Detection and Classification of Hysteroscopic Images Using Deep Learning.

BACKGROUND: Although hysteroscopy with endometrial biopsy is the gold standard in the diagnosis of endometrial pathology, the gynecologist experience is crucial for a correct diagnosis. Deep learning (DL), as an artificial intelligence method, might help to overcome this limitation. Unfortunately, only preliminary findings are available, with the absence of studies evaluating the performance of DL models in identifying intrauterine lesions and the possible aid related to the inclusion of clinical factors in the model. AIM: To develop a DL model as an automated tool for detecting and classifying endometrial pathologies from hysteroscopic images. METHODS: A monocentric observational retrospective cohort study was performed by reviewing clinical records, electronic databases, and stored videos of hysteroscopies from consecutive patients with pathologically confirmed intrauterine lesions at our Center from January 2021 to May 2021. Retrieved hysteroscopic images were used to build a DL model for the classification and identification of intracavitary uterine lesions with or without the aid of clinical factors. Study outcomes were DL model diagnostic metrics in the classification and identification of intracavitary uterine lesions with and without the aid of clinical factors. RESULTS: We reviewed 1500 images from 266 patients: 186 patients had benign focal lesions, 25 benign diffuse lesions, and 55 preneoplastic/neoplastic lesions. For both the classification and identification tasks, the best performance was achieved with the aid of clinical factors, with an overall precision of 80.11%, recall of 80.11%, specificity of 90.06%, F1 score of 80.11%, and accuracy of 86.74 for the classification task, and overall detection of 85.82%, precision of 93.12%, recall of 91.63%, and an F1 score of 92.37% for the identification task. CONCLUSION: Our DL model achieved a low diagnostic performance in the detection and classification of intracavitary uterine lesions from hysteroscopic images. Although the best diagnostic performance was obtained with the aid of clinical data, such an improvement was slight.

Horizontal versus vertical direction of posterior vaginal wall suture after eradication of rectovaginal endometriosis: A multicenter study.

OBJECTIVE: To compare safety and effectiveness of two-different directions of suturing the posterior vaginal breach (horizontal [Ho] vs vertical [Ve]) in women undergoing recto-vaginal endometriosis (RVE) nodule resection. METHODS: A multicenter, retrospective, observational, cohort study was performed including all women of reproductive age undergoing RVE nodule resection between March 2013 and December 2018 at our tertiary centers. Patients included in the present study were divided into two groups based on the direction in suturing the posterior vaginal fornix defect, for comparisons in terms of rate of postoperative complications, pain relief, pain and anatomical recurrence, and length of hospital stay. Univariate comparisons were performed adopting the t test or the Mann-Whitney test for continuous data and the chi-square test or the Fisher exact test for categorical data, with a significant P value set to <0.05. RESULTS: A total of 101 women were included: 67 in the Ho-group and 34 in the Ve-group. The two groups did not significantly differ in length of hospital stay (6.7 ± 6.9 vs 6.6 ± 3.3 days; P = 0.95), overall postoperative complications (32.8% vs 14.7%; P = 0.05), pain recurrence (35.8% vs 26.5%; P = 0.34) and anatomical recurrence rate (19.4% vs 23.5%; P = 0.62). Conversely, grade III complications were significantly more common in the Ho-group than in the Ve-group (22.7% vs 20%, P = 0.009), while pain relief in terms of deep dyspareunia, dyschezia, dysuria and chronic pelvic pain was more consistent in the Ve-group patients (P = 0.04, 0.04, 0.05, 0.004, respectively). CONCLUSION: In symptomatic women undergoing RVE nodule resection, Ho suturing of the vaginal breach appears more commonly associated with severe postoperative complications and a worse pain control.

Prevalence of sonographic signs in women with uterine sarcoma: a systematic review and meta-analysis.

