Three-Dimensional Measurement of the Uterus on Magnetic Resonance Images: Development and Performance Analysis of an Automated Deep-Learning Tool.

Uterus measurements are useful for assessing both the treatment and follow-ups of gynaecological patients. The aim of our study was to develop a deep learning (DL) tool for fully automated measurement of the three-dimensional size of the uterus on magnetic resonance imaging (MRI). In this single-centre retrospective study, 900 cases were included to train, validate, and test a VGG-16/VGG-11 convolutional neural network (CNN). The ground truth was manual measurement. The performance of the model was evaluated using the objective key point similarity (OKS), the mean difference in millimetres, and coefficient of determination R2. The OKS of our model was 0.92 (validation) and 0.96 (test). The average deviation and R2 coefficient between the AI measurements and the manual ones were, respectively, 3.9 mm and 0.93 for two-point length, 3.7 mm and 0.94 for three-point length, 2.6 mm and 0.93 for width, 4.2 mm and 0.75 for thickness. The inter-radiologist variability was 1.4 mm. A three-dimensional automated measurement was obtained in 1.6 s. In conclusion, our model was able to locate the uterus on MRIs and place measurement points on it to obtain its three-dimensional measurement with a very good correlation compared to manual measurements.

Pregnancy in a rudimentary horn: multicenter's MRI features of a rare condition.

PURPOSE: To describe the MRI features of rudimentary horn pregnancy (RHP) with surgical correlations. METHODS: Nine women with a RHP underwent preoperative pelvic MRI. MRI protocol included T2- (n = 9), T1- (n = 7), and fat-suppressed contrast-enhanced T1-weighted sequences (n = 4). Two pelvic radiologists retrospectively analyzed MR images to assess the following MRI features: presence of a myometrium around the gestational sac (GS) and characteristics of its wall, GS surrounded by myometrium in contact with the round ligament, communication of the GS with the endometrial cavity of the main horn, continuity of the GS with the cervix, fibrous or muscular GS attachment to the main horn, lateral deviation, and endometrial thickness of the main horn. Ovaries and tubes were also assessed. MRI features were correlated with surgical findings. RESULTS: Seven of the nine women [29 ± 6 SD years (range 16-37 years)] underwent surgical management. The first US diagnosed RHP in only 1/9 patients. All pregnancies were diagnosed using MRI. RHP was all located in the rudimentary horn of a unicornuate uterus. All the GS was surrounded by myometrium in contact with the round ligament. None of the RHP displayed communication with the endometrial cavity of the main horn nor with the cervix. An attachment between the RHP and the main horn was seen in 3/9 patients. All the main horns were lateralized and empty. CONCLUSION: MRI diagnosed RHP in all patients by identifying the GS surrounded by myometrium in contact with the round ligament and the absence of continuity between the GS and the cervix. LEVEL OF EVIDENCE: IV-retrospective study.

Impact of artificial intelligence in breast cancer screening with mammography.

OBJECTIVES: To demonstrate that radiologists, with the help of artificial intelligence (AI), are able to better classify screening mammograms into the correct breast imaging reporting and data system (BI-RADS) category, and as a secondary objective, to explore the impact of AI on cancer detection and mammogram interpretation time. METHODS: A multi-reader, multi-case study with cross-over design, was performed, including 314 mammograms. Twelve radiologists interpreted the examinations in two sessions delayed by a 4 weeks wash-out period with and without AI support. For each breast of each mammogram, they had to mark the most suspicious lesion (if any) and assign it with a forced BI-RADS category and a level of suspicion or "continuous BI-RADS 100". Cohen's kappa correlation coefficient evaluating the inter-observer agreement for BI-RADS category per breast, and the area under the receiver operating characteristic curve (AUC), were used as metrics and analyzed. RESULTS: On average, the quadratic kappa coefficient increased significantly when using AI for all readers [κ = 0.549, 95% CI (0.528-0.571) without AI and κ = 0.626, 95% CI (0.607-0.6455) with AI]. AUC was significantly improved when using AI (0.74 vs 0.77, p = 0.004). Reading time was not significantly affected for all readers (106 s without AI and vs 102 s with AI; p = 0.754). CONCLUSIONS: When using AI, radiologists were able to better assign mammograms with the correct BI-RADS category without slowing down the interpretation time.

O-RADS MRI Classification of Indeterminate Adnexal Lesions: Time-Intensity Curve Analysis Is Better Than Visual Assessment.

