Susceptibility-weighted magnetic resonance imaging for the evaluation of deep infiltrating endometriosis: preliminary results

Abstract

Background

Knowledge of the precise sites of deep infiltrating endometriosis (DIE) lesions is essential for preoperative workup and treatment. Susceptibility-weighted imaging (SWI) has high sensitivity for blood products and have recently been applied in abdominal imaging.

Purpose

To determine the value of SWI in the diagnosis of DIE.

Material and Methods

Forty-three clinically suspected DIE patients with sonographically diagnosed ovarian endometriomas who had tenderness or palpable nodule(s) on rectovaginal examination were referred to pelvic magnetic resonance imaging (MRI) including SWI. Two patients were excluded from the study because of low quality of SWI series. Twenty-eight patients who were offered laparoscopic endometriosis surgery (LES) preferred medical treatment over surgical approach. Thirteen out of 41 participants had LES. Lesions were evaluated for their locations, signal intensities on T1-weighted (T1W) and T2-weighted (T2W) images, and presence of signal voids on SWI using 3T MRI and correlated with LES findings.

Results

A total of 18 endometriosis foci were laparoscopically removed from 13 patients. DIE lesions removed at laparoscopy were located at the uterosacral ligament (9/18), rectovaginal region (4/18), retrocervical region (2/18), and fallopian tubes (3/18). Eleven out of 18 (61%) DIE foci were detected by their high-signal intensities on T1W images whereas 16 out of 18 (89%) DIE foci were detected by signal voids on SWI.

Conclusion

SWI imaging with its high sensitivity to blood products, contributes to the diagnosis of DIE by depicting different phases of hemorrhage not seen by conventional MRI sequences.

Keywords

Endometriosis susceptibility-weighted imaging (SWI)deep infiltrating endometriosis (DIE)magnetic resonance imaging (MRI)

Introduction

Endometriosis is defined as the presence of functional endometrial glands and stroma outside the uterine cavity affecting 5–20% of women at reproductive age (1). Deep infiltrating endometriosis (DIE) is defined as subperitoneal invasion by endometriotic lesions that exceed 5 mm in depth (1 –3). The most common DIE sites are the retrocervical region, uterosacral ligaments, rectum, rectovaginal septum, vagina, urinary tract, and other extra-peritoneal pelvic sites (2). Some patients with extensive disease and large endometriomas may have relatively few symptoms, whereas others with small deep endometriotic implants can suffer from severe symptoms such as dysmenorrhea, dyspareunia, urinary tract symptoms, infertility, and pelvic pain.

Determining the extension of DIE is difficult with physical examination and laparoscopic exploration alone. Palpation of tender nodules in cul-de-sac in gynecologic examination has poor accuracy and is limited in predicting the disease extension in cases with pelvic adhesions (4). Laparoscopy has limitations in detecting DIE sites hidden by adhesions. Since the standard treatment for DIE is surgical excision of the endometriotic foci, knowledge of the precise sites of the lesions is essential for preoperative workup (5).

Sonography is the first choice of imaging examination in the diagnosis of ovarian endometriomas. Although transvaginal sonographic findings associated with posterior cul-de-sac endometriosis appear to be specific, rectal carcinoma or peritoneal metastasis should be excluded (6). Investigators have reported that the sensitivity of transvaginal ultrasound was best for intestinal and bladder disease location and slightly less accurate for uterosacral and rectovaginal involvement (7,8). Rectal endoscopic sonography has been recommended for detection of rectal, rectovaginal, uterosacral, and rectosigmoid DIE locations, but poor penetration limits diagnostic accuracy (9,10). Magnetic resonance imaging (MRI) has been shown to have higher accuracy in detecting DIE lesions. Addition of fat-suppressed sequences has been reported to facilitate the detection of small endometriotic foci (11,12). Small non-hemorrhagic foci in superficial peritoneal endometriosis are often not detectable with MRI (3,13). DIE lesions often have poorly defined margins and appear hypointense on T2-weighted (T2W) imaging secondary to fibrosis. Although most endometriomas can be recognized on the basis of their T1 high-signal intensity, solid subcentimeter foci of DIE can easily be overlooked. Diagnosis can be challenging because some solid DIE lesions contain only punctate foci of high-signal intensity on T1 sequences representing minimal cyclical bleeding surrounded by fibrosis and smooth muscle hypertrophy (12,14,15).

