Value of enhanced fluid-attenuated inversion-recovery T1-weighted imaging sequences combined with the three-dimensional modulated flip-angle technique in diagnosing brain metastases

Introduction

With the progressive advancements made in systemic treatment strategies for malignant tumors, the overall survival time of patients with cancer has been prolonged, leading to an increase in the incidence of brain metastases. Brain metastases are the most common intracranial tumors in adults, occurring in approximately 20–40% of patients with primary malignant tumors (1). The prognosis for patients with brain metastases is generally poor, with an estimated 2-year overall survival rate of 8.1% and a 5-year overall survival rate of 2.5% for all tumor types (2). Systemic therapy is crucial for treating patients with advanced cancer and brain metastases. However, a variety of drug types are unable to penetrate the blood–brain barrier, limiting their effectiveness against brain metastases (3). A growing body of evidence suggests that stereotactic radiotherapy (SRT) offers unique advantages in the treatment of brain metastases. Whether used alone or in combination with whole-brain radiotherapy (WBRT), SRT has become a standard modality for treating brain metastases (4). For SRT, it is critical to determine the presence, number, and boundaries of brain metastases (5-7).

Contrast-enhanced magnetic resonance imaging (CE-MRI) is typically employed with T1-weighted imaging (T1WI) sequences. However, due to its thick slices and wide intervals, there is a risk of small lesions being missed. Reducing slice thickness and interval can significantly improve the diagnostic accuracy and detection rate for brain metastases, but it also reduces the signal-to-noise ratio (SNR), making the differentiation of small metastases from small blood vessels in cortical or deep structures difficult (8,9). To selectively suppress vascular signals and effectively differentiate between brain metastases and small blood vessels enhanced by contrast agents, black-blood MRI was developed (10). The three-dimensional fast-spin-echo (3D FSE) black-blood sequence, employing the modulated flip-angle technique in refocusing imaging with extended echo trains (MATRIX), involves an iterative method that modulates optimal protocols from an inversion angle database, effectively suppressing blood flow signals within vessels (11). Compared to traditional FSE sequences, the MATRIX sequence can accelerate imaging by using longer echo train lengths (ETLs) and shorter echo spacing (ESP). Studies have shown that using the MATRIX sequence shortens the acquisition time of routine clinical knee MRI without compromising image quality, SNR, or contrast-to-noise ratio (CNR) (11).

The objective of this study was to evaluate the effectiveness of the MATRIX CE-T1FLAIR sequence in detecting brain metastases, with the goal of optimizing scanning protocols. By investigating the capabilities of this advanced imaging technique, the study aimed to provide insights that could lead to improved diagnostic accuracy and better clinical outcomes for patients with brain metastases. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-941/rc).

Methods

Clinical data

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and received approval from the Medical Ethics Committee of Henan Cancer Hospital (approval No. 2022-KY-0063-001). The requirement for informed consent was waived due to retrospective nature of the analysis and the absence of any additional interventions or procedures beyond routine clinical care. This prospective study consecutively included patients from Henan Cancer Hospital who were pathologically diagnosed with malignant tumors and underwent CE 3.0T MRI scans due to suspected brain metastases between October 2023 and February 2024. A total of 326 patients were initially enrolled. Lesion detection and radial line measurements were based on the linked smart brain metastases artificial intelligence (AI)-assisted detection system (12) and clinical follow-up, through which 176 patients were clinically confirmed to have parenchymal brain metastases. The inclusion and exclusion criteria for the study participants are detailed in Figure 1.

Figure 1 Patient flowchart. AI, artificial intelligence; LM, leptomeningeal metastasis; MRI, magnetic resonance imaging.

MRI examination protocol

The MRI examinations were performed with a 3.0T uMR770-R002 MRI scanner (United Imaging Healthcare, Shanghai, China) equipped with a 16-channel head-and-neck coil. During the CE scans, a routine dose of gadolinium-based contrast agent was administered at a dose of 0.1 mmol/kg body weight via a high-pressure injector at a flow rate of 2.5 mL/s. Specifically, all scans were performed via gadodiamide injection (Omniscan, GE HealthCare, Chicago, IL, USA). The MRI scans included both unenhanced and enhanced sequences. The unenhanced sequences consisted of axial T1 fluid-attenuated inversion recovery (FLAIR), axial T2WI, axial T2FLAIR, sagittal T1FLAIR, and axial diffusion-weighted imaging (DWI). The enhanced sequences included sagittal T1FLAIR, coronal T2FLAIR, axial MATRIX CE-T1FLAIR, three-dimensional magnetization-prepared rapid gradient echo (3D GRE) fast-spin echo (FSE) with angle flow-sensitive preparation (fsp) CE-T1FLAIR, and 2D FSE CE-T1FLAIR. The sequence acquisition order was randomized across patients to eliminate order bias. The scanning sequence and parameters for the cranial MRI sequences are detailed in Table 1.