Background The MRI Ovarian-Adnexal Reporting and Data System (O-RADS) enables risk stratification of sonographically indeterminate adnexal lesions, partly based on time-intensity curve (TIC) analysis, which may not be universally available. Purpose To compare the diagnostic accuracy of visual assessment with that of TIC assessment of dynamic contrast-enhanced MRI scans to categorize adnexal lesions as benign or malignant and to evaluate the influence on the O-RADS MRI score. Materials and Methods The European Adnex MR Study Group, or EURAD, database, a prospective multicenter study of women undergoing MRI for indeterminate adnexal lesions between March 2013 and March 2018, was queried retrospectively. Women undergoing surgery for an adnexal lesion with solid tissue were included. Solid tissue enhancement relative to outer myometrium was assessed visually and with TIC. Contrast material washout was recorded. Lesions were categorized according to the O-RADS MRI score with visual and TIC assessment. Per-lesion diagnostic accuracy was calculated. Results A total of 320 lesions (207 malignant, 113 benign) in 244 women (mean age, 55.3 years ± 15.8 [standard deviation]) were analyzed. Sensitivity for malignancy was 96% (198 of 207) and 76% (157 of 207) for TIC and visual assessment, respectively. TIC was more accurate than visual assessment (86% [95% CI: 81, 90] vs 78% [95% CI: 73, 82]; P < .001) for benign lesions, predominantly because of higher specificity (95% [95% CI: 92, 98] vs 76% [95% CI: 68, 81]). A total of 21% (38 of 177) of invasive lesions were rated as low risk visually. Contrast material washout and high-risk enhancement (defined as earlier enhancement than in the myometrium) were highly specific for malignancy for both TIC (97% [95% CI: 91, 99] and 94% [95% CI: 90, 97], respectively) and visual assessment (97% [95% CI: 92, 99] and 93% [95% CI: 88, 97], respectively). O-RADS MRI score was more accurate with TIC than with visual assessment (area under the receiver operating characteristic curve, 0.87 [95% CI: 0.83, 0.90] vs 0.73 [95% CI: 0.68, 0.78]; P < .001). Conclusion Time-intensity curve analysis was more accurate than visual assessment for achieving optimal diagnostic accuracy with the Ovarian-Adnexal Reporting and Data System MRI score. Clinical trial registration no. NCT01738789 © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Vargas and Woo in this issue. An earlier incorrect version appeared online. This article was corrected on March 7, 2022.

Ovarian-Adnexal Reporting Data System Magnetic Resonance Imaging (O-RADS MRI) Score for Risk Stratification of Sonographically Indeterminate Adnexal Masses.

Importance: Approximately one-quarter of adnexal masses detected at ultrasonography are indeterminate for benignity or malignancy, posing a substantial clinical dilemma. Objective: To validate the accuracy of a 5-point Ovarian-Adnexal Reporting Data System Magnetic Resonance Imaging (O-RADS MRI) score for risk stratification of adnexal masses. Design, Setting, and Participants: This multicenter cohort study was conducted between March 1, 2013, and March 31, 2016. Among patients undergoing expectant management, 2-year follow-up data were completed by March 31, 2018. A routine pelvic MRI was performed among consecutive patients referred to characterize a sonographically indeterminate adnexal mass according to routine diagnostic practice at 15 referral centers. The MRI score was prospectively applied by 2 onsite readers and by 1 reader masked to clinical and ultrasonographic data. Data analysis was conducted between April and November 2018. Main Outcomes and Measures: The primary end point was the joint analysis of true-negative and false-negative rates according to the MRI score compared with the reference standard (ie, histology or 2-year follow-up). Results: A total of 1340 women (mean [range] age, 49 [18-96] years) were enrolled. Of 1194 evaluable women, 1130 (94.6%) had a pelvic mass on MRI with a reference standard (surgery, 768 [67.9%]; 2-year follow-up, 362 [32.1%]). A total of 203 patients (18.0%) had at least 1 malignant adnexal or nonadnexal pelvic mass. No invasive cancer was assigned a score of 2. Positive likelihood ratios were 0.01 for score 2, 0.27 for score 3, 4.42 for score 4, and 38.81 for score 5. Area under the receiver operating characteristic curve was 0.961 (95% CI, 0.948-0.971) among experienced readers, with a sensitivity of 0.93 (95% CI, 0.89-0.96; 189 of 203 patients) and a specificity of 0.91 (95% CI, 0.89-0.93; 848 of 927 patients). There was good interrater agreement among both experienced and junior readers (κ = 0.784; 95% CI, 0.743-0824). Of 580 of 1130 women (51.3%) with a mass on MRI and no specific gynecological symptoms, 362 (62.4%) underwent surgery. Of them, 244 (67.4%) had benign lesions and a score of 3 or less. The MRI score correctly reclassified the mass origin as nonadnexal with a sensitivity of 0.99 (95% CI, 0.98-0.99; 1360 of 1372 patients) and a specificity of 0.78 (95% CI, 0.71-0.85; 102 of 130 patients). Conclusions and Relevance: In this study, the O-RADS MRI score was accurate when stratifying the risk of malignancy in adnexal masses.