Susceptibility-weighted imaging (SWI) has higher sensitivity to blood products compared to conventional MRI sequences (16). The magnetic susceptibility effects generated by local inhomogeneity of the magnetic field are visualized as signal voids. SWI not only detects subacute blood products like methemoglobin, but also acute and chronic blood products such as deoxyhemoglobin and hemosiderin, respectively (12,17). Although it has been mostly used in neuroimaging studies (17 –20), SWI is recently applied in body imaging of liver, spleen, and prostate gland (21 –23). Only a few studies in the literature have reported their findings for the usefulness of SWI in the diagnosis of endometriomas and extra-ovarian endometriosis (12,17,24). Our aim in this study was to analyze the contribution of SWI in diagnosis of DIE by comparing imaging and laparoscopic findings.

Material and Methods

Patients

This prospective study was approved by the Ethics Committee of our institution. Informed written consent was obtained from all individual participants included in the study. Patients were referred from the Obstetrics and Gynecology Department between April 2013 and March 2015.

Forty-three clinically suspected DIE patients with sonographically diagnosed ovarian endometriomas who had tenderness or palpable nodule(s) on rectovaginal examination were submitted to pelvic MRI including SWI. Two patients were excluded from the study because of low quality of SWI series. Thirteen out of 41 patients had laparoscopic endometriosis surgery (LES) to obtain definitive diagnosis confirmed by histology, and treatment of pelvic pain. Twenty-eight patients who were offered LES preferred medical treatment over surgical approach. Mean age of the group that underwent LES was 31.2 years. (age range, 26–41years). None of the patients had previous endometriosis surgery.

Imaging protocol

Examinations were performed regardless of the stage of the menstrual cycle. MRI was performed with moderate repletion of the bladder. No bowel preparation or intraluminal opacification of bowel or vagina were used. Images were acquired on a 3T MR scanner (Magnetom Verio, Siemens, Erlangen, Germany) equipped with an 8-channel torso coil. Standard pelvic MRI protocol included three projections: axial, sagittal, and coronal T2W and T1-weighted (T1W) pulse sequences before and after fat suppression. The additional time for SWI sequence was approximately 3 min. SWI sequences consisting of magnitude and phase images were obtained in axial and sagittal planes. Postprocessing was applied to the magnitude images multiplied with a phase mask generated from the filtered phase data. Acquisition parameters for SWI were: TR/TE, 28/20 ms; flip angle, 15°; field of view (FOV) 35 cm; matrix, 202 × 320; slice thickness, 1.8 mm; gap, 0.36 mm; and scan time, 3 min 11 s.

Image interpretation

All images were reviewed on a picturing archiving and communication system (Novapacs, Novarad Corporation, American Fork, UT, USA) by two radiologists with 10 and 15 years of expertise in body MRI. Patients underwent laparoscopy following the completion of imaging and the histopathological diagnosis was confirmed on surgical specimens. The radiologists were blinded to LES and the histopathological results. All patients were reviewed independently by the radiologists and any discrepancy was resolved by consensus. Lesions were evaluated regarding their location, signal intensity on T1W and T2W images, and presence of signal void on SWI. The lesion signal intensity was considered hypointense or hyperintense compared to the signal intensity of adjacent pelvic muscles.

Results

Thirteen out of 41 patients were treated with LES. A total of 18 endometriosis foci were removed laparoscopically from the 13 patients. DIE lesions removed at laparoscopy were located at the uterosacral ligament (9/18), rectovaginal region (4/18), retrocervical region (2/18), and fallopian tubes (3/18). Nine patients had both ovarian endometriomas and DIE foci whereas four patients had only ovarian endometriomas. Eleven out of 18 (61%) DIE foci were detected by their high-signal intensity on T1W images whereas 16 out of 18 (89%) DIE foci were detected by signal void on SWI. Patient complaints and MRI findings are summarized in Table 1.

Table 1.

Clinical and MRI findings of patients.

