, P<0.005). Patient age was not a significant covariate.

For T2 relaxation time, formalin fixation made a significant difference, but only between the fresh and <6 h formalin groups (+26 ms, P<0.001). No significant differences were observed between <6 h and >24 h timepoints (P=0.13), suggesting that most of the T2 shortening occurs early in the fixation process. No covariates—including clot age, BMI, or patient age—were significantly associated with T2.

T2* relaxation time was the least affected by fixation. Although fixation showed a trend toward reduced values, neither the <6 h nor the >24 h groups differed significantly from the fresh condition. Clot age of 8 days was the only covariate that approached achieving statistical significance (4.94 ms, P=0.054). No other covariates significantly impacted T2*.

Collectively considered, these findings show that formalin fixation has the greatest effect on T1 relaxation times, followed by effects caused by clot age and BMI. T2 only responds to acute fixation effects, while T2* is relatively unchanging, irrespective of fixation conditions as well as variables dependent on a patient.

To determine how quickly thrombus samples need to be scanned to obtain accurate and biologically representative relaxation time measurements, three clot samples were imaged in their fresh state (within 1 hour of surgical removal) and then re-scanned after 6 and 24 hours without any fixation or immersion in liquid. We observed a marked decline in both T1 and T2 values over time (T1: 1,715±119.85→973±104.53→350±84.57 ms; T2: 165±113.46→57±41.33→35±23.89 ms), consistent with progressive water loss and tissue desiccation. These changes likely reflect the collapse of the hydrated clot matrix and reduced proton mobility, resulting in faster relaxation. T2* values showed a less linear trajectory (18±14.55→19±15.27→12±10.66 ms), potentially influenced by microstructural degradation and hemoglobin oxidation. While this highlights dehydration as a major confounding factor in ex vivo imaging, we did not evaluate whether temporary immersion in saline could mitigate these effects. These findings emphasize the importance of scanning clots in their fresh state—ideally within the first hour after surgical removal, to ensure biologically accurate relaxation measurements.

Discussion

Our study demonstrates that formalin fixation has a substantial and time-dependent effect on T1 and T2 relaxation times in ex vivo thrombus samples, with T1 showing progressive shortening across fixation stages and T2 displaying a rapid initial decline within the first 6 hours. In contrast, T2* values remained relatively stable. These findings align with and extend the existing literature on fixation-induced MR relaxation changes in other tissue types, including animal heart, liver and spleen and human heart, brain and cartilage tissues (15-26).

A main observation across prior studies is the time-dependent reduction in T1 relaxation following formalin exposure. Raman et al. showed that in formalin-fixed human brain tissue, T1 values dropped significantly over six months, initially rapidly, then gradually stabilized, following a biphasic decay pattern (15). Ebata et al. wrote about a similar pattern in postmortem human heart tissue, with T1 decreasing from ~405 ms in fresh samples to 175 ms after prolonged fixation (16). In both studies, T2 also decreased, although at a slower and less variable rate. Our findings in thrombus tissue also reflect similar biphasic T1 relaxation time behavior during this 24–48 h fixation period (15,16).

The mechanism behind T1 shortening has been widely attributed to formaldehyde-induced protein cross-linking, which prevents molecular motion and reduces the mobility of water protons. This was demonstrated by Birkl et al. in their examination performed in 2016, which confirmed a 76% reduction in T1 relaxation time values in fixed brain slices (17), and also by Fishbein et al., who linked T2 shortening in cartilage to irreversible structural changes of collagen following fixation. In thrombus tissue, where fibrin, cellular debris, and trapped plasma form the bulk matrix, protein cross-linking likely alters the proton environment in a similar manner, decreasing longitudinal relaxation efficiency (18).

Rivlin et al. provided informative molecular-level evidence by showing that proton exchange between water and methylene glycol (the hydrated form of formaldehyde) contributes significantly to T2 relaxation in formalin solutions (19). This chemical exchange process suggests that T2 shortening may occur even before deep tissue cross-linking, helping to explain the rapid decline in T2 relaxation times within the first 6 hours seen in our study. Similarly, Thickman et al. reported early fixation-induced T1 and T2 changes in rat liver and spleen, aligning with this observation that relaxation alterations begin almost immediately after formalin exposure (20).

Our result, that T2* remained relatively stable across fixation stages, however, contrasts with findings provided by Birkl et al. in their later work in 2018, who demonstrated that T2* decreased in fixed brain tissue and was sensitive to both formalin concentration and vendor formulation. This discrepancy may be due to differences in tissue susceptibility. Thrombi contain high concentrations of deoxyhemoglobin and hemosiderin, which dominate T2* relaxation through local magnetic field inhomogeneities. Because formalin does not oxidize iron or eliminate blood degradation products, T2* in thrombi may be inherently resistant to fixation-induced changes (21).

