Transvaginal ultrasonography associated with colour Doppler energy in the diagnosis of hydrosalpinx

S. Guerriero1, S. Ajossa, M.P. Lai, V. Mais, A.M. Paoletti and G.B. Melis

Department of Obstetrics and Gynaecology of the University of Cagliari, Cagliari, Italy


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The aims of this prospective study were to investigate the accuracy of B-mode transvaginal ultrasonography alone, using the typical finding of the presence of an elongated shaped mass with incomplete septa, in the screening of hydrosalpinx in women undergoing surgery for gynaecological diseases, and to determine the predictive value of this method combined with colour Doppler energy (CDE) imaging evaluation and CA125 concentrations in differentiating hydrosalpinx from other adnexal masses. In the first part of the study, 378 consecutive pre-menopausal non-pregnant women were submitted to transvaginal ultrasonography alone before surgery. In the second part of the study, 256 adnexal masses underwent transvaginal ultrasonography combined with CDE imaging evaluation associated with spectral Doppler analysis and plasma concentrations of CA125. Sensitivity and specificity for the ultrasonographic screening were 84.6 and 99.7% respectively, calculated for each adnexum (n = 756) and 93.3 and 99.6% respectively, calculated for each mass, for differentiating hydrosalpinx from other adnexal masses. The CDE imaging and the evaluation of CA125 plasma concentrations do not seem to increase the accuracy of B-mode transvaginal ultrasonography. Inter- and intra-observer agreement, expressed in terms of k-values, was high (0.87 and 0.93 respectively). In conclusion, transvaginal ultrasonography alone is a useful method of detection of hydrosalpinx.

Key words: CA125/colour Doppler energy/hydrosalpinx/pelvic inflammatory disease/transvaginal ultrasound


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Several recent studies have shown that the presence of hydrosalpinges adversely affects clinical pregnancy rate achieved with IVF and embryo transfer (Blazar et al., 1997Go; Murray et al., 1998Go). Hydrosalpinx fluid may be toxic to the endometrium or the embryo (Mukherjee et al., 1996Go), or may mechanically interfere with the implantation (Granot et al., 1998Go). Laparoscopy is the gold standard in the diagnosis of this disease, but less invasive techniques should be proposed to allow appropriate patient counselling of further diagnostic or therapeutic interventions and to reduce the number of unnecessary laparoscopies. Differing results are present in English publications on the use of techniques less invasive than transvaginal ultrasonography. Although several authors have investigated the ultrasonographic appearance of hydrosalpinx (Timor-Tritsch and Rottem, 1987Go; Lande et al., 1988Go; Tessler et al., 1989Go; Patten et al., 1990Go; Cacciatore et al., 1992Go; Kupesic et al., 1995Go) only a few prospective studies have been performed and all included patients with acute symptoms (Patten et al., 1990Go; Cacciatore et al., 1992Go) or ectopic pregnancies (Lande et al., 1988Go). On the contrary, Atri et al. (Atri et al., 1994Go), in a study on chronic pelvic inflammatory disease reported a sensitivity of only 34% in patients who subsequently underwent hysterosalpingography but not laparoscopy. This sensitivity was also reported in another study (Boardman et al., 1997Go) but in acute pelvic inflammatory disease, while Tukeva et al. (Tukeva et al., 1999Go) reported a higher value (81%) in detection of the same disease with a lower specificity. Recently, Timor-Tritsch et al. (Timor-Tritsch et al., 1998Go) reviewed the spectrum of transvaginal sonographic findings of tubal inflammatory disease but did not analyse the specificity and the sensitivity of this technique. To our knowledge, transvaginal ultrasonography alone or combined with colour Doppler energy has never been prospectively evaluated in the diagnosis of this kind of tubal disease in a population of women without evidence of acute genital inflammation. The aims of this prospective study were to investigate the accuracy of transvaginal ultrasonography alone in the screening of hydrosalpinx in women undergoing surgery for gynaecological diseases and to determine the predictive value of this method combined with colour Doppler energy (CDE) imaging and CA125 determinations in differentiating hydrosalpinx from other adnexal masses.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Study design
This study has been reviewed and approved by the ethical committee of the Department of Obstetrics and Gynaecology of the University of Cagliari. The study was divided into two sections. In the first one, 378 consecutive pre-menopausal non-pregnant women underwent surgery at the Department of Obstetrics and Gynaecology of the University of Cagliari between February 1995 and February 1999; they were enrolled to investigate the role of transvaginal ultrasonography in the screening for hydrosalpinx. The mean (±SD) age of the study population was 32.3 ± 6.9 years (range 14–54). The patients underwent surgery for infertility, pelvic pain, uterine fibroids, endometrial hyperplasia or adnexal masses. Women with evidence of acute genital inflammation, previous bilateral salpingo-oophorectomy, previously treated ovarian carcinoma, and endometrial or cervical carcinomas were excluded. Within 2 days before surgery, all patients underwent transvaginal ultrasonography using an Acuson XP/10 OB (Acuson Inc., Mountain View, CA, USA) with a 7 MHz transvaginal probe. Using B-mode ultrasonography a hydrosalpinx was diagnosed when an elongated shaped mass with incomplete septa (Timor-Tritsch et al., 1998Go) was visualized (Figure 1Go). We defined, in agreement with Timor-Tritsch et al. (Timor-Tritsch et al., 1998Go), the incomplete septa as hyperechoic septa that originate as a triangular protrusion from one of the walls, but do not reach the opposite wall (Figure 1Go). As additional evidence, we evaluated the presence or absence of the `beads-on-a-string' sign (Figure 2Go). We defined (Timor-Tritsch et al., 1998Go) the `beads-on-a-string' sign as hyperechoic mural nodules measuring about 2–3 mm and seen on cross-section of the fluid-filled distended structure (Figure 2Go).