Patient	Age (years)	Complaint	Ovarian endometrioma	DIE site	High intensity on T1W	Intensity on T2W	SWI
1	32	Dysmenorrhea, dyspareunia	Bilateral	Negative	–	–	–
2	41	Pelvic pain	Bilateral	Negative	–	–	–
3	26	Infertility, dyspareunia	Bilateral	Negative	–	–	-
4	36	Infertility	Bilateral	Negative	–	–	–
5	31	pelvic pain, dyspareunia	Bilateral	Right uterosacral ligament	Negative	Low	Negative
				Left uterosacral ligament	Positive	Low	Signal void
6	33	Pelvic pain, dyspareunia	Right	Rectovaginal	Negative	Low	Negative
7	27	Pelvic pain, dysmenorrhea	Left	Right uterosacral ligament	Positive	High	Signal void
				Left uterosacral ligament	Negative	Low	Signal void
8	29	Dysmenorrhea, dyspareunia	Bilateral	Retrocervical	Positive	High	Signal void
				Rectovaginal	Negative	Low	Signal void
9	26	Infertility, dyspareunia	Left	Left uterosacral ligament	Negative	Low	Signal void
				Left fallopian tube	Positive	Low	Signal void
10	28	Dysmenorrhea, infertility	Bilateral	Right uterosacral ligament	Positive	Low	Signal void
				Left uterosacral ligament	Positive	Low	Signal void
11	37	Pelvic pain, infertility	Bilateral	Fallopian tubes	Positive	Low	Signal void
				Rectovaginal	Negative	Low	Signal void
12	33	Dysmenorrhea, dyspareunia	Left	Rectovaginal	Negative	Low	Signal void
13	27	Dysmenorrhea, dyspareunia	Bilateral	Right uterosacral ligament	Positive	Low	Signal void
				Left uterosacral ligament	Positive	Low	Signal void
				Fallopian tubes	Positive	Low	Signal void
				Retrocervical	Positive	High	Signal void

Six out of nine uterosacral ligament endometriosis foci had high-signal intensity on T1W images whereas signal void were detected in eight of these lesions (Figs. 1 and 2). Two retrocervical endometriosis foci were hyperintense on T1W and T2W images and both had signal voids on SWI (Fig. 3). None of the rectovaginal endometriosis foci had high-signal intensity on T1W images and signal void were present in three out of four of these lesions with SWI (Fig. 4). Three patients had dilated fallopian tubes with hemorrhagic content, and signal void foci on the surface of the tubes were seen with SWI which were not detected on conventional MR images (Fig. 5). Site mapping of MRI-defined DIE foci in LES group with laparoscopic correlation are summarized in Table 2.

Fig. 1.

Bilateral ovarian endometriomas and endometrial implant on the uterosacral ligament in a 27-year-old woman with dysmenorrhea and dyspareunia. Axial fat-suppressed T2W (a) and T1W (b) images show fluid in cul-de-sac and posterior displacement of “kissing ovaries”. Bilateral endometriomas have shading on T2W image (a). Hemorrhagic endometrial implants show high intensity on T1W image (arrow) (b) and signal voids on SWI (arrow) (c). Note the punctuate and curved linear signal voids along the cyst walls revealing hemosiderin deposition (c).

Fig. 2.

Left ovarian endometrioma and endometriotic involvement of the uterosacral ligament in a 27-year-old woman with pelvic pain and dysmenorrhea. Sagittal T2W image shows low-intensity nodular thickening of the proximal uterosacral ligament (dashed arrow) (a). SWI shows punctuate signal voids within the fibrotic nodule (dashed arrows) (b, d). The lesion is not visible on coronal fat-suppressed T1W image, and high-intensity endometrioma is present (c). Photograph taken during laparoscopy shows fibrotic nodular focus on the uterosacral ligament (dashed arrow) (e).

Fig. 3.

Endometriotic involvement of the retrocervical area in a patient with dyspareunia. The lesion is hyperintense on T1W and T2W fat-suppressed images (arrows) (a, b) and the foci have signal voids on SWI (arrows) (c).

Fig. 4.

Endometrial focus at the rectovaginal area in a 29-year-old woman with dyspareunia and dysmenorrhea presented with bilateral endometrioma and tenderness on rectovaginal examination. Axial T1W fat-suppressed image shows no high-intensity hemorrhagic focus (a), whereas prominent punctuate signal voids are present in SWI (arrows) (b, c).

Fig. 5.

Bilateral hematosalpinges in a 37-year-old woman with pelvic pain and infertility. T1W and T2W hyperintensity within dilated fallopian tubes consistent with subacute hemorrhagic content suggesting endometriosis (a, b). Prominent multiple endometrial foci on the serosal surfaces of the tubes seen as signal voids on SWI (arrow) (c).

Table 2.