The recent work of Velasco Gonzalez et al. further supports this, showing that T2* and T1 relaxation time values in artificially created clot analogs are both influenced by iron and fibrin content, respectively. They reported that T2* decreased with increasing iron content, while within the same fixation state, higher fibrin fraction was associated with longer T1; across states, formalin fixation caused a larger global T1 reduction that overrides composition effects (22). These findings are consistent with our observation that fresh thrombi showed longer T1 than fixed samples due to the dominant global effect of fixation on T1. Dumitriu LaGrange et al. similarly demonstrated significant variation in relaxation times depending on clot composition on in vitro clot alternatives. They also highlighted the role of field strength when evaluating relaxation time parameters: their work confirmed that higher field strength (7 T) enhances the sensitivity of MRI to microstructural clot components, including iron and fibrin, thus supporting our observation of susceptibility-related stabilization of T2* in thrombi (23). Moreover, their data correlate with our findings of how fixation and intrinsic tissue properties influence T1 and T2, reaffirming the need to account for field strength and fixation status when interpreting MR parameters in thrombosis research. Bainbridge et al. [2004] also highlighted that fixation reduced T1 but increased T2 in brain tissue with multiple sclerosis (MS) imaged with a 7 T scanner, possibly due to high-field susceptibility effects (24). This emphasizes the effect of field strength dependence when interpreting ex vivo findings, as demonstrated by Bottomley et al. (25).

Notably, our study showed that clot age and BMI influenced T1 values, with older and higher-BMI samples exhibiting shorter T1 relaxation time values. These results suggest that intrinsic factors such as systemic metabolic or inflammatory states may influence thrombus microstructure and water dynamics, potentially through mechanisms like altered fibrin organization, plasma composition, or red cell degradation, ultimately altering water dynamics and proton mobility. While not directly addressed in the studies by Shatil et al. or Fishbein et al., their findings of progressive T1 and T2 signal alteration and the structural reorganization of collagen over time also support this hypothesis (18,26).

Lastly, Bilgin et al. demonstrated that histologic quality of thrombus that was preserved for up to two years after fixation, suggesting that MRI signal changes do not necessarily indicate tissue degradation, but rather biochemical reorganization (3). Together, these findings support the interpretative value of T1 and T2 as indirect biomarkers of clot age, fixation status, and composition, particularly as supplementary to clinical covariates. The summary of the existing literature is represented by Table 5.

Limitations

This study has several limitations that should be acknowledged. First, the sample size was limited, comprising 19 clot samples from 13 patients. While linear mixed-effects modeling was used to accommodate repeated measures and incomplete timepoint data, the small cohort reduced statistical power and limited the ability to assess interactions or control for confounders. Continuous covariates such as BMI, clot age, and patient age were included as fixed effects, but categorical variables like sex and clot type were excluded from final models due to insufficient subgroup sizes, which made reliable estimation of their independent effects statistically infeasible.

Second, not all samples underwent imaging at every fixation timepoint (fresh, <6 h formalin, >24 h formalin), primarily due to tissue availability and time constraints during scanning. This introduces some inconsistency across the dataset, although mixed-effects modeling allowed us to retain partially scanned samples and still draw comparisons within and between groups. Nonetheless, uniform scanning at all stages would have strengthened internal validity.

Third, we used 10% neutral buffered formalin for fixation, consistent with standard histopathological practice, but we did not compare its effects to other fixatives (e.g., paraformaldehyde) or non-fixative storage (e.g., saline). While formalin is known to alter proton mobility and relaxation times, the degree to which different fixatives or preservation techniques might affect clot tissue MR characteristics remains unexplored and should be addressed in future studies.

Fourth, clot type classification was based on clinical and procedural context rather than histological confirmation. While practical in real-world thrombectomy workflows, this introduces potential for misclassification, especially in mixed thrombi or complex presentations. Similarly, clot age estimation was based on symptom onset and intervention timing, which may not reflect true biological age or remodeling stage of the thrombus. Also, we did not directly quantify red-cell or hemoglobin loss during fixation; future work will pair relaxation mapping with compositional assays to test this mechanism.

Fifth, our segmentation methodology relied primarily on intensity thresholding and manual refinement using OsiriX. While this approach was practical and reproducible for the high-resolution data, it may be limited by user subjectivity and tissue-formalin contrast variability. More advanced segmentation methods, such as machine learning-based tools or signal deconvolution techniques, may improve accuracy in future studies.