Image analysis

All images were evaluated and analyzed in a double-blind manner by three radiologists, each with at least 5 years of diagnostic experience. Specialized analysis software (linked smart brain metastases AI-assisted detection system) (12) on the postprocessing workstation was employed to process and analyze images from the MATRIX CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and 2D FSE CE-T1FLAIR sequences. The 2D FSE-CE-T1FLAIR sequence was specifically included as a clinical reference standard, given its widespread use and established sensitivity in detecting brain metastases in routine practice. The final number of metastases was determined through the following steps: (I) lesions were detected on the united imaging AI-assisted brain metastases detection system across the MATRIX CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and 2D FSE CE-T1FLAIR images; (II) the two radiologists manually confirmed and measured the lesions; (III) lesions showing vascular cross-sections or nonmetastatic enhancement or those that disappeared or remained unchanged in size during follow-up were excluded; (IV) the final decision for any discrepancies was made by a senior neuroradiologist with 25 years of experience. The image analysis included lesion identification, lesion count, and diameter measurement. Specific image evaluation criteria encompassed the clarity of lesion edges, the prominence of contrast enhancement, and the visibility of surrounding tissues. Lesion diameter measurements were performed using the linked smart brain metastases AI-assisted detection system. Based on their diameters, lesions were categorized into three groups: <5, 5–10, and >10 mm. The radiologists recorded the lesion count and measured each lesion’s maximum diameter. In cases of discrepancies, a senior radiologist with 10 years of diagnostic experience made the final determination to ensure consistency and accuracy in the evaluations.

Statistical analysis

Statistical analyses were performed with SAS 9.4 software (SAS Institute, Cary, NC, USA). The differences in lesion detection rates between the three sequences were compared with the χ² test or the Fisher exact test, with multiple comparisons conducted via the Bonferroni correction method. The Kappa consistency test was used to assess the interrater reliability for measurements between the two radiologists, with a Kappa value ≥0.6 indicating good consistency. A P value of <0.05 was considered statistically significant, and the Bonferroni-corrected P value was set to 0.05/n, where n represents the number of statistical comparisons.

Results

A total of 176 patients were included in this study, with an average age of 59.2±10.6 years. As shown in Table 2, the cohort comprised 73 males and 103 females. The primary sources of tumors were as follows: lung cancer in 115 cases (65.34%), breast cancer in 51 cases (28.98%), liver cancer in 3 cases (1.7%), pancreatic cancer in 3 cases (1.7%), ovarian cancer in 2 cases (1.14%), and sigmoid colon cancer in 2 cases (1.14%). The distribution of brain metastases was predominantly in the gray-white matter junction (49.83%), followed by the superficial area of the brain’s convexity (32.47%), deep cerebral lobe (8.45%), cerebellum (7.67%), and brainstem (1.58%). The average number of brain metastases per patient was 5.04±3.40. Notably, 2 patients had more than 10 brain metastases.

Consistency in total lesion count of brain metastases across three sequence images

The consistency of total lesion count of brain metastases among the three radiologists was excellent across the MATRIX CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and FSE CE-T1FLAIR sequence images, with Kappa values of 0.83 [95% confidence interval (CI): 0.78–0.88], 0.79 (95% CI: 0.74–0.84), and 0.77 (95% CI: 0.72–0.82), respectively, with the MATRIX CE-T1FLAIR sequence demonstrating the highest consistency.