Site mapping of MRI-defined DIE foci with laparoscopic correlation.

DIE sites	MRI		LES
	High intensity on T1W	Signal void on SWI
Uterosacral ligaments	6	8	9
Rectovaginal region	0	3	4
Retrocervical region	2	2	2
Fallopian tubes	3	3	3
Total (n)	11	16	18

Discussion

Endometriosis is an important cause of chronic pelvic pain and infertility associated with a spectrum of imaging findings ranging from subcentimeter implants to cystic collections and fibrotic nodules. Laparoscopy is the gold standard for evaluation and treatment of DIE lesions. MRI can accurately detect multiple sites of DIE and evaluate areas otherwise inaccessible by laparoscopy in the presence of dense adhesions without the use of ionizing radiation (2,3).

SWI is a relatively recently developed MRI technique where the image contrast is mainly based on magnetic susceptibility effects. The natural property of tissues reflects the magnetic response of a substance to an external magnetic field. The difference in susceptibility between substances leads to local magnetic field inhomogeneities, which results in faster T2* relaxation and signal loss on MR sequences sensitive to T2* effects. Ferromagnetic substances such as iron, paramagnetic substances including blood products and MR contrast agents, and diamagnetic substances such as calcium exhibit a strong susceptibility effect and are more easily detected using SWI (25). SWI employs a high resolution 3D gradient-recalled echo sequence and by utilizing both magnitude and phase information, the sensitivity in the detection of identifying blood products increase (25 –27). Another advantage of SWI over the current gradient echo sequences is the ability to differentiate between blood products and calcification. Blood products are paramagnetic and increase the local magnetic field, whereas calcium is diamagnetic and reduces the magnetic field, thus appearing different on phase images (28). Although used mainly in neuroimaging studies, SWI images have recently been applied in abdominal imaging.

In our study hyperintensity on T1W images were detected in 61% (11/18) laparoscopically removed DIE foci. With the addition of SWI to conventional MRI, detection rate of foci increased to 89% (16/18) due to the signal voids. The number of foci was higher and more prominent on SWI compared to T1W imaging. Because the basis for MRI contrast of these DIE lesions is susceptibility-related signal-intensity loss due to their ferromagnetic composition, it is not surprising that SWI being sensitive to these signal-intensity properties and a higher field strength with 3T would increase their detection. Our results correlate well with two studies reported by Takeuchi et al. (12,17). In their first study the contribution of SWI to the diagnosis of endometrioma by depicting hemosiderin deposition in the endometrial cyst wall was reported, and the number of implants already seen on classical MRI sequences increased with SWI (17). Takeuchi et al. also reported their SWI findings in a case series of eight extra-ovarian endometriosis namely bladder, bowel, and abdominal wall endometriosis where addition of SWI to conventional MRI increased their detection rate of lesions from 50% to 100% (12).

The uterosacral ligaments are considered the most common location for solid endometriosis. The ligaments may be involved along their entire length, especially the portion along the posterolateral margin of the cervix. Irregularity, asymmetry, and nodular thickening of the ligaments suggest DIE. Retroflexed uterus can limit the visualization of these ligaments (29). Diagnosis with routine MRI has a reported sensitivity and specificity of 69% and 90%, respectively (3). In our study, half of the lesions were located at the uterosacral ligament. Six of these nine lesions were diagnosed with high intensity on T1W images whereas eight foci had signal voids on SWI.

Rectovaginal endometriosis often remains undiagnosed for long periods of time because of its anatomical relationship with nerves. Extension of the disease in this anatomic area does not necessarily correlate with the severity of pain or infertility (30). Infiltrating fibrous lesions at this region together with the collapsed vaginal walls and fibrofatty structure of rectovaginal septum limit the diagnostic accuracy of conventional MR sequences. In a meta-analysis study of accuracy of MRI in DIE, sensitivity and specificity for the rectovaginal region was reported as 82% and 77%, respectively (2). In our study, none of the four rectovaginal endometriosis foci had high-signal intensity on T1W images and signal voids were prominent on SWI in three of these cases.

Most retroperitoneal endometriotic lesions are found in the retrocervical area. Retrocervical involvement is mostly associated with uterosacral ligament involvement (1). Pelvic MRI has a higher sensitivity and specificity for diagnosis of DIE in the pouch of Douglas (89% and 94%, respectively) (2). Our results were also in line with these studies. Both endometriotic foci were hyper-intense on T1W images reflecting subacute hemorrhage and had signal void on SWI. Loculated fluid collections in the pouch of Douglas suggesting adhesions were also noted.