View larger version (129K):
[in this window]
[in a new window]
 
Figure 1. The elongated shaped mass with incomplete septa (arrows) considered as B-mode ultrasonographic findings of hydrosalpinx.

 


View larger version (99K):
[in this window]
[in a new window]
 
Figure 2. The `beads-on-a-string' sign (arrows) considered as additional evidence of the presence of hydrosalpinx.

 
Colour Doppler assessment and CA125 determination
In the second part of the study, 239 consecutive pre-menopausal non-pregnant women underwent surgery for 256 adnexal masses between December 1996 and February 1999 and were enrolled to determine the predictive value of transvaginal ultrasonography alone or combined with other methods, such as colour Doppler and CA125 determinations, in differentiating hydrosalpinx from other adnexal masses. The mean (±SD) age of the study population was 33.5 ± 9.2 years (range 14–54). In this section of the study, all transvaginal scans performed within 2 days before surgery were completed by transvaginal CDE imaging using the previously described ultrasonographic equipment and blood samples were collected the same day from all the patients to measure serum concentrations of CA125.

The ultrasonographic unit was equipped with a colour pulsed Doppler ultrasonography system upgraded with CDE imaging. To avoid the risks of bias, conventional colour Doppler imaging evaluation was performed neither before nor after CDE imaging evaluation. The machine settings were fixed at the following parameters: log compression (dynamic range of energy signal) of 35–40 dB, mix 4 (= most transparent), CDE post-processing 6, power <500, preprocessing 1, persistence 3 (= medium amount of smoothing), filter 3, gate setting 2. The Doppler study was performed by looking for colour signals along the wall and within the septa. When colour signals were detected, the pulsed Doppler gate was superimposed, and the pulsatility index (PI) and resistance index (RI) were electronically computed. When multiple signals were obtained from the same mass, the lowest PI and RI values were used for the statistical analysis. The intra-observer coefficient of variation was determined by analysing three sets of five consecutive waveforms from the vessel with the lowest PI and RI in the first 10 masses studied. The intra-observer variabilities for RI and PI were 3 and 4% respectively. Through B-mode ultrasonography, a hydrosalpinx was diagnosed using the previously described criteria (Timor-Tritsch et al., 1998Go) (Figure 1Go). Using CDE imaging (CDE imaging evaluation of vessel distribution), the presence of an elongated shaped mass with incomplete septa with `poor' and peripheral vascularization was considered to indicate the likelihood of hydrosalpinx (Figure 3Go). In addition a PI >1.0 or an RI >0.4 were considered characteristic of benign adnexal mass (Guerriero et al., 1998aGo) and used to reduce the risk of misdiagnosis with ovarian cancer.