Comparison of total detection rates and intergroup detection rates for brain metastases of different diameters across three sequences

The overall detection rates of brain metastases differed significantly between the MATRIX CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and FSE CE-T1FLAIR sequence images (P<0.001). The total detection rate for MATRIX CE-T1FLAIR was 98.9%, significantly higher than that for 3D GRE fsp CE-T1FLAIR at 93.5% and FSE CE-T1FLAIR at 80.6% (Figure 2). When categorization was conducted according to lesion diameter, the detection rates for lesions <5 and 5–10 mm were 98.8% and 98.0% for MATRIX CE-T1FLAIR, respectively, significantly higher than those for 3D GRE fsp CE-T1FLAIR and FSE CE-T1FLAIR. For lesions >10 mm, there was no significant difference in detection rates between the three sequences (P>0.05) (Table 3, Figures 3,4).

Figure 2 A 66-year-old male patient with brain metastases from lung cancer. In the MATRIX CE-T1FLAIR sequence (A), multiple small enhancing nodules (white arrows) were observed in the bilateral temporal lobes, left occipital lobe, and brainstem. However, the 3D GRE fsp CE-T1FLAIR sequence (B) showed only two small enhancing foci (white arrows). Conversely, the FSE CE-T1FLAIR sequence (C) did not demonstrate clear findings. However, upon reassessment 3 months after second-line treatment, the FSE CE-T1FLAIR sequence (D) revealed multiple small metastases in bilateral temporal lobes, left occipital lobe, cerebellum, and brainstem (white arrows). 3D GRE, three-dimensional gradient echo; CE-T1FLAIR, contrast-enhanced T1-weighted fluid-attenuated inversion recovery; FSE, fast-spin echo; MATRIX, modulated flip angle technique in refocused imaging with extended echo trains.

Figure 3 A comparative statistical chart of lesion and microlesion detection numbers across three sequences. MATRIX CE-T1FLAIR exhibited a significantly higher detection rate of brain metastases with long diameters <5 mm (A) and 5–10 mm (B) and a higher overall detection rate (C) compared to 3D GRE fsp CE-T1FLAIR and FSE CE-T1FLAIR (P<0.001). **, P<0.01; ****, P<0.0001. 3D GRE, three-dimensional gradient echo; CE-T1FLAIR, contrast-enhanced T1-weighted fluid-attenuated inversion recovery; FSE, fast-spin echo; fsp, flow-sensitive preparation; MATRIX, modulated flip angle technique in refocused imaging with extended echo trains.

Figure 4 A 53-year-old male patient with brain metastases from breast cancer. In the MATRIX CE-T1FLAIR sequence (A), multiple enhancing nodules with diameters <5 mm (white arrows) were observed in the brainstem and cerebellar vermis. However, the 3D GRE fsp CE-T1FLAIR sequence (B) had indistinct findings (white arrows). Brainstem lesions were not clearly visualized on the FSE CE-T1FLAIR sequence (C), and a single enhancing nodule in the left cerebellar vermis appeared blurry and difficult to distinguish from vascular cross-sections (white arrow). Upon reassessment 3 months later, the FSE CE-T1FLAIR sequence (D) revealed enlarged and ring-enhancing nodules in the brainstem and cerebellar vermis, along with multiple new enhancing nodules within the brain parenchyma (white arrows), confirming metastatic tumors. 3D GRE, three-dimensional gradient echo; CE-T1FLAIR, contrast-enhanced T1-weighted fluid-attenuated inversion recovery; FSE, fast-spin echo; fsp, flow-sensitive preparation; MATRIX, modulated flip angle technique in refocused imaging with extended echo trains.

Comparison of detection rates for brain metastases in different subgroups based on the distribution across three sequence images

Significant differences were observed in the detection rates for brain metastases in the superficial brain convexity, gray-white matter junction, basal ganglia, cerebellum, and brainstem subgroups across the MATRIX CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and FSE CE-T1FLAIR sequence images (P<0.001). Pairwise comparisons showed that the detection rates for all the subgroups were significantly higher for MATRIX CE-T1FLAIR and 3D GRE fsp CE-T1FLAIR as compared to FSE CE-T1FLAIR (P<0.001). Specifically, MATRIX CE-T1FLAIR had significantly higher detection rates in the superficial brain convexity (P=0.002), gray-white matter junction (P<0.001), and cerebellum (P=0.015) compared to 3D GRE fsp CE-T1FLAIR. However, no significant differences were found between MATRIX CE-T1FLAIR and 3D GRE fsp CE-T1FLAIR in the basal ganglia and brainstem (P>0.05) (Table 4, Figures 5,6).