Kim et al. and Gougoutas et al. reported that 30% of patients with endometriosis had tubal involvement at laparoscopy (31,32). Women with endometriosis develop hydrosalpinges secondary to endometrial implants on the surfaces of or within the tubes. Hematosalpinges may be the only finding at MRI suggestive of endometriosis. Kim et al. reported that approximately 40% of endometriosis patients with dilated tubes have hematosalpinges (31). In our study, three patients had hematosalpinges on MRI. Numerous signal void foci on the surface of the tubes were present on SWI, which were missing or very few in numbers on conventional MR images.

Two foci detected at laparoscopy had neither high-signal intensity on T1W images nor signal void on SWI. One focus removed from the uterosacral ligament was beige of color and solid macroscopically and had no blood degraded products on histopathological examination. The second focus removed from the posterior rectovaginal fornix had blood degraded products surrounded by fibrosis on histopathological examination. We suggest that the prominent susceptibility artifacts caused by the rectum limited the visibility of these lesions.

One of the limitations of our study is that not all of the patients underwent LES. Studies with larger groups are needed to support our findings. Second, susceptibility artifacts caused by intestinal gas is a disadvantage since it can be indistinguishable from signal voids caused by blood products within the small DIE foci. Susceptibility effect increases with field strength (17,33). Takeuchi et al. reported that signal void due to hemosiderin deposition were more prominent at 3T. Susceptibility artifacts caused by intestinal gas in the pelvic region also increase with 3T MRI when compared to 1.5T (17). Two patients were excluded from the study due to prominent susceptibility artifacts caused by colonic overdistention, which limited the visibility of the pelvis. Three lesions were reevaluated by two radiologists to resolve discrepancies caused by these susceptibility artifacts. Chamié et al. reported that motion artifacts and presence of intestinal residues may reduce optimal MRI performance in DIE assessment (34) and bowel preparation could have been helpful when specifically looking for DIE foci. In light of these facts, further comparative studies are needed to determine which field strength is superior for evaluation of DIE. Third, the inter-observer agreement was not determined.

In conclusion, the accurate preoperative assessment of DIE is extremely important in order to establish the precise location and ensure complete surgical removal. SWI detects different phases of endometriosis-associated hemorrhagic changes and provides additional value to improve diagnostic ability of pelvic MRI in DIE patients.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Coutinho

Jr Bittencourt

Pires

. MR imaging in deep pelvic endometriosis: a pictorial essay. Radiographics 2011; 31: 549–567.

Medeiros

Rosa

Silva

. Accuracy of magnetic resonance in deeply infiltrating endometriosis: a systematic review and meta-analysis. Arch Gynecol Obstet 2015; 291: 611–621.

Siegelman

Oliver

. MR imaging of endometriosis: ten imaging pearls. Radiographics 2012; 32: 1675–1691.

Hudelist

Oberwinkler

Singer

. Combination of transvaginal sonography and clinical examination for preoperative diagnosis of pelvic endometriosis. Hum Reprod 2009; 24: 1018–1024.

Abrao

Gonçalves

Dias

Jr . Comparison between clinical examination, transvaginal sonography and magnetic resonance imaging for the diagnosis of deep endometriosis. Hum Reprod 2007; 22: 3092–3097.

Hensen

JH1

Puylaert

. Endometriosis of the posterior cul-de-sac: clinical presentation and findings at transvaginal ultrasound. Am J Roentgenol 2009; 192: 1618–1624.

Benacerraf

Groszmann

. Sonography should be the first imaging examination done to evaluate patients with suspected endometriosis. J Ultrasound Med 2012; 31: 651–653.

Benacerraf

Groszmann

Hornstein

. Deep infiltrating endometriosis of the bowel wall: the comet sign. J Ultrasound Med 2015; 34: 537–542.

Roseau

Dumontier

Palazzo

. Rectosigmoid endometriosis: endoscopic ultrasound features and clinical implications. Endoscopy 2000; 32: 525–530.

10.

Chapron

Dubuisson

. Management of deep endometriosis. Ann N Y Acad Sci 2001; 943: 276–280.

11.

Togashi

Nishimura

Kimura

. Endometrial cysts: diagnosis with MR imaging. Radiology 1991; 180: 73–78.