OBJECTIVE: To assess the prevalence of sonographic signs in women with uterine sarcoma. MATERIALS AND METHODS: A systematic review and meta-analysis were performed. Five electronic databases were searched from inception to June 2022 for all studies allowing calculation of the prevalence of sonographic signs in women with uterine sarcoma. Pooled prevalence with 95% confidence intervals was calculated for each sonographic sign and was a priori defined as "very high" when it was ≥ 80%, "high" when it ranged from 80% to 70%, and less relevant when it was ≤ 70%. RESULTS: 6 studies with 317 sarcoma patients were included. The pooled prevalence was: · 25.0% (95%CI:15.4-37.9%) for absence of visibility of the myometrium. · 80.5% (95%CI:74.8-85.2%) for solid component. · 78.3% (95%CI:59.3-89.9%) for inhomogeneous echogenicity of solid component. · 47.9% (95%CI:41.1-54.8%) for cystic areas. · 80.7% (95%CI:68.3-89.0%) for irregular walls of cystic areas. · 72.3% (95%CI:16.7-97.2%) for anechoic cystic areas. · 54.8% (95%CI:34.0-74.1%) for absence of shadowing. · 73.5% (95%CI:43.3-90.9%) for absence of calcifications. · 48.7% (95%CI:18.6-79.8%) for color score 3 or 4. · 47.3% (95%CI:37.0-57.8%) for irregular tumor borders. · 45.4% (95%CI:27.6-64.3%) for endometrial cavity not visualizable. · 10.9% (95%CI:3.5-29.1%) for free pelvic fluid. · 6.4% (95%CI:1.1-30.2%) for ascites. · 21.2% (95%CI:2.1-76.8%) for intracavitary process. · 81.5% (95%CI:56.1-93.8%) for singular lesion.. CONCLUSION: Solid component, irregular walls of cystic areas, and singular lesions are signs with very high prevalence, while inhomogeneous echogenicity of solid component, anechoic cystic areas, and absence of calcifications are signs with high prevalence. The remaining signs were less relevant.

"A Space Odyssey" on Laparoscopic Segmental Rectosigmoid Resection for Deep Endometriosis: A Seventeen-year Retrospective Analysis of Outcomes and Postoperative Complications among 3050 Patients Treated in a Referral Center.

STUDY OBJECTIVE: To evaluate the feasibility of laparoscopic rectosigmoid resection for bowel endometriosis (RSE), reporting surgical and short-term postoperative outcomes in a consecutive large series of patients. DESIGN: A retrospective cohort study. SETTING: Third-level national referral center for deep endometriosis (DE). PATIENTS: 3050 patients with symptomatic RSE requiring surgical treatment. INTERVENTIONS: Nerve-sparing laparoscopic resection for RSE perfomed by a multidisciplinary team. After collecting intraoperative surgical characteristics, postoperative complications were collected by evaluating the risk factors associated with their onset. MEASUREMENTS AND MAIN RESULTS: Clavien-Dindo IIIb postoperative complications were noted in 13.1% of patients, with anastomotic leakage and rectovaginal fistula accounting for 3.0% and 1.9%, respectively. Postoperative bladder impairment was observed in 13.9% of patients during hospital discharge but spontaneously decreased to 4.5% at the first evaluation after 30 days, alongside a statistically significant change towards global symptom improvement. Multivariate analyses were done to identify the risk factors for segmental bowel resection in terms of occurrence of postoperative major complications. Ultralow (≤5 cm from the anal verge), low rectal anastomosis (<8 cm, >5 cm), parametrectomy, vaginal resection, and previous surgeries seemed more related to anastomotic leakage, rectovaginal fistula, and bladder retention. CONCLUSIONS: Laparoscopic rectosigmoid resection for RSE seems an effective and feasible procedure. The surgical complication rate is not negligible but could be reduced by implementing a multidisciplinary approach, an endless improvement in nerve-sparing techniques and surgical anatomy, as well as technological enhancements. Real future challenges will be to reduce the time for the first diagnosis of DE and the likelihood of surgical indications.

Diagnostic accuracy of HIK1083 and MUC6 as immunohistochemical markers of endocervical gastric-type adenocarcinoma: A systematic review and meta-analysis.