Aggressive angiomyxoma of the perineal region.

Aggressive angiomyxoma is a rare tumour that frequently involves the perineal region with a high risk of local recurrence. This is a case report of a 24-year-old female patient with a genital prolapse. We performed a surgical treatment. Histological examination found an aggressive angiomyxoma. The tumour recurred 1 year after surgery. Long-term follow-up is necessary.

Pelvic MRI: Is Endovaginal or Rectal Filling Needed?

Magnetic resonance imaging is the optimal modality for pelvic imaging. It is based on T2-weighted magnetic resonance (MR) sequences allowing uterine and vaginal cavity assessment as well as rectal evaluation. Anatomical depiction of these structures may benefit from distension, and conditions either developing inside the lumen of cavities or coming from the outside may then be better delineated and localized. The need for distension, either rectal or vaginal, and the way to conduct it are matters of debate, depending on indication for which the MR examination is being conducted. In this review, we discuss advantages and potential drawbacks of this technique, based on literature and our experience, in the evaluation of various gynecological and rectal diseases.

Underestimation Rate at MR Imaging-guided Vacuum-assisted Breast Biopsy: A Multi-Institutional Retrospective Study of 1509 Breast Biopsies.

Purpose To assess the rate of underestimation of atypical ductal hyperplasia (ADH) and ductal carcinoma in situ (DCIS) at magnetic resonance (MR) imaging-guided vacuum-assisted breast biopsy and to explore the imaging, demographic, and histologic characteristics associated with lesion upgrade after surgery. Materials and Methods This retrospective study had institutional review board approval, and the need to obtain informed patient consent was waived. A total of 1509 MR imaging-guided vacuum-assisted biopsy procedures were performed in nine centers. A diagnosis of ADH was obtained after biopsy in 72 cases, and a diagnosis of DCIS was obtained in 118 cases. Pearson χ2 and Fisher tests were used to assess the association between demographic, MR imaging, and biopsy features and lesion upgrade. Univariate statistical analyses were performed, and each significant parameter was entered into a multivariate logistic regression analysis. Results Surgical excision was performed in 66 of the 72 ADH cases and in 117 of 118 DCIS cases. The ADH and DCIS underestimation rates were 25.8% (17 of 66) and 23.1% (27 of 117), respectively. Underestimation was 5.6-fold (odds ratio [OR] = 5.6; 95% confidence interval [CI]: 1.7, 18.3) and 3.6-fold (OR = 3.6; 95% CI: 1.2, 10) more likely in mass (n = 20 for ADH and n = 20 for DCIS) than in non-mass (n = 46 for ADH and n = 97 for DCIS), compared with nonunderestimation, in ADH and DCIS respectively. At multivariate analysis, the use of a 9- or 10-gauge needle versus a 7- or 8-gauge needle was also an independently associated with underestimation when a diagnosis of ADH was made at MR imaging-guided biopsy. No other parameters were associated with of ADH or DCIS upgrade at surgery. Conclusion The rates of underestimation in ADH and DCIS diagnosed at MR imaging-guided vacuum-assisted biopsy were high, at around 25%, and were significantly associated with the presence of a mass at MR imaging. © RSNA, 2016.

Ultrasound-guided high-intensity focused ultrasound treatment of breast fibroadenoma-a multicenter experience.