12.

Takeuchi M, Matsuzaki K, Harada M. Susceptibility-weighted MRI of extra-ovarian endometriosis: preliminary results. Abdom Imaging 2015 Feb 15 [Epub ahead of print].

13.

Coccia

Rizzello

. Ultrasonographic staging: a new staging system for deep endometriosis. Ann N Y Acad Sci 2011; 1221: 61–69.

14.

Siegelman

Outwater

Wang

. Solid pelvic masses caused by endometriosis: MR imaging features. Am J Roentgenol 1994; 163: 357–361.

15.

Busard

Mijatovic

Lüchinger

. MR imaging of bladder endometriosis and its relationship with the anterior uterine wall: experience in a tertiary referral centre. Eur J Radiol 2012; 81: 2106–2111.

16.

Nair

Van Hecke

De Belder

. High-resolution susceptibility-weighted imaging at 3 T with a 32-channel head coil: technique and clinical applications. Am J Roentgenol 2010; 195: 1007–1014.

17.

Takeuchi

Matsuzaki

Nishitani

. Susceptibility-weighted MRI of endometrioma: preliminary results. Am J Roentgenol 2008; 191: 1366–1370.

18.

Haacke

Cheng

. Susceptibility weighted imaging (SWI). Magn Reson Med 2004; 52: 612–618.

19.

Sehgal

Delproposto

Haddar

. Susceptibility-weighted imaging to visualize blood products and improve tumor contrast in the study of brain masses. J Magn Reson Imaging 2006; 24: 41–51.

20.

Dessy

Buccheri

Chiummariello

. Umbilical endometriosis, our experience. In Vivo 2008; 22: 811–815.

21.

Zhou

. Differentiation of bland from neoplastic thrombus of the portal vein in patients with hepatocellular carcinoma: application of susceptibility-weighted MR imaging. BMC Cancer 2014; 14: 590–590.

22.

Zhou

Zhao

. Magnetic resonance susceptibility-weighted imaging versus other imaging modalities in detecting splenic siderotic lesions. PLoS One 2013; 8: e73626–e73626.

23.

Bai

Wang

Han

. Susceptibility weighted imaging: a new tool in the diagnosis of prostate cancer and detection of prostatic calcification. PLoS One 2013; 8: e53237–e53237.

24.

Solak

Sahin

Genç

. Diagnostic value of susceptibility-weighted imaging of abdominal wall endometriomas during the cyclic menstrual changes: a preliminary study. Eur J Radiol 2013; 82: e411–416.

25.

Krishnan

Lansley

Jäger

. New vistas in clinical practice: susceptibility-weighted imaging. Quant Imaging Med Surg 2015; 5: 448–452.

26.

Haacke

Mittal

. Susceptibility-weighted imaging: technical aspects and clinical applications, part 1. Am J Neuroradiol 2009; 30: 19–30.

27.

Ayaz

Boikov

Haacke

. Imaging cerebral microbleeds using susceptibility weighted imaging: one step toward detecting vascular dementia. J Magn Reson Imaging 2010; 31: 142–148.

28.

Mittal

Kish

. Identification of calcification with MRI using susceptibility-weighted imaging: a case study. J Magn Reson Imaging 2009; 29: 177–182.

29.

Bazot

Darai

Hourani

. Deep pelvic endometriosis: MR imaging for diagnosis and prediction of extension of disease. Radiology 2004; 232: 379–389.

30.

Griffiths

Koutsouridou

Penketh

. Rectovaginal endometriosis – a frequently missed diagnosis. J Obstet Gynaecol 2007; 27: 605–607.

31.

Kim

Rha

. MR Imaging findings of hydrosalpinx: a comprehensive review. Radiographics 2009; 29: 495–507.

32.

Gougoutas

Siegelman

Hunt

. Pelvic endometriosis: various manifestations and MR imaging findings. Am J Roentgenol 2000; 175: 353–358.

33.

Nandigam

Viswanathan

Delgado

. MR imaging detection of cerebral microbleeds: effect of susceptibility-weighted imaging, section thickness, and field strength. Am J Neuroradiol 2009; 30: 338–343.

34.

Chamié

Blasbalg

Gonçalves

. Accuracy of magnetic resonance imaging for diagnosis and preoperative assessment of deeply infiltrating endometriosis. Int J Gynaecol Obstet 2009; 106: 198–201.