INTRODUCTION: HIK1083 and MUC6 have been used as immunohistochemical markers to differentiate gastric-type adenocarcinoma (GTAC) from other endocervical adenocarcinomas. We aimed to assess their diagnostic accuracy through a systematic review and meta-analysis. METHODS: Three electronic databases were searched from their inception to July 2022 for all studies assessing the expression in endocervical GTAC vs other endocervical adenocarcinomas. Diagnostic accuracy was assessed as sensitivity, specificity, positive likelihood ratio (LR+), negative likelihood ratio (LR-), diagnostic odds ratio (DOR), and area under the curve (AUC) on SROC curves. RESULTS: Four studies with 343 patients were included. HIK1083 showed sensitivity= 0.64, specificity= 0.94, LR+ =8.30, LR-= 0.38, DOR= 33.36, AUC= 89.9%. MUC6 showed sensitivity= 0.51, specificity= 0.74, LR+ =1.96, LR-= 0.71, DOR= 3.48, AUC= 72.8%. CONCLUSION: HIK1083 showed high specificity and low sensitivity as a marker of GTAC, with moderate overall accuracy; MUC6 showed moderate specificity and low sensitivity, with low overall accuracy.

Clinical, laboratory and ultrasonographic risk factors for uterine artery arteriosclerosis: Insights for uterus transplantation.

OBJECTIVE: To assess uterine artery arteriosclerosis prevalence in women undergoing total hysterectomy for benign diseases, and any associations between clinical, laboratory, and ultrasound factors and uterine artery arteriosclerosis occurrence. METHODS: A single center, observational, prospective, cohort study was performed enrolling all symptomatic patients scheduled for total hysterectomy from May to December 2021. Our outcomes were: uterine artery arteriosclerosis prevalence and the difference in clinical, laboratory, and ultrasonographic factors between patients with and without uterine artery arteriosclerosis. RESULTS: Forty women were included. Uterine artery arteriosclerosis prevalence was 70%. Body mass index (BMI) (P = 0.04) and triglycerides (P = 0.04) were significantly higher in patients with uterine artery arteriosclerosis than in patients without; while high-density lipoprotein (HDL) (P = 0.02) was significantly lower. Uterine arteries with arteriosclerosis showed higher peak systolic velocity (PSV) values compared with vessels without arteriosclerosis (P = 0.05). CONCLUSIONS: Uterine artery arteriosclerosis shows a high prevalence in women undergoing total hysterectomy for benign diseases. Higher BMI, serum triglycerides, and PSV, and lower serum HDL appear as risk factors for uterine artery arteriosclerosis.

Sentinel Lymph Node Biopsy in Surgical Staging for High-Risk Groups of Endometrial Carcinoma Patients.

BACKGROUND: In endometrial carcinoma (EC) patients, sentinel lymph node (SLN) biopsy has shown the potential to reduce post-operative morbidity and long-term complications, and to improve the detection of low-volume metastasis through ultrastaging. However, while it has shown high sensitivity and feasibility in low-risk EC patient groups, its role in high-risk groups is still unclear. AIM: To assess the role of SLN biopsy through the cervical injection of indocyanine green (ICG) in high-risk groups of early-stage EC patients. MATERIALS AND METHODS: Seven electronic databases were searched from their inception to February 2021 for studies that allowed data extraction about detection rate and accuracy of SLN biopsy through the cervical injection of ICG in high-risk groups of early-stage EC patients. We calculated pooled sensitivity, false negative (FN) rate, detection rate of SLN per hemipelvis (DRh), detection rate of SLN per patients (DRp), and bilateral detection rate of SLN (DRb), with 95% confidence interval (CI). RESULTS: Five observational cohort studies (three prospective and two retrospective) assessing 578 high risk EC patients were included. SLN biopsy sensitivity in detecting EC metastasis was 0.90 (95% CI: 0.03-0.95). FN rate was 2.8% (95% CI: 0.6-11.6%). DRh was 88.4% (95% CI: 86-90.5%), DRp was 96.6% (95% CI: 94.7-97.8%), and DRb was 80% (95% CI: 75.4-83.9). CONCLUSION: SLN biopsy through ICG cervical injection may be routinely adopted instead of systematic pelvic and para-aortic lymphadenectomy in surgical staging for high-risk groups of early-stage EC patients, as well as in low-risk groups.

Impact of nerve-sparing posterolateral parametrial excision for deep infiltrating endometriosis on postoperative bowel, urinary, and sexual function.