Figure 5 A comparative statistical chart of brain metastasis detection numbers across different subgroups for the three sequences. MATRIX CE-T1FLAIR showed a higher detection rate for brain metastases in the superficial cortical area (A), the gray-white matter junction (B), and the cerebellum (D) compared to 3D GRE fsp CE-T1FLAIR and FSE CE-T1FLAIR, with differences being statistically significant (P<0.05). However, in the basal ganglia (C) and brainstem (E), the detection rate differences between MATRIX CE-T1FLAIR and 3D GRE fsp CE-T1FLAIR were not statistically significant (P>0.05). ns, no statistically significant difference (P>0.05); *, statistically significant difference (P<0.05); **, P<0.01; ****, P<0.0001. 3D GRE, three-dimensional gradient echo; CE-T1FLAIR, contrast-enhanced T1-weighted fluid-attenuated inversion recovery; FSE, fast-spin echo; fsp, flow-sensitive preparation; MATRIX, modulated flip angle technique in refocused imaging with extended echo trains.

Figure 6 A 66-year-old female patient with newly diagnosed lung cancer and a history of hypertension and occasional dizziness who underwent preoperative examination to clarify the intracranial condition. The MATRIX CE-T1FLAIR sequence (A) showed a tiny enhancing nodule near the confluence of sinuses on the left parietal convexity (white arrow), which was not clearly visible on the 3D GRE fsp CE-T1FLAIR sequence (B) and FSE CE-T1FLAIR sequence (C). The patient was transferred to internal medicine for treatment based on MRI indications of brain metastasis and underwent targeted therapy. After 3 months, follow-up examination with the FSE CE-T1FLAIR sequence (D) showed that the previously noted tiny enhancing nodule near the confluence of sinuses on the left parietal convexity had increased in size, displaying ring enhancement. Follow-up confirmed the diagnosis of brain metastasis (white arrow). 3D GRE, three-dimensional gradient echo; CE-T1FLAIR, contrast-enhanced T1-weighted fluid-attenuated inversion recovery; FSE, fast-spin echo; fsp, flow-sensitive preparation; MATRIX, modulated flip angle technique in refocused imaging with extended echo trains; MRI, magnetic resonance imaging.

Discussion

CE-MRI is considered to be the most effective method for detecting brain metastases. Metastatic tumors within the brain parenchyma frequently occur at the gray-white matter junction in the territory of the middle cerebral artery. This predilection is a result of the high vascularity and small vessel diameter in the cortex-medulla junction and sulcal regions, where blood flow is relatively slow. After the injection of contrast agents, enhanced vessels appear as small nodular or punctate hyperintense signals on conventional T1WI. However, small metastatic tumors are often difficult to distinguish from small vascular cross-sections due to their minimal or absent edema and small size (13,14). The primary drawbacks of CE-T1WI include thick slices, interslice gaps, and the inability to perform multiplanar reconstructions, which can result in the omission of small lesions (<5 mm). Moreover, cerebrospinal fluid (CSF) motion and patient movement artifacts can obscure lesions in the cortical sulci, cisterns, and meninges, leading to false positives and negatives.

The MATRIX sequence is a 3D FSE black-blood sequence based on variable flip-angle technology. It uses iterative methods to optimize flip angles from a database, effectively suppressing intravascular blood flow signals, thus providing superior SNR and black-blood capabilities. This sequence clearly visualizes vessel walls, black-blood backgrounds, and intraluminal abnormal signal thrombi and allows for thin-slice continuous scanning and multiplanar reconstruction (11). Although MATRIX sequences have been reported for detecting deep vein thrombosis, diagnosing cartilage injuries, and imaging the cranial nerve, their application in detecting and diagnosing brain metastases is less common. Sui et al. (15) demonstrated that MATRIX, as a novel MR black-blood imaging sequence, reduces scan time by 30% without compromising image quality or diagnostic performance as compared to conventional 2D sequences. Li et al. (16) found that for knee joint MRI, the MATRIX sequence had high consistency in evaluating cartilage, subchondral bone, and ligaments as compared to 2D FSE or proton density sequences. Bae et al. (17) confirmed that 3D CE-T1W FSE imaging could selectively suppress vessels and achieve high a CNR, making it highly suitable for detecting brain metastases ≤3 mm. Cao et al. (18) discovered that 3D CE-T1W flow-sensitive black-blood sequences could identify more intratumoral vessels compared to convention sequences, reporting a positive linear relationship between the number of intratumoral vessels and microbleeds. Compared with 2D sequences, 3D FSE sequences such as MATRIX CE-T1FLAIR have several advantages that contribute to higher detection sensitivity. These advantages include a higher SNR, better image quality, and the ability to reconstruct images in any plane while preserving resolution. These features help reduce partial volume effects, improve lesion visibility, and allow for more accurate detection of small metastases.