View larger version (46K):
[in this window]
[in a new window]
 
Figure 3. The typical colour Doppler energy findings of hydrosalpinx.

 
We also evaluated the distribution of vessels (central or peripheral) and the intensity of CDE signal (`poor vascularization' or `rich vascularization'). An adnexal mass was defined as `poorly vascularized' when there was an absence of signal or only two to four small colour signals were visualized along the wall.

The CA125 assays were performed with an immunoradiometric assay method by using two monoclonal antibodies (CIS Bio International, Gif sur Yvette, France). The intra- and interassay coefficients of variation were 3.9 and 4.2% respectively; the sensitivity was <0.5 IU/ml. A CA125 cut-off value of 25 IU/ml was considered to be characteristic of hydrosalpinx because in a previous retrospective study on 10 hydrosalpinges (unpublished data), we found that a value of CA125 >25 IU/ml was present in 100% of this type of adnexal mass.

The sonograms of each adnexal mass were obtained prospectively to evaluate the presence of hydrosalpinx. At surgery, all adnexa were carefully observed by two of the authors (V.M. and G.B.M.) and all adnexal masses were removed. The ultrasonographic impressions and the CA125 values were then compared with the final visual and histopathological diagnosis of hydrosalpinx which was defined as a liquid filled tube (Bloeche, 1999).

Statistics
In the first part of the study, to investigate the role of transvaginal ultrasonography in screening for hydrosalpinx, the sensitivity, specificity, and positive and negative predictive values of transvaginal ultrasonography were calculated for each visualized adnexum (Mais et al., 1993Go; Guerriero et al., 1998bGo). To evaluate the overall agreement between a test result and the actual outcome, the kappa index was calculated according to a previously described method (Fleiss, 1981Go); kappa-values ranging between 0.40 and 0.75 were assumed to indicate a strong agreement.

In the second part of the study, the sensitivity, specificity, positive and negative predictive values, and kappa-value of transvaginal ultrasonography and all combined methods were calculated for each adnexal mass (Mais et al., 1993Go; Guerriero et al., 1998bGo). The z statistic for the comparison of two proportions (Glantz, 1981Go) was used to evaluate the results. To evaluate the reproducibility of B-mode findings and assess the inter- and intra-observer variability in the interpretation of images (expressed in terms of k-values) (Mol et al., 1996Go), the sonograms of each adnexal mass were independently reviewed by two gynaecologists (S.G. and S.A.) with different experience in transvaginal ultrasonography (11 and 5 years of experience respectively). In addition, for comparison of different percentages of distribution of vessels and the intensity of CDE signal in hydrosalpinges, other benign adnexal masses and malignant masses, the {chi}2 statistic was used.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Both adnexae were visualized in all patients. In the first part of the study, 26 out of 378 subjects were found to have one hydrosalpinx confirmed by pathology (prevalence, 7%). Of the 24 hydrosalpinges suspected by transvaginal ultrasonography 22 were confirmed by laparoscopy. Of 732 ultrasonographic diagnoses of absence of hydrosalpinx, 728 were confirmed by surgical evaluation. In the 22 hydrosalpinges confirmed by laparoscopy the mean diameter (±SD) was 45.5 ± 14.2 mm, while the mean (±SD) of maximum diameters was 58.5 ± 15.3 mm. We found the presence of the `beads-on-a-string' at ultrasonography in only three hydrosalpinges confirmed by laparoscopy (Timor-Tritsch et al., 1998Go).

The ultrasonographic findings of the two false-positive cases were similar to findings considered characteristic for hydrosalpinx but pathological work-up revealed two serous cystadenomas. In the four false negative cases of B-mode ultrasonography, the presence of hydrosalpinx was missed but an alteration of normal utero–ovarian relationship was always suspected due to the presence of pelvic adhesions. Therefore the sensitivity of transvaginal ultrasonography in the diagnosis of hydrosalpinx was 84.6%, with a specificity of 99.7%, and positive and negative predictive values of 91.7 and 99.4% respectively. The kappa index of 0.87 suggested a strong agreement between transvaginal ultrasonography and surgery.