Lastly, all imaging was performed ex vivo at ultra-high field strength (9.4 T). Although this provides excellent resolution and sensitivity, the relaxation times measured under these conditions may not directly translate to clinical field strengths (e.g., 1.5 or 3 T) or in vivo applications. Future validation studies using both clinical imaging platforms and histological correlation will be essential for translational implementation.

Conclusions

This study demonstrates that formalin fixation has a substantial and time-dependent effect on the MRI relaxation properties of human thrombi.

Our findings are consistent with prior reports in brain, heart, and cartilage tissues, reinforcing that both cross-linking of matrix proteins and early chemical interactions substantially impact proton mobility and relaxation dynamics. Importantly, biological covariates should also be taken into consideration for possible influence, suggesting that endogenous clot maturation and systemic factors can further modulate MRI relaxation signatures independently of fixation effects.

Formalin fixation and passive dehydration were associated with significant T1 and T2 shortening, indicating that sample handling can obscure or mimic intrinsic clot features. T2* changed little across fixation stages; thus, susceptibility-weighted metrics appeared comparatively less affected under our protocol, although confirmation with compositional assays is needed. Based on these findings, ex vivo thrombus MRI studies prioritize imaging as close to the time of surgical removal as possible, ideally within the first hour, is strongly recommend it in order to preserve native relaxation properties and avoid dehydration-related distortion. Although ex vivo MRI of human thrombi has been reported, prior work has not systematically quantified the time-dependent effects of formalin fixation on T1, T2, and T2 relaxation times; thus, our results provide a reference point for future investigations as well as interpreting existing animal studies and artificial clot analog models. Therefore, tissue handling conditions should be carefully considered in future study designs and when comparing across experimental platforms, and while fixation allows histologic correlation, its effect on relaxation times limits its use in developing quantitative MRI biomarkers.

Acknowledgments

We gratefully acknowledge the contributions of Janak Lamichhane, whose support with consenting and data collection was essential to this study.

Footnote

Data Sharing Statement: Available at https://atm.amegroups.com/article/view/10.21037/atm-25-108/dss

Peer Review File: Available at https://atm.amegroups.com/article/view/10.21037/atm-25-108/prf