Our study demonstrated significant differences in the overall detection rates of brain metastases between the MATRIX CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and 2D FSE CE-T1FLAIR sequences, with MATRIX CE-T1FLAIR achieving the highest overall detection rate. Specifically, for lesions <5 and 5–10 mm, MATRIX CE-T1FLAIR outperformed 3D GRE fsp CE-T1FLAIR and 2D FSE CE-T1FLAIR in terms of detection rate. Furthermore, the distribution of brain metastases primarily occurred in the superficial brain convexity and cortex-medulla junction (727 lesions, 82.0%), consistent with the anatomical and vascular characteristics that contribute to the lodging of tumor emboli in these regions. Significant differences in detection rates were observed for the superficial brain convexity, gray-white matter junction, basal ganglia, cerebellum, and brainstem subgroups across the three sequences. Pairwise comparisons revealed that MATRIX CE-T1FLAIR and 3D GRE fsp CE-T1FLAIR had significantly higher detection rates than did 2D FSE CE-T1FLAIR across all subgroups. MATRIX CE-T1FLAIR demonstrated significantly higher detection rates than did 3D GRE fsp CE-T1FLAIR in the superficial brain convexity, gray-white matter junction, and cerebellum, while no significant differences were found in the basal ganglia or brainstem (19). The imaging characteristics of the MATRIX CE-T1FLAIR sequence, such as its iterative method for optimal flip-angle modulation and effective blood flow signal suppression, enhance SNR and black-blood capabilities, thereby preventing the misdiagnosis and omission of small vascular cross-sections on high-resolution T1W enhanced images. These technical strengths make MATRIX CE-T1FLAIR especially suitable for detecting small-sized, subtly enhanced metastases in regions prone to vascular interference.

In addition, spatial resolution plays a critical role in the sensitivity of brain metastasis detection. The MATRIX sequence provides high isotropic resolution, which allows for better delineation of small lesions and enables multiplanar reformation without loss of detail. In contrast, the conventional 2D FSE sequence, with thicker slices and anisotropic voxels, may limit the detection of subcentimeter lesions due to partial volume effects. This difference in resolution likely contributes to the superior performance of the MATRIX sequence in identifying smaller metastases.

The significant differences in detection rates between various anatomical locations and imaging sequences may be attributed to several factors. First, the anatomical characteristics and susceptibility to artifacts vary across brain regions. For instance, the posterior fossa, including the cerebellum and brainstem, is prone to magnetic susceptibility and CSF pulsation artifacts, which can degrade image quality and obscure small lesions. In contrast, superficial cortical areas and the gray-white matter junction tend to have higher lesion detectability, possibly due to better contrast and fewer structural interferences. Second, partial volume effects in small deep-brain structures such as the basal ganglia may reduce detection sensitivity, especially for lesions near vascular structures or ventricles. Additionally, some lesions in periventricular regions may be masked by adjacent high-signal vascular or CSF-related flow artifacts, particularly in sequences lacking sufficient blood suppression. Finally, MATRIX CE-T1FLAIR showed superior performance in detecting small lesions in superficial and cerebellar regions, likely due to its optimized flip-angle modulation, black-blood effect, and higher isotropic spatial resolution, which together improve lesion conspicuity and reduce confusion with vascular signals. These technical advantages may explain its significantly higher detection rates for certain anatomical subregions when compared to the 3D GRE fsp and 2D FSE sequences. Moreover, in our literature review, we found that many studies (20-22) have reported differences in the number of small-volume brain metastases detected at varying postcontrast time points. Several studies recommend performing scanning within 10–15 minutes after contrast agent injection to optimize detection, as this window appears to enhance both lesion visibility and lesion volume. Conversely, other research (9) suggests that postinjection delay does not significantly affect lesion conspicuity when a standard single-dose of gadolinium contrast is used. Given this inconsistency in findings, we did not adopt a fixed acquisition order for our postcontrast sequences. Instead, to minimize potential bias and reflect real-world imaging workflow, we randomized the order of acquiring the 2D FSE CE-T1FLAIR, 3D GRE fsp CE-T1FLAIR, and MATRIX CE-T1FLAIR sequences across patients.