In the second part of the study, the sensitivity of transvaginal ultrasonography in the differential diagnosis of hydrosalpinx was 93.3%, with a specificity of 99.6% and positive and negative predictive values of 93.3 and 99.6% respectively. The kappa index of 0.93 suggests a strong agreement between transvaginal ultrasonography and surgery. No increase in diagnostic accuracy was achieved by using colour Doppler or CA125 plasma concentration determinations (Table IGo). Kappa-values for reproducibility within observers and between observers were almost perfect (0.93 and 0.87 respectively).


View this table:
[in this window]
[in a new window]
 
Table I. Accuracy of association of B-mode ultrasonography and colour Doppler energy (CDE) imaging, CA125 >25 IU/ml and lowest pulsatility index (PI) and resistance index (RI) in the differential diagnosis of hydrosalpinx
 
We evaluated the PI, RI and CA125 in patients with adnexal masses. The median (±SD) PI value was 1.62 ± 0.75 in hydrosalpinges (range 0.84–3.19), 0.89 ± 0.61 in other benign masses (range: 0.31–4.21) and 0.675 ± 0.30 in malignant masses (range 0.14–1.33). The median (±SD) RI value was 0.71 ± 0.11 in hydrosalpinges (range 0.57–0.89), 0.57 ± 0.14 in other benign adnexal masses (range 0.25–1.05) and 0.44 ± 0.16 in malignant masses (range: 0.13–0.73). The median CA125 value was 42.2 ± 33.0 IU/ml in hydrosalpinx (range 8.6–140.2 IU/ml), 25.6 ± 99.6 IU/ml in other benign adnexal masses (range: 3.41–1000), and 100.25 ± 187.5 IU/ml in malignant masses (range 10.9–710). Intratumoral arterial blood was visualized in 93.3% (14/15) of hydrosalpinges, in 91.4% (202/221) of other benign adnexal masses and in 100% (20/20) of malignant masses (not significant). The analysis of vessel distribution showed the presence of colour in the periphery of the mass in 100% of hydrosalpinges, in 94.1% (208/221) of other benign adnexal masses (not significant) and in 0% of malignant masses (P < 0.001).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study suggests the important role of transvaginal ultrasonography in the evaluation of hydrosalpinx in non-pregnant women without evidence of acute genital inflammation with the combined use of colour Doppler and CA125 plasma concentrations. In fact, no increase in accuracy was observed. It must be considered that the good predictive capacity of transvaginal ultrasonography demonstrated in a population of patients with chronic pelvic pain and/or infertility and/or pelvic masses undergoing surgery may not be true in the general population. Our study population has a higher risk of the presence of hydrosalpinx in comparison to asymptomatic women. This may lead to higher sensitivity and positive predictive values in the population studied than if we had looked at another population such as an asymptomatic one. Otherwise, as stated previously in a study of pelvic adhesions (Guerriero et al., 1997Go), these `at risk' patients are the group for which the transvaginal ultrasonographic evaluation could be very useful in allowing appropriate counselling for further diagnostic or therapeutic interventions.

The different results observed between a previous study (Atri et al., 1994Go) and the present study probably arise because the first study uses hysterosalpingography as the `gold standard' with the aim of detecting the distally occluded tube, which is a different entity from hydrosalpinx with different reproductive outcome (Bloeche, 1999).

The role of ultrasonography seems crucial also because only hydrosalpinges visible on ultrasound are associated with reduced implantation and pregnancy rates after in-vitro fertilization (de Wit et al., 1998Go; Strandell et al., 1999Go). Theoretically, as a tube filled with fluid should be recognized as a cystic mass, all hydrosalpinges should be visible by ultrasound. Our study firstly demonstrated that this is not always true because some hydrosalpinges can be missed by ultrasound. As previously demonstrated in the diagnosis of dermoid cyst (Mais et al., 1995Go), hydrosalpinx could also be missed because of the presence of faecal material and gas within the bowel in a pelvis distorted by other abnormalities. For these reasons the overall accuracy of ultrasonography is slightly reduced but, due to its higher sensitivity, this technique shows greater accuracy in comparison to magnetic resonance imaging (Outwater et al., 1998Go) with lower costs. In addition, our study demonstrates the high reproducibility of the ultrasonographic findings. As stated previously (Timor-Tritsch et al., 1998Go), it is critical to differentiate chronic tubal inflammatory disease from an ovarian malignant tumour. In our experience, this is easy and false negatives of ovarian cancer are absent because the presence of incomplete septa almost uniformly indicates the diagnosis of a Fallopian tube, since the true septa of ovarian tumours are very seldom, if ever, incomplete.