Funding: This work was supported by the Houston Methodist Research Institute Clinician-Scholar program, the Jerold B. Katz Academy of Translational Science under project number 15790002 (recipient’s name: Trisha Roy), the Cornerstone President Award number 15790006 (recipient’s name: Trisha Roy), the American Heart Association Transformational Award number 23TPA1077601 (recipient’s name: Trisha Roy), the National Institutes of Health Research Project grant (R01) under award number R01HL174587 (recipient’s name: Trisha Roy).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://atm.amegroups.com/article/view/10.21037/atm-25-108/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Houston Methodist Hospital (PRO00027258) and informed consent was obtained from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Toussaint JF, LaMuraglia GM, Southern JF, et al. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo. Circulation 1996;94:932-8. [Crossref] [PubMed]
2. Douglas A, Fitzgerald S, Pandit A, et al. Storage of blood clots for histological analysis: How long is too long in saline and paraformaldehyde? Histol Histopathol 2020;35:313-20. [Crossref] [PubMed]
3. Bilgin C, Dai D, Johnson C, et al. Quality assessment of histopathological stainings on prolonged formalin fixed thrombus tissues retrieved by mechanical thrombectomy. Front Neurol 2023;14:1223947. [Crossref] [PubMed]
4. Liebeskind DS, Sanossian N, Yong WH, et al. CT and MRI early vessel signs reflect clot composition in acute stroke. Stroke 2011;42:1237-43. [Crossref] [PubMed]
5. Ramachandran H, Girdhar S, Sreedharan SE, et al. Does thrombus imaging characteristics predict the degree of recanalisation after endovascular thrombectomy in acute ischaemic stroke? J Stroke Cerebrovasc Dis 2022;31:106621. [Crossref] [PubMed]
6. Vidmar J, Blinc A, Kralj E, et al. An MRI study of the differences in the rate of thrombolysis between red blood cell-rich and platelet-rich components of venous thrombi ex vivo. J Magn Reson Imaging 2011;34:1184-91. [Crossref] [PubMed]
7. Hendley SA, Dimov A, Bhargava A, et al. Assessment of histological characteristics, imaging markers, and rt-PA susceptibility of ex vivo venous thrombi. Sci Rep 2021;11:22805. [Crossref] [PubMed]
8. Christiansen SD, Liu J, Bullrich MB, et al. Deep learning prediction of stroke thrombus red blood cell content from multiparametric MRI. Interv Neuroradiol 2024;30:541-9. [Crossref] [PubMed]
9. Tratar G, Blinc A, Podbregar M, et al. Characterization of pulmonary emboli ex vivo by magnetic resonance imaging and ultrasound. Thromb Res 2007;120:763-71. [Crossref] [PubMed]
10. Shah KJ, Roy TL. Catheter-Directed Interventions for the Treatment of Lower Extremity Deep Vein Thrombosis. Life (Basel) 2022;12:1984. [Crossref] [PubMed]
11. Alkarithi G, Duval C, Shi Y, et al. Thrombus Structural Composition in Cardiovascular Disease. Arterioscler Thromb Vasc Biol 2021;41:2370-83. [Crossref] [PubMed]
12. Chernysh IN, Nagaswami C, Kosolapova S, et al. The distinctive structure and composition of arterial and venous thrombi and pulmonary emboli. Sci Rep 2020;10:5112. [Crossref] [PubMed]
13. Nosaka M, Ishida Y, Kimura A, et al. Time-dependent appearance of intrathrombus neutrophils and macrophages in a stasis-induced deep vein thrombosis model and its application to thrombus age determination. Int J Legal Med 2009;123:235-40. [Crossref] [PubMed]
14. Fineschi V, Turillazzi E, Neri M, et al. Histological age determination of venous thrombosis: a neglected forensic task in fatal pulmonary thrombo-embolism. Forensic Sci Int 2009;186:22-8. [Crossref] [PubMed]
15. Raman MR, Shu Y, Lesnick TG, et al. Regional T(1) relaxation time constants in Ex vivo human brain: Longitudinal effects of formalin exposure. Magn Reson Med 2017;77:774-8. [Crossref] [PubMed]
16. Ebata K, Noriki S, Inai K, et al. Changes in magnetic resonance imaging relaxation time on postmortem magnetic resonance imaging of formalin-fixed human normal heart tissue. BMC Med Imaging 2021;21:134. [Crossref] [PubMed]
17. Birkl C, Langkammer C, Golob-Schwarzl N, et al. Effects of formalin fixation and temperature on MR relaxation times in the human brain. NMR Biomed 2016;29:458-65. [Crossref] [PubMed]
18. Fishbein KW, Gluzband YA, Kaku M, et al. Effects of formalin fixation and collagen cross-linking on T2 and magnetization transfer in bovine nasal cartilage. Magn Reson Med 2007;57:1000-11. [Crossref] [PubMed]
19. Rivlin M, Eliav U, Navon G. NMR studies of proton exchange kinetics in aqueous formaldehyde solutions. J Magn Reson 2014;242:107-12. [Crossref] [PubMed]
20. Thickman DI, Kundel HL, Wolf G. Nuclear magnetic resonance characteristics of fresh and fixed tissue: the effect of elapsed time. Radiology 1983;148:183-5. [Crossref] [PubMed]
21. Birkl C, Soellradl M, Toeglhofer AM, et al. Effects of concentration and vendor specific composition of formalin on postmortem MRI of the human brain. Magn Reson Med 2018;79:1111-5. [Crossref] [PubMed]
22. Velasco Gonzalez A, Buerke B, Görlich D, et al. Making acute ischemic stroke thrombi visible in MRI imaging. Sci Rep 2024;14:12325. [Crossref] [PubMed]
23. Dumitriu LaGrange D, Xin L, Lazeyras F, et al. MRI characterization of in vitro clots at 3T and 7T: A technical note. J Neuroradiol 2024;51:38-42. [Crossref] [PubMed]
24. Bainbridge A, Schmierer K, Miller DH, et al. Changes in T1 and T2 in Formalin fixed tissue compared to fresh tissue at 7 Tesla. Proceedings of the 12th Annual Meeting of ISMRM, Kyoto, Japan. 2004:2316.
25. Bottomley PA, Foster TH, Argersinger RE, et al. A review of normal tissue hydrogen NMR relaxation times and relaxation mechanisms from 1-100 MHz: dependence on tissue type, NMR frequency, temperature, species, excision, and age. Med Phys 1984;11:425-48. [Crossref] [PubMed]
26. Shatil AS, Uddin MN, Matsuda KM, et al. Quantitative Ex Vivo MRI Changes due to Progressive Formalin Fixation in Whole Human Brain Specimens: Longitudinal Characterization of Diffusion, Relaxometry, and Myelin Water Fraction Measurements at 3T. Front Med (Lausanne) 2018;5:31. [Crossref] [PubMed]