Certain limitations to this study should be noted. To begin, we employed a prospective consecutive case series design within a specific time frame, which might have introduced bias due to the short duration. Second, meningeal metastases were not analyzed or discussed. Future studies should expand the sample size and extend follow-up periods to further verify the detection and diagnostic capabilities of the MATRIX CE-T1FLAIR sequence for meningeal metastases. Additionally, multicenter, multivendor studies to enhance the generalizability of our findings. Although follow-up scans were used to confirm lesion validity, lesion-by-lesion comparison over time and interreader spatial agreement was beyond the scope of this study and will be explored in future work. Moreover, we relied on a double-blind approach with three independent radiologists to assess lesion detection. Although the linked smart brain metastases AI-assisted detection system helped standardize lesion identification, variability in reader interpretation might nonetheless have been present. The interrater reliability was good, as demonstrated by the Kappa values ranging from 0.77 to 0.83. However, further work is needed to improve consistency in identifying lesions, particularly in cases with subtle characteristics, through enhanced training and more advanced AI tools.

Conclusions

The MATRIX CE-T1FLAIR sequence significantly improves the detection rate and diagnostic accuracy for brain metastases, especially for small lesions as compared to traditional sequences. Its superior imaging quality and ability to reduce misdiagnosis make it a valuable tool in neuroimaging for the early and precise identification of brain metastases.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-941/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-941/dss

Funding: This work was supported by the Henan Province Science and Technology Key Project (No. 232102311052).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-941/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. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study received approval from the Medical Ethics Committee of Henan Cancer Hospital (No. 2022-KY-0063-001). Informed consent was waived because the study was retrospective in nature, and did not involve any additional interventions or procedures beyond routine clinical care.