Several articles in the last 2 years have been specifically addressed to investigate and review the modern management of hydrosalpinx (Aboulghar et al., 1998Go; Lass, 1999Go). For these reasons the diagnosis of hydrosalpinx with less invasive methods has aroused new interest associated with unexpected and controversial results. The use of B-mode transvaginal ultrasonography should be a preliminary test in the evaluation of an infertile couple because of the large amount of available information about the presence of pelvic adhesions (Guerriero et al., 1997Go). The results of this study show that hydrosalpinges can also be suspected and identified with sufficient accuracy.

In the present study, we demonstrate that abnormal results at transvaginal ultrasound examination accurately identify patients with hydrosalpinx with a positive predictive value of 93.3%. Therapeutic considerations in this clinical situation include operative laparoscopy to remove the mass or IVF. In our opinion, operative laparoscopy is the first therapeutic choice due to the effect of hydrosalpinx on the results of IVF. Normal ultrasonography is also reliable because surgery generally confirmed the absence of hydrosalpinx. We found the false negative rate to be 2.6% and the negative predictive value to be 99.6%. These patients, if they desire pregnancy and the tubal patency is present, can postpone diagnostic laparoscopy for three to six cycles of in-vivo inseminations (Melis et al., 1995Go).


    Notes
 
1 To whom correspondence should be addressed at: Department of Obstetrics and Gynaecology of the University of Cagliari, Ospedale San Giovanni di Dio, Via Ospedale 46, 09124, Cagliari, Italy. Email: gimeli{at}tin.it Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Aboulghar, M.A.A., Mansour, R.T. and Serour, G.I. (1998) Controversies in the modern management of hydrosalpinx. Hum. Reprod. Update, 4, 882–890.[Abstract/Free Full Text]

Atri, M., Tran, C.N., Bret, P.M. et al. (1994) Accuracy of endovaginal sonography for the detection of Fallopian tube blockage. J. Ultrasound Med., 13, 429–434.[Abstract]

Blazar, A.S., Hogan, J.W., Seifer, D.B. et al. (1997) The impact of hydrosalpinx on successful pregnancy in tubal factor infertility treated by in vitro fertilization. Fertil. Steril., 67, 517–520.[ISI][Medline]

Bloechle, M. (1999) What is a hydrosalpinx? A plea for the use of a proper terminology in scientific discussion. Hum. Reprod., 14, 578.[Free Full Text]

Boardman, L.A., Peipert, J.F., Brody, J.M. et al. (1997) Endovaginal sonography for the diagnosis of upper genital tract infection. Obstet. Gynecol., 90, 54–57.[Abstract/Free Full Text]

Cacciatore, B., Leminen, A., Ingman-Friberg, S. et al. (1992) Transvaginal sonographic findings in ambulatory patients with suspected pelvic inflammatory disease. Obstet. Gynecol., 80, 912–916.[Abstract]

De Wit, W., Gowrising, C.J., Kuik, D.J. et al. (1998) Only hydrosalpinges visible on ultrasound are associated with reduced implantation and pregnancy rates after in-vitro fertilization. Hum. Reprod., 13, 1696–1701.[Abstract]

Fleiss, J.L. (1981) Statistical Methods for Rates and Proportions. Wiley, New York.

Glantz, S.A. (1981) Primer of Biostatistics. McGraw-Hill, New York.