BACKGROUND: The aim of our multicenter study was to assess the clinical outcome and safety of ultrasound (US)-guided high-intensity focused ultrasound (HIFU) in patients with breast fibroadenoma (FA). METHODS: From May 2011 to February 2013, 42 women with 51 FA in one or both breasts were selected for treatment with US-guided HIFU. Eight of 51 FA were treated twice. Patients' age ranged from 16 to 52 years (mean 32 years). All patients with FA underwent core needle biopsy with histological confirmation. HIFU treatment was performed as an outpatient procedure under conscious sedation. Exclusion criteria were pregnant or lactating women, microcalcifications within the lesion at mammogram, history of breast cancer, previous laser or radiation therapy, and breast implant in the same breast. All patients signed written informed consent. After the treatment, follow-up US with volume evaluation was performed at 2, 6, and 12 months. RESULTS: The FA mean baseline volume was 3.89 ml (0.34-19.66 ml). At 2-month follow-up, the mean volume reduction was 33.2% ± 19.1% and achieved significance at 6-month (59.2% ± 18.2%, p < 0.001) and 12-month (72.5% ± 16.7%, p < 0.001) follow-up. Related side effects as superficial skin burn with blister-like aspect in three patients and hyperpigmentation over the treated area in one patient were transient and resolved spontaneously. In one patient, asymptomatic subcutaneous induration persisted at the end of the study. CONCLUSIONS: US-guided HIFU treatment is an effective noninvasive method for the treatment of breast FA and well tolerated by the patients. Preliminary results are encouraging and show that HIFU could be an alternative to surgery for breast FA.

Reconciling the definitions of polycystic ovary syndrome: the ovarian follicle number and serum anti-Müllerian hormone concentrations aggregate with the markers of hyperandrogenism.

CONTEXT: It is still debated whether clinical and/or biological indices of hyperandrogenism (HA) should be present to qualify a patient as having polycystic ovary syndrome (PCOS). We hypothesized that excessive follicle number (FN) assessed by ovarian ultrasonography and/or serum anti-Müllerian hormone (AMH) concentrations may be used as surrogates for the classical markers of HA. DESIGN AND METHODS: Data were obtained from a database of clinical, hormonal, and ultrasound features that were consecutively recorded in 270 women with PCOS (defined using the Rotterdam Criteria) and 217 infertile nonhyperandrogenic normoovulatory women. These variables were submitted to principal component analysis, a multivariable statistical procedure that transforms a number of possibly correlated variables into a smaller number of uncorrelated variables called principal components (PC). Variables that aggregate in the same PC capture the same information. RESULTS: In the control group, as expected, three independent PCs were identified: 1) the markers of the metabolic (i.e. insulin resistance) status; 2) those of the androgen status; and 3) those of the follicle status. In the PCOS group, the metabolic variables also aggregated in a first PC. Ovarian androgen and follicle markers aggregated in a second independent PC, with FN and serum AMH having the strongest correlation coefficients. A third PC summarized the adrenal contribution to the HA of PCOS. In both groups, the free androgen index correlated equally to the first and second PCs. CONCLUSIONS: The similarity of the first PC between controls and PCOS supports the hypothesis that the metabolic anomaly of PCOS is neither intrinsic nor specific. Conversely, by gathering the androgen and follicle variables, the second PC in PCOS may be viewed as summarizing a specific ovarian anomaly. Because both FN and/or serum AMH were strongly correlated to the second PC along with androgens, they may be used equally as surrogates for the classical markers of ovarian HA. This reconciles the Rotterdam Consensus and other definitions for PCOS, especially in women having the Rotterdam PCOS phenotype without HA. We thus propose a simple strategy for the diagnosis of PCOS in clinical practice.

Breast cancer during pregnancy.

Breast cancer in pregnancy is an uncommon situation but poses dilemmas for patients and their physicians. There is a paucity of prospective studies regarding diagnosis and treatment of breast cancer during pregnancy. Women diagnosed with breast cancer during pregnancy have similar disease characteristics to age-matched controls. Current evidence suggests that diagnosis may be carried out with limitations regarding staging. Surgical treatment may be performed as for non-pregnant women. Radiotherapy and endocrine or antibody treatment should be postponed until after delivery. Chemotherapy is allowed after the first trimester. Physicians should be aggressive in the workup of breast symptoms in the pregnant population to expedite diagnosis and allow multidisciplinary treatment without delay.

Dynamic contrast-enhanced MRI of anterior prostate cancer: morphometric assessment and correlation with radical prostatectomy findings.