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. Yan Q, Li R, Yang J, Bai X, Guo X, Yang X, Song J. Efficacy and safety evaluation of combined therapies incorporating whole-brain radiotherapy in patients with brain metastases: a systematic review and meta-analysis. Clin Transl Oncol 2024;26:3020-36. [Crossref] [PubMed]
2. Hall WA, Djalilian HR, Nussbaum ES, Cho KH. Long-term survival with metastatic cancer to the brain. Med Oncol 2000;17:279-86. [Crossref] [PubMed]
3. Parrish KE, Sarkaria JN, Elmquist WF. Improving drug delivery to primary and metastatic brain tumors: strategies to overcome the blood-brain barrier. Clin Pharmacol Ther 2015;97:336-46. [Crossref] [PubMed]
4. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF, Kornguth DG, Arbuckle RB, Swint JM, Shiu AS, Maor MH, Meyers CA. Neurocognition in patients with brain metastases treated with radiosurgery or radiosurgery plus whole-brain irradiation: a randomised controlled trial. Lancet Oncol 2009;10:1037-44. [Crossref] [PubMed]
5. Yamamoto M, Serizawa T, Shuto T, Akabane A, Higuchi Y, Kawagishi J, et al. Stereotactic radiosurgery for patients with multiple brain metastases (JLGK0901): a multi-institutional prospective observational study. Lancet Oncol 2014;15:387-95. [Crossref] [PubMed]
6. Yamamoto M, Kawabe T, Sato Y, Higuchi Y, Nariai T, Watanabe S, Kasuya H. Stereotactic radiosurgery for patients with multiple brain metastases: a case-matched study comparing treatment results for patients with 2-9 versus 10 or more tumors. J Neurosurg 2014;121:16-25. [Crossref] [PubMed]
7. Eaton DJ, Lee J, Paddick I. Stereotactic radiosurgery for multiple brain metastases: Results of multicenter benchmark planning studies. Pract Radiat Oncol 2018;8:e212-20. [Crossref] [PubMed]
8. Reichert M, Morelli JN, Runge VM, Tao A, von Ritschl R, von Ritschl A, Padua A, Dix JE, Marra MJ, Schoenberg SO, Attenberger UI. Contrast-enhanced 3-dimensional SPACE versus MP-RAGE for the detection of brain metastases: considerations with a 32-channel head coil. Invest Radiol 2013;48:55-60. [Crossref] [PubMed]
9. Jeon JY, Choi JW, Roh HG, Moon WJ. Effect of imaging time in the magnetic resonance detection of intracerebral metastases using single dose gadobutrol. Korean J Radiol 2014;15:145-50. [Crossref] [PubMed]
10. Yang S, Nam Y, Kim MO, Kim EY, Park J, Kim DH. Computer-aided detection of metastatic brain tumors using magnetic resonance black-blood imaging. Invest Radiol 2013;48:113-9. [Crossref] [PubMed]
11. Mugler JP 3rd. Optimized three-dimensional fast-spin-echo MRI. J Magn Reson Imaging 2014;39:745-67. [Crossref] [PubMed]
12. Luo X, Yang Y, Yin S, Li H, Zhang W, Xu G, et al. False-negative and false-positive outcomes of computer-aided detection on brain metastasis: Secondary analysis of a multicenter, multireader study. Neuro Oncol 2023;25:544-56. [Crossref] [PubMed]
13. Tomura N, Narita K, Takahashi S, Otani T, Sakuma I, Yasuda K, Nishii T, Watarai J. Contrast-enhanced multi-shot echo-planar FLAIR in the depiction of metastatic tumors of the brain: comparison with contrast-enhanced spin-echo T1-weighted imaging. Acta Radiol 2007;48:1032-7. [Crossref] [PubMed]
14. Sun L, Luan S, Ye X, Chen J, Shi J, Zhu H, Dong H, Tao G, Liu X, Zhu L, Yu H. Combination of Contrast-enhanced FlAIR and Contrast-enhanced T1WI: A Quick and Efficient Method in Detecting Brain Metastases of Lung Cancers. Curr Med Imaging 2024;20:e15734056288838. [Crossref] [PubMed]
15. Sui H, Li J, Liu L, Lv Z, Zhang Y, Dai Y, Mo Z. Accelerating Knee MRI: 3D Modulated Flip-Angle Technique in Refocused Imaging with an Extended Echo Train and Compressed Sensing. J Pain Res 2022;15:577-90. [Crossref] [PubMed]
16. Li G, Wu D, Xu Z, Zuo X, Li X, Chang S, Dai Y. Evaluation of an accelerated 3D modulated flip-angle technique in refocused imaging with an extended echo-train sequence with compressed sensing for imaging of the knee: comparison with routine 2D MRI sequences. Clin Radiol 2021;76:158.e13-8. [Crossref] [PubMed]
17. Bae YJ, Choi BS, Lee KM, Yoon YH, Sunwoo L, Jung C, Kim JH. Efficacy of Maximum Intensity Projection of Contrast-Enhanced 3D Turbo-Spin Echo Imaging with Improved Motion-Sensitized Driven-Equilibrium Preparation in the Detection of Brain Metastases. Korean J Radiol 2017;18:699-709. [Crossref] [PubMed]
18. Cao X, Lv K, Yin X, Cao Y, Xu S, Feng Z, Han Y, Tang Y, Geng D, Zhang J. Preoperative Assessment of Blood Vessels and Intratumoral Microbleeds in Brain Tumors Based on a 3D Contrast-Enhanced T(1) -Weighted Flow-Sensitive Black-Blood Sequence. J Magn Reson Imaging 2023;57:1543-51. [Crossref] [PubMed]
19. Jun C, Shuhua L, Xue Z, Chunqing B, Mingli H. Application of motion-sensitized driven equilibrium based black blood 3D TSE sequence in the detection of brain metastases. Magn Reson Imaging 2022;93:145-8. [Crossref] [PubMed]
20. Chen M, Wang P, Guo Y, Yin Y, Wang L, Su Y, Gong G. The effect of time delay for magnetic resonance contrast-enhanced scan on imaging for small-volume brain metastases. Neuroimage Clin 2022;36:103223. [Crossref] [PubMed]
21. Kushnirsky M, Nguyen V, Katz JS, Steinklein J, Rosen L, Warshall C, Schulder M, Knisely JP. Time-delayed contrast-enhanced MRI improves detection of brain metastases and apparent treatment volumes. J Neurosurg 2016;124:489-95. [Crossref] [PubMed]
22. Du S, Gong G, Chen M, Liu R, Meng K, Yin Y. The effect of time-delayed contrast-enhanced scanning in determining the gross tumor target volume of large-volume brain metastases. Radiother Oncol 2024;197:110330. [Crossref] [PubMed]

(English Language Editor: J. Gray)