Granot, I., Dekel, N., Segal, I. et al. (1998) Is hydrosalpinx fluid cytotoxic? Hum. Reprod., 13, 1620–1624.[Abstract]

Guerriero, S., Ajossa, S., Lai, M.P. et al. (1997) Transvaginal ultrasonography in the diagnosis of pelvic adhesions. Hum. Reprod., 12, 2649–2653.[Abstract]

Guerriero S., Ajossa S., Risalvato A. et al. (1998a) Diagnosis of adnexal malignancies by using color Doppler energy imaging as a secondary test in persistent masses. Ultrasound Obstet. Gynecol., 11, 277–282.[ISI][Medline]

Guerriero S., Ajossa S., Mais V. et al. (1998b) The diagnosis of endometriomas using colour Doppler energy imaging. Hum. Reprod., 13, 1691–1695.[Abstract]

Kupesic, S., Kurjak, A., Pasalic, L. et al. (1995) The value of transvaginal color Doppler in the assessment of pelvic inflammatory disease. Ultrasound Med. Biol., 21, 733–738.[ISI][Medline]

Lande, I.M., Hill, M., Cosco, F.E. et al. (1988) Adnexal and cul-de-sac abnormalities: transvaginal sonography. Radiology, 166, 325–332.[Abstract]

Lass, A. (1999) What effect does hydrosalpinx have on assisted reproduction? What is the preferred treatment for hydrosalpinges? The ovary's perspective. Hum. Reprod., 14, 1674–1677.[Free Full Text]

Mais, V., Guerriero, S., Ajossa, S. et al. (1993) The efficiency of transvaginal ultrasonography in the diagnosis of endometrioma. Fertil. Steril., 60, 776–780.[ISI][Medline]

Mais, V., Guerriero, S., Ajossa, S. et al. (1995) Transvaginal ultrasonography in the diagnosis of cystic teratoma. Obstet. Gynecol., 85, 48–52.[Abstract/Free Full Text]

Melis, G.B., Paoletti, A.M., Ajossa, S. et al. (1995) Ovulation induction with gonadotropins as sole treatment in infertile couples with open tubes: a randomized prospective comparison between intrauterine insemination and timed vaginal intercourse. Fertil. Steril., 64, 1088–1093.[ISI][Medline]

Mol, B.W.J., Swart, P., Bossuyt, P.M.M. et al. (1996) Reproducibility of the interpretation of hysterosalpingography in the diagnosis of tubal pathology. Hum. Reprod., 11, 1204–1208.[Abstract]

Mukherjee, T., Copperman, A.B., McCaffrey, C. et al. (1996) Hydrosalpinx fluid has embryotoxic effects on murine embryogenesis: a case for prophylactic salpingectomy. Fertil. Steril., 66, 851–853.[ISI][Medline]

Murray, D.L., Sagoskin, A.W., Widra, E.A. et al. (1998) The adverse effect of hydrosalpinges on in vitro fertilization pregnancy rates and the benefit of surgical correction. Fertil. Steril., 69, 41–45.[ISI][Medline]

Outwater, E.K., Siegelman, E.S., Chiowanich, P., Kilger, A.M. et al. (1998) Dilated Fallopian tubes: MR imaging characteristics. Radiology, 208, 463–469[Abstract]

Patten, R.M., Vincent, L.M., Wolner-Hanssen, P. et al. (1990) Pelvic inflammatory disease. Endovaginal sonography with laparoscopic correlation. J. Ultrasound Med., 9, 681–689.[Abstract]

Strandell, A., Lindhard, A., Waldenstrom, U. et al. (1999) Hydrosalpinx and IVF outcome: a prospective, randomized multicentre trial in Scandinavia on salpingectomy prior to IVF. Hum. Reprod., 14, 2762–2769.[Abstract/Free Full Text]

Tessler, F.L., Perrella, R.R., Fleischer, A.C. and Grant, E.G. (1989) Endovaginal sonographic diagnosis of dilated Fallopian tubes. Am. J. Roentgenol., 153, 523–525.[ISI][Medline]

Timor-Tritsch, I.E. and Rottem, S. (1987) Transvaginal ultrasonographic study of the Fallopian tube. Obstet. Gynecol., 70, 424–428.[Abstract]

Timor-Tritsch, I.E., Lerner, J.P., Monteagudo, A. et al. (1998) Transvaginal sonographic markers of tubal inflammatory disease. Ultrasound Obstet. Gynecol., 12, 56–66.[ISI][Medline]

Tukeva, T.A., Aronen, H.J., Karjalainen, P.T. et al. (1999) MR imaging in pelvic inflammatory disease: comparison with laparoscopy. Radiology, 210, 209–216.[Abstract/Free Full Text]

Submitted on December 2, 1999; accepted on March 29, 2000.