The purpose of the study was to relate morphometric features of prostate cancers in the anterior compartment of the prostate by dynamic contrast-enhanced (DCE) MRI to subsequent histopathologic findings. We prospectively performed DCE-MRI before biopsy in patients with suspected prostate cancer and selected those showing both a suspicious lesion at MRI and positive biopsies in the anterior compartment of the gland. Tumor contours, margins, largest surface areas and volumes were assessed at MRI and histopathology, when available. Anterior compartment tumors were classified according to transition zone (TZ) boundaries with the peripheral zone (PZ) or with the anterior fibromuscular stroma (SFMA). Forty-three patients were included in this study [median PSA 12.7 ng/ml (3.6-72)]. Whole-mount radical prostatectomy specimens were available in 27 cases. Of the anterior cancers, 89% had ill-defined margins at T2-weighted MRI. Cancer location and contour established at MRI agreed well with histopathology in the 27 cases. Median largest surface area and volume were 1.38 cm(2) (0.35-5.82) and 1.01 cc (0.15-7.4) for MRI versus 1.86 cm(2) (0.2-14) and 2.84 cc (0.33-28.92) for histopathology with respective correlation coefficients (r(2)) of 0.73 and 0.69. The site of origin could be accurately determined for the 15 tumors of less than 3 cc. We found a good relationship between DCE-MRI and histopathology for localization, morphologic description and volume assessment of anterior prostate cancers.

Changes in serum anti-mullerian hormone level during low-dose recombinant follicular-stimulating hormone therapy for anovulation in polycystic ovary syndrome.

CONTEXT: We previously hypothesized that the excess of anti-müllerian hormone (AMH) at the level of selectable follicles could be involved in the follicular arrest of polycystic ovary syndrome (PCOS), mainly through inhibition of FSH effect on aromatase expression. OBJECTIVE: In this study, we investigated whether a decrease in the serum AMH level was concomitant to the appearance of a dominant follicle induced by administration of mild amounts of exogenous FSH in women with PCOS. DESIGN: A total of 30 women with PCOS in whom anovulation was resistant to clomiphene citrate received recombinant FSH using the low-dose step-up protocol during only one cycle. Serum levels of estradiol, AMH, LH, FSH, inhibin B, and ultrasound parameters were assessed twice a week until 3 d after the appearance of one or more dominant follicle(s). RESULTS: The day of dominance (d 0) was defined by the appearance of at least one follicle more than 10 mm growing 2 mm/d. From d -14 before dominance to d +3, the mean serum AMH level and the 2- to 5-mm follicle number at ultrasound declined steadily, although not significantly by ANOVA. Mean AMH relative values (100% being the value at d 0) declined significantly (P = 0.04), from 125 +/- 32% at d -14 to 105 +/- 15% at d -4. Within the same time lag, the mean FSH relative values increased from 91 +/- 17% to 107 +/- 19% (P = 0.013). In the 87 samples obtained from d -14 to -4, absolute values of AMH were positively and negatively associated with those of LH and FSH, respectively, in an independent manner (P = 0.009 and P = 0.03, respectively). In the 55 samples collected at d 0 and +3, they were negatively correlated to those of estradiol (r = -0.272; P < 0.05). CONCLUSIONS: These data suggest that in anovulatory women with PCOS, gently increasing the serum FSH level reduces the AMH excess, thus relieving the inhibition from the latter on aromatase expression by selectable follicles and allowing the emergence of a dominant follicle.

The excess in 2-5 mm follicles seen at ovarian ultrasonography is tightly associated to the follicular arrest of the polycystic ovary syndrome.

BACKGROUND: We previously hypothesized that the excess of 2-5 mm follicles seen at ovarian ultrasonography might be involved in the follicular arrest (FA) of polycystic ovary syndrome (PCOS), independently from the main putative contributors of FA, namely hyperandrogenism and hyperinsulinism. METHODS: A multivariate statistical analysis was applied retrospectively to clinical, biological and ultrasound data that were consecutively collected during 5 years in 457 patients with polycystic ovaries and in 188 age-matched non-hyperandrogenic and regularly cycling controls without PCO at ultrasound. RESULTS: Stepwise discriminant analysis indicated that in PCOS the 2-5 mm follicle number (FN) gave the strongest correlation to severity of the FA, followed by age and then by fasting insulin level. The other variables [waist circumference (WC), 6-9 mm FN, serum testosterone, FSH, LH and ovarian area] were rejected by the analysis. Multiple linear regression indicated a significant and independent negative relationship between the 2-5 and 6-9 mm FN in the PCOS (r = - 0.186, P < 0.01) and control groups (r = - 0.281, P < 0.01). In PCOS only, the 6-9 mm FN was negatively and independently related to the WC (r = - 0.108, P < 0.05). CONCLUSIONS: The size of the 2-5 mm follicle pool is an independent and important contributor to the FA of PCOS. This result could be explained by an exaggerated physiological inhibitory effect from this pool on the terminal follicle growth. The metabolic derangement of PCOS that also contributes to the FA would act through a different mechanism.