OBJECTIVE: To evaluate the functional outcomes of nerve-sparing surgery for deep infiltrating endometriosis (DIE) with or without posterolateral parametrectomy. METHODS: A multicenter, observational, retrospective, cohort study was performed including all symptomatic women who underwent nerve-sparing laparoscopic excision of DIE and preoperative and postoperative assessment of functional outcomes through validated questionnaires between April 2019 and March 2020. Women with posterolateral parametrial DIE (P-group) and women with no parametrial involvement (NP-group) were compared in terms of preoperative and postoperative functional outcomes related to pelvic organs assessed through validated questionnaires (KESS and GIQLI for bowel function, BFLUTS for urinary function, and FSFI for sexual function); pain symptoms at 3-month follow up assessed through an 11-point visual analogue scale (VAS) for dyschezia, dysmenorrhea, dyspareunia and chronic pelvic pain; surgical outcomes; and rate of urinary voiding dysfunction at 3-month follow up. RESULTS: One-hundred patients were included: 69 in the P-group and 31 in the NP-group. Preoperative and postoperative values of questionnaires, pain symptoms, and postoperative complication rates were comparable between the two groups, except for postoperative dyspareunia and sexual dysfunction, which were statistically higher in the P-group. Only patients in the P-group experienced urinary voiding dysfunction, but no statistical significance was reached (P = 0.173). CONCLUSION: Posterolateral parametrectomy for DIE appears to be associated with a higher risk of postoperative dyspareunia and sexual dysfunction.

Clinical Characteristics of Patients with Endometrial Cancer and Adenomyosis.

A better endometrial cancer (EC) prognosis in patients with coexistent adenomyosis has been reported. Unfortunately, it is still unclear if this better prognosis is related to a more favorable clinical profile of adenomyosis patients. We aimed to evaluate differences in the clinical profiles of EC patients with and without adenomyosis. A systematic review and meta-analysis was performed by searching seven electronics databases for all studies that allowed extraction of data about clinical characteristics in EC patients with and without adenomyosis. Clinical characteristics assessed were: age, Body Mass Index (BMI), premenopausal status, and nulliparity. Mean difference in mean ± standard deviation (SD) or odds ratio (OR) for clinical characteristics between EC patients with and without adenomyosis were calculated for each included study and as a pooled estimate, and graphically reported on forest plots with a 95% confidence interval (CI). The Z test was used for assessing the overall effect by considering a p value < 0.05 as significant. Overall, eight studies with 5681 patients were included in the qualitative analysis, and seven studies with 4366 patients in the quantitative analysis. Pooled mean difference in mean ± SD between EC women with and without adenomyosis was -1.19 (95% CI: -3.18 to 0.80; p = 0.24) for age, and 0.23 (95% CI: -0.62 to 1.07; p = 0.60) for BMI. When compared to EC women without adenomyosis, EC women with adenomyosis showed a pooled OR of 1.53 (95% CI: 0.92 to 2.54; p = 0.10) for premenopausal status, and of 0.60 (95% CI: 0.41 to 0.87; p = 0.007) for nulliparity. In conclusion, there are not significant differences in clinical characteristics between EC patients with and without adenomyosis, with the exception for nulliparity. Clinical features seem to not underlie the better EC prognosis of patients with adenomyosis compared to patients without adenomyosis.

Ovarian Reserve after Chemotherapy in Breast Cancer: A Systematic Review and Meta-Analysis.

Background: Worldwide, breast cancer (BC) is the most common malignancy in the female population. In recent years, its diagnosis in young women has increased, together with a growing desire to become pregnant later in life. Although there is evidence about the detrimental effect of chemotherapy (CT) on the menses cycle, a practical tool to measure ovarian reserve is still missing. Recently, anti-Mullerian hormone (AMH) has been considered a good surrogate for ovarian reserve. The main objective of this paper is to evaluate the effect of CT on AMH value. METHODS: A systematic review and meta-analysis were conducted on the PubMed and Scopus electronic databases on articles retrieved from inception until February 2021. Trials evaluating ovarian reserves before and after CT in BC were included. We excluded case reports, case-series with fewer than ten patients, reviews (narrative or systematic), communications and perspectives. Studies in languages other than English or