It has been suggested that an inadequate blood supply caused by uterine fibroids may lead to decreasing fertility. Therefore, a quantitative evaluation of blood flow in the uterus might be a good tool for infertility treatments. For the first step, the ability to perform arterial spin labeling (ASL)-MRI in pelvic organs was examined by measuring blood flow in the uterine muscle layer.
Results
Three normal volunteer women, seven patients with one uterine fibroid and four patients treated with GnRH analogue for uterine fibroids, were enrolled in this study. Perfusion of normal uterine myometrium was examined using non-enhanced ASL-MRI. The region of interest was set in the uterine muscle layer, with a point in the iliopsoas or gluteus muscle. The ASL perfusion index was calculated as (ASL value in uterus—ASL value in iliopsoas/gluteus muscle). The ASL perfusion indexes in the secretory phase of all 3 volunteers were significantly lower than the indexes in the proliferative phases (P < 0.05). In patients with fibroids, all three types of fibroids (subserosal, intramural and submucosal types) were included. In seven patients harboring a single uterine fibroid, the ASL perfusion indexes of myometrium on the fibroid-positive side increased 4.9 fold compared with that of the fibroid-negative side. With GnRH analogue treatment, ASL perfusion in myometrium decreased to 39% on average (P < 0.05).
Conclusion
We utilized the ASL-MRI technique to evaluate perfusion of uterine myometrium. For clinical use, an inadequate blood supply caused by uterine fibroids is known to lead to decreasing fertility. The ASL-MRI technique might be useful to evaluate blood supply as a quantitative measurement of fertility in patients with uterine fibroids.
In the uterus, adequate blood flow is important to maintain fertility, while vascular abnormalities are associated with clinical symptoms such as abnormal bleeding, implantation failure and gestation discontinuation (Deligdish and Loewenthal 1970; Buttram and Reiter 1981). In fact, it has been suggested that an inadequate blood supply caused by uterine fibroids may lead to decreasing fertility (Ng and Ho 2002). Therefore, a quantitative evaluation of blood flow in the uterus might be a good tool for infertility treatments. To date, ultrasound techniques, i.e. laser Doppler fluxmetry and power Doppler (Gannon et al. 1997; Raine-Fenning et al. 2004a) have been examined to evaluate blood supply, but a standardized method has not been established.
Recently, arterial spin labeling (ASL) techniques have been examined to assess tissue perfusion without contrast media. ASL is a Magnetic Resonance (MR) perfusion method using magnetic-labeled arterial blood water as an endogenous tracer. ASL is thought to be especially useful for evaluating cerebral blood flow (Telischak et al. 2015), but the use of ASL to evaluate blood flow in pelvic organs has not been established. This non-invasive technique might be useful to evaluate blood flow in the uterus to improve the outcome of infertility treatments.
In the present study, as a first step, we used the ASL technique to examine blood flow in the uterine muscle layer in three separate groups: healthy volunteers, patients with uterine fibroids, and patients with fibroids before and after treatment with GnRH analogues (GnRHa).
Methods
Subjects
From August 2012 to September 2013, non-enhanced MR perfusion studies using ASL were performed on volunteers and patients with fibroids as part of the routine pre-operative MRI examination at Yaesu Clinic. This study was approved by the institutional reviewed board, and informed consent was obtained from the patients for the examinations.
Study design
MR perfusion protocols
Uterus perfusion measurements were acquired in an axial slice on a 3 T Philips Achieva (Philips Medical Systems, Best, the Netherlands) with 32ch Cardiac Torso coil using a FAIR single shot Turbo spin-echo acquisition scheme. Briefly, labeling pulse was positioned to cover bilateral common iliac arteries. From 1 to 3 s after labeling, signal intensities were measured at uterus. Detail Parameters were as followed; repetition time/echo time 3600/5.9 ms, echo train length 21, matrix 128 × 96, FOV 230 × 230 mm2, imaging/tagging slice thickness 10/200 mm. Imaging time for each inversion time was 90 s. The slice which exhibited the maximum area of uterine fibroid was used.
Determination of the optimal scanning time of ASL and a region of interest in the pelvis
In a pilot study, we examined various scan timings of the ASL signal after labeling. As shown in Fig. 1, the ASL signals in the uterus were observed at 1.0 s after labeling, and increased until 1.5 or 2.0 s, then attenuated. We set the optimal scanning time as 1.5 s after labeling, which was used for subsequent studies.
Fig. 1
Sequential ASL signals in the uterus. ASL signals in the uterus were observed at 1.0 ms after labeling, and increased until 1.5 or 2.0 ms, then attenuated. We set the optimal scanning time as 1.5 ms after labeling
A visual evaluation of the colored perfusion map calculated on the MR scanner using software provided by the MR vendor and a region of interest (ROI) analysis were performed by two experienced radiologists. The ROI consisted of a 5 mm2 diameter circle. We set 4–5 points of the ROI in the uterine muscle layer, a point in the iliopsoas/gluteus muscle, which was shown in the same image, and a point in the air as background. The perfusion index was calculated as (ASL value in uterus—ASL value in iliopsoas/gluteus muscle). To compare data of different time points, data were corrected by adjusting background ASL value. We examined various subjects as described below.
Examination 1) Physiological status of perfusion in the myometrium
To examine the effect of hormonal status, the ASL perfusion indexes of uterine muscle were obtained in proliferative and secretory phase, respectively, from three healthy volunteers who had regular menstrual cycles and normal uterus and ovaries (evaluated via MRI). The uterine cavity was divided into five equal parts in the image, and five points of ROI were set at muscle layer just outside of the junctional zone. ROIs were set at the same locations in both proliferative and secretory phases.
Examination 2) Perfusion in the myometrium of patients with fibroids
To examine the effect of uterine fibroids, we evaluated seven patients who each possessed one fibroid. In this examination, all subjects were examined during the proliferative phase. We set 2 ROIs at muscle layer, just outside of the junctional zone, on the fibroid-positive side and -negative side, respectively. We then determined the average ASL perfusion index on each side, and compared the averages for the fibroid-positive and -negative side.
Examination 3) Perfusion in myometrium around the fibroid before and after GnRHa treatment
Four patients with uterine fibroids treated with GnRHa were enrolled in this study (one patient with a single fibroid and three patients with multiple fibroids). In this examination, the subjects underwent their first MRI during the proliferative phase, and ROIs were set at myometrium, just outside of the junctional zone. After two or three rounds of GnRHa, the patients underwent a second MRI. ROIs were set at the same locations in myometrium before and after GnRHa treatment, and ASL perfusion indexes were compared.
Statistical analysis
Statistical analysis was performed by using JMP Pro 11 software (SAS institute Inc., Cary, NC). ASL perfusion indexes are shown as the mean + S.E.M relative to an adjusted value of 1.0 for the mean value of the other one. The data were analyzed by the student’s t test for paired comparison. A P value of less than 0.05 was considered significant.
Results
Examination 1) Physiological status of perfusion in myometrium
The ASL perfusion indexes of myometrium in normal uterus were measured in the proliferative and secretory phases. Signal intensities during the secretory phase were evaluated as a relative ratio to that of the proliferative phase. As shown in Fig. 2, the ASL perfusion indexes in the secretory phase of all three volunteers were significantly lower than the indexes in the proliferative phase (P < 0.05).
Fig. 2
ASL perfusion indexes of normal myometrium in the proliferative and secretory phases. Three healthy volunteers who had normal MRI appearances in the uterus and ovaries with regular menstrual cycles were enrolled in this study. ASL perfusion indexes were calculated as (ASL value in uterus—ASL value in iliopsoas/gluteus muscle) and data were normalized with ASL values in background. Data were shown as the mean + S.E.M relative to an adjusted value of 1.0 for the mean value of the proliferative phase. ASL perfusion indexes in the secretory phase were significantly lower than the levels in the proliferative phase (*P < 0.05)
Examination 2) The effect of uterine fibroids on perfusion in normal myometrium
Figure 3 shows an example of the ROI in this examination. ROIs were set at uterine muscle layer just outside of the junctional zone. This patient had an intramural fibroid arising from the uterine posterior wall. We set 5 points of ROI (ROI①: uterine right posterior wall beneath the fibroid, ROI②: left uterine posterior wall beneath the fibroid, ROI③: right uterine anterior wall, ROI④: left uterine anterior wall and gluteus muscle). We defined ROI① and ② as the fibroid-positive side, and ROI③ and ④ as the fibroid-negative side. As shown in Table 1, the ASL data on the fibroid-positive side were comparable (56.6 and 56.7), and those taken on the fibroid negative side were 2.7 and 8.7. ASL perfusion indexes on the fibroid-positive side were around 9.9 times higher than those measured on the fibroid-negative side.
Fig. 3
Region of interest of ASL perfusion in a 41-year-old patient with fibroid. The fibroid was subserosal-mural type arising from uterine posterior wall. ROI①: right uterine posterior wall beneath the fibroid, ROI②: left uterine posterior wall beneath the fibroid, ROI③: uterine anterior wall, ROI④: uterine anterior wall, and gluteus muscle. We defined ROI① and ROI② as fibroid positive side, and ROI③ and ROI④ as fibroid negative side. ASL perfusion indexes were calculated as ASL value in uterus—ASL value in gluteus muscle
Table 1
ASL values in a 41-year-old patient with fibroid
ROI①
ROI②
ROl③
ROI④
Gluteus muscle
ASL measured value
707.9
708
654
660
651.3
ASL perfusion indexes (ROI-muscle)
56.6
56.7
2.7
8.7
ASL perfusion indexes were calculated as ASL value in uterus—ASL value in gluteus muscle
Table 2 shows the summary of patients’ background and ASL values. All three types of fibroids (subserosal, intramural and submucosal type) were included. Higher ASL perfusion indexes (mean value 4.9 fold increases) were demonstrated on the fibroid-positive side than on the fibroid-negative side (Fig. 4).
Table 2
Characteristics and ASL perfusion indexes of patients with uterine fibroids
Age
Size of uterine fibroid (mm)
Type of fibroid
ASL perfusion inc (ROI-muscle)
Fibroid (+)
Fibroid (−)
(+)/(−) Fold
Case 1
41
106 × 56
Subserosal
56.7
5.7
9.9
Case 2
39
68 × 59
Intramural
62.4
7.1
8.6
Case 3
37
68 × 51
Intramural
13.9
1.7
7.8
Case 4
42
70 × 65
Subserosal
175
96.4
1.8
Case 5
40
44 × 36
Intramural
21.1
12.2
1.7
Case 6
38
35 × 28
Submucosal
23.4
4.9
4.6
Case 7
47
72 × 52
Intramural
199
127
1.5
Patients who possessed one fibroid underwent MRI during the proliferative phase. Two ROIs of myometrium on the fibroid-positive side and fibroid-negative side were set. ASL perfusion indexes were calculated as ASL value in uterus—ASL value in iliopsoas/gluteus muscle
Fig. 4
Comparison of ASL perfusion indexes on the fibroid-positive or -negative side. ASL perfusion indexes were calculated as ASL value in uterus—ASL value in iliopsoas/gluteus muscle. Data are shown as the mean ± S.E.M relative to an adjusted value of 1.0 for the mean value of the fibroid negative side. Perfusion of the fibroid-positive side was significantly higher than that of the fibroid-negative side (*P < 0.05)
Examination 3) The effect of GnRHa treatment on perfusion in myometrium around fibroids
The effect of GnRHa on ASL perfusion in myometrium was evaluated. ASL data obtained from 4 ROIs were averaged. The fibroid size and ASL perfusion index of all patients decreased after GnRHa treatment (Table 3). The percentage change in the ASL perfusion index ranged from 6.8 to 73%, giving a decrease to 39% of basal levels on average (P < 0.05).
Table 3
Characteristics and ASL perfusion indexes of patients with fibroids treated with GnRH analogue
Age
Diameter of fibroid (mm)
Before ⇒ after GnRH treatment (reduction ratio)
Type of fibroid
Calibrated ASL perfusion index
(%) After ⇒ before
Case 1
39
86 ⇒ 80 (7%)
72 ⇒ 72 (0%)
56 ⇒ 42 (25%)
52 ⇒ 52 (0%)
Subserosal
Subserosal
Subserosal
Intramural
15% (8.0 ⇒ 1.2)
Case 2
39
64 ⇒ 64 (0%)
40 ⇒ 25 (38%)
Intramural
intramural
63% (9.9 ⇒ 6.3)
Case 3
44
18 ⇒ 16 (11%}
Submucosal
6.8% (12.9 ⇒ 0.88)
Case 4
45
96 ⇒ 70 (27%)
42 ⇒ 30 (29%)
Intramural
Intramural
73% (11.7 ⇒ 8.6)
Mean 39%
The subjects underwent their first MRI during the proliferative phase. After two or three rounds of GnRH analogue administration, the patients underwent a second MRI. ROIs were set at the same locations in myometrium beneath fibroids. ASL perfusion indexes before and after GnRH analogue treatments were calculated as (ASL value in uterus—ASL value in iliopsoas/gluteus muscle). Data were calibrated with ASL values in background
There is no correlative relationship between the decreasing ratio of fibroid size and change in ASL index before and after GnRHa.
Discussion
In the present study, we utilized the ASL-MRI technique to evaluate perfusion of myometrium, and showed the ability to perform the technique by the findings that (1) normal myometrium perfusion during the secretory phase was significantly lower than that of the proliferative phase, (2) perfusion of myometrium on the fibroid-positive side was significantly higher than that of the fibroid-negative side and (3) with GnRHa treatment, ASL perfusion in myometrium decreased to 39% on average.
The two most common methods for measuring perfusion with MRI are based on dynamic susceptibility contrast (DSC) and arterial spin labeling (ASL) (Wolf and Detre 2007). As MR perfusion using ASL does not need contrast medium, and ASL signals disappear in few seconds, it is a non-invasive examination and can be performed repeatedly even in renal failure patients (Telischak et al. 2015). ASL MRI is useful to diagnose various cerebrovascular diseases which cause a disorder of perfusion (Detre et al. 2012), and it has also been used to examine normal brain function (Wang et al. 2003; Telischak et al. 2015). We applied ASL-MRI to the female pelvis in order to evaluate perfusion in the myometrium. In our study, ASL perfusion in normal myometrium during the secretory phase was lower than in the proliferative phase. In accordance with our findings, ultrasound techniques designed to measure blood flow also revealed that myometrium showed lower tissue blood flow during the secretory phase compared to the proliferative phase by laser Doppler fluxmetry and power Doppler angiography (Gannon et al. 1997; Raine-Fenning et al. 2004a). During the proliferative phase, small vessels, including the spiral and straight arterioles in the inner myometrium develop (Blackwell and Fraser 1988). Additionally, the increase of serum estradiol levels and the decrease in microvascular resistance during the proliferative phase are known to increase blood flow (de Vries et al. 1990).
Secondly, we used the ASL-MRI technique to demonstrate the influence of fibroids in myometrium perfusion. Doppler ultrasound examination revealed that in women with fibroids, blood flow in the uterine artery was higher than in women without fibroids (Sladkevicius et al. 1996; Kurjak et al. 1992; Alataş et al. 1997). However, it is difficult to evaluate perfusion of local myometrium. Using the ASL technique, we were able to evaluate perfusion in local regions by setting the region of interest (ROI). In our study, we evaluated the perfusion of myometrium beneath the endometrium, and found that all three types, submucosal, intramural, and subserosal fibroids, increased blood perfusion in the adjacent myometrium. Importantly, the ASL signal in the myometrium on the fibroid-positive side was higher than that of fibroid-negative side, implying that fibroids may cause an imbalance of blood distribution in the myometrium.
In recent studies, fibroids have been shown to exhibit high levels of pro-angiogenic factors, which cause an increase in angiogenesis and vascular density in the normal myometrium surrounding fibroids (Tal and Segars 2014). Vascular abnormalities due to the presence of fibroids might be associated with clinical symptoms such as abnormal bleeding, implantation failure and gestation discontinuation (Deligdish and Loewenthal 1970; Buttram and Reiter 1981). In another report, endometrial and sub-endometrial perfusion was impaired in women with unexplained subfertility (Raine-Fenning et al. 2004b). These data imply that an adequate blood supply to the endometrium is required for implantation. In some patients with fibroids, one of the symptoms is infertility. Meta-analyses showed that submucosal but not subserosal fibroids have a negative effect on pregnancy rate, while conclusions regarding intramural lesions have been conflicting (Somigliana et al. 2007; Pritts et al. 2009). Therefore, subserosal and intramural fibroids are not routinely removed for infertility treatment. However, in some patients, removal of fibroids could improve pregnancy rate and decrease miscarriage (Brady et al. 2013). Gynecologists need to accurately assess the effectiveness of surgical intervention of uterine fibroids. In our study, although it included a small numbers of patients, all three types of fibroids caused an imbalance of perfusion in uterine myometrium. Moreover, there was no correlation between ASL perfusion indexes and types or the size of fibroid. Further study is needed to investigate the relationship between uterine perfusion and reproductive outcomes.
Finally, we confirmed that the ASL-MRI technique could detect the decrease of myometrial blood supply with GnRHa treatment. The use of GnRHa prior to fibroid surgery reduces fibroid volume, and also improves pre-operative anemia, and intra-operative blood loss (Lethaby et al. 2001). GnRHa treatment led to a decrease in micorovascular density in tissue of uterine fibroids (Khan et al. 2010), and a decrease in angiogenesis-related factors such as VEGF, bFGF, and PDGF (Di Lieto et al. 2005). Consistent with our ASL-MRI finding, there are some reports that GnRHa treatment reduced the uterine blood flow evaluated with Doppler ultrasound (Matta et al. 1988; Aleem and Predanic 1995). However, in the present study, there is no correlative relationship between the decreasing ratio of fibroid size and change in ASL index before and after GnRHa, suggesting that other factors might be involved in decreasing fibroid size.
In the present study, we could not compare ASL signal data with perfusion data obtained by using contrast agents, due to the clinical limitations in which contrast agents were not used for patients with fibroids in our clinic. In the brain MRI, it appears that blood flow can be estimated correctly in the grey matter of the brain by using ASL techniques (Wintermark et al. 2005). Further study is needed to determine whether ASL signal actually correlates to the blood flow in the uterus.
Conclusion
We evaluated myometrium perfusion using ASL-MRI. This technique has successfully been performed in the uterus to assess tissue perfusion without contrast media. Future uses of this technique may include fibroid analysis in the investigation into infertility and assessment of malignant tumors.
Declarations
Authors’ contributions
NT: Data management, Manuscript writing. OY: Data analysis, Protocol/project development, Data management, and Manuscript writing. OH: Data collection. EM: Data analysis and Manuscript writing. MN: Data management, Analysis. MH: Data management, Protocol/project development. MH, TF: Data collection, Manuscript editing. KK: Data collection. SS: Manuscript editing, Protocol/project development. YO: Data collection, Manuscript editing, Protocol/project development. All authors read and approved the final manuscript.
Acknowledgements
The authors thank Dr. Heather M. Martinez for her helpful discussion and critical reading of the manuscript. Dr. Yasushi Hirota, Dr. Tetsuya Hirata, Dr. Masashi Takamura and Dr. Gen Izumi for their great help to this work.
Competing interests
The authors declare that they have no competing interests.
Ethical approval
This study involved human participants. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional (The University of Tokyo and Yaesu Clinic) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent
Informed consent was obtained from all individual participants included in the study (Yaesu Clinic).
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Authors’ Affiliations
(1)
Department of Obstetrics and Gynecology, The University of Tokyo
(2)
Department of Obstetrics and Gynecology, The University of Toyama
(3)
Department of Radiology, The University of Tokyo
(4)
Philips Electronics Japan
(5)
Department of Radiology, The University of Juntendo
References
Alataş G, Aksoy E, Akarsu C et al (1997) The effect of uterine volume on uterine artery Doppler velocimetry in the myomatous state. Gynecol Obstet Invest 43:55–59View ArticlePubMedGoogle Scholar
Aleem FA, Predanic M (1995) The hemodynamic effect of GnRH agonist therapy on uterine leiomyoma vascularity: a prospective study using transvaginal color Doppler sonography. Gynecol Endocrinol 9:253–258View ArticlePubMedGoogle Scholar
Blackwell PM, Fraser IS (1988) A morphometric and ultrastructural study of the microvessels of the functional zone of normal human endometrium with some notions on possible secretory functions of the endothelial cells. Asia Ocean J Obstet Gynaecol 14:233–250View ArticleGoogle Scholar
Brady PC, Stanic AK, Styer AK (2013) Uterine fibroids and subfertility: an update on the role of myomectomy. Curr Opin Obs Gynecol 25:255–259View ArticleGoogle Scholar
de Vries K, Lyons EA, Ballard G et al (1990) Contractions of the inner third of the myometrium. Am J Obstet Gynecol 162:679–682View ArticlePubMedGoogle Scholar
Di Lieto A, De Falco M, Mansueto G et al (2005) Preoperative administration of GnRH-a plus tibolone to premenopausal women with uterine fibroids: evaluation of the clinical response, the immunohistochemical expression of PDGF, bFGF and VEGF and the vascular pattern. Steroids 70:95–102View ArticlePubMedGoogle Scholar
Gannon BJ, Carati CJ, Verco CJ (1997) Endometrial perfusion across the normal human menstrual cycle assessed by laser Doppler fluxmetry. Hum Reprod 12:132–139View ArticlePubMedGoogle Scholar
Khan KN, Kitajima M, Hiraki K et al (2010) Changes in tissue inflammation, angiogenesis and apoptosis in endometriosis, adenomyosis and uterine myoma after GnRH agonist therapy. Hum Reprod 25:642–653View ArticlePubMedGoogle Scholar
Kurjak A, Kupesic-Urek S, Miric D (1992) The assessment of benign uterine tumor vascularization by transvaginal color Doppler. Ultrasound Med Biol 18:645–649View ArticlePubMedGoogle Scholar
Lethaby A, Vollenhoven B, Sowter M (2001) Pre-operative GnRH analogue therapy before hysterectomy or myomectomy for uterine fibroids. Cochrane Database Syst Rev (2):CD000547Google Scholar
Matta WH, Stabile I, Shaw RW, Campbell S (1988) Doppler assessment of uterine blood flow changes in patients with fibroids receiving the gonadotropin-releasing hormone agonist Buserelin. Fertil Steril 49:1083–1085View ArticlePubMedGoogle Scholar
Ng EHY, Ho PC (2002) Doppler ultrasound examination of uterine arteries on the day of oocyte retrieval in patients with uterine fibroids undergoing IVF. Hum Reprod 17:765–770View ArticlePubMedGoogle Scholar
Pritts EA, Parker WH, Olive DL (2009) Fibroids and infertility: an updated systematic review of the evidence. Fertil Steril 91:1215–1223View ArticlePubMedGoogle Scholar
Raine-Fenning NJ, Campbell BK, Kendall NR et al (2004a) Quantifying the changes in endometrial vascularity throughout the normal menstrual cycle with three-dimensional power Doppler angiography. Hum Reprod 19:330–338View ArticlePubMedGoogle Scholar
Raine-Fenning NJ, Campbell BK, Kendall NR et al (2004b) Endometrial and subendometrial perfusion are impaired in women with unexplained subfertility. Hum Reprod 19:2605–2614View ArticlePubMedGoogle Scholar
Sladkevicius P, Valentin L, Marsál K (1996) Transvaginal Doppler examination of uteri with myomas. J Clin Ultrasound 24:135–140View ArticlePubMedGoogle Scholar
Somigliana E, Vercellini P, Daguati R et al (2007) Fibroids and female reproduction: a critical analysis of the evidence. Hum Reprod Update 13:465–476View ArticlePubMedGoogle Scholar
Tal R, Segars JH (2014) The role of angiogenic factors in fibroid pathogenesis: potential implications for future therapy. Hum Reprod Update 20:194–216View ArticlePubMedGoogle Scholar
Telischak NA, Detre JA, Zaharchuk G (2015) Arterial spin labeling MRI: clinical applications in the brain. J Magn Reson Imaging 41:1165–1180View ArticlePubMedGoogle Scholar
Wang J, Licht DJ, Jahng G-H et al (2003) Pediatric perfusion imaging using pulsed arterial spin labeling. J Magn Reson Imaging 18:404–413View ArticlePubMedGoogle Scholar
Wintermark M, Sesay M, Barbier E et al (2005) Comparative overview of brain perfusion imaging techniques. Stroke 36:e83–e99View ArticlePubMedGoogle Scholar
Wolf RL, Detre JA (2007) Clinical neuroimaging using arterial spin-labeled perfusion magnetic resonance imaging. Neurotherapeutics 4:346–359View ArticlePubMedPubMed CentralGoogle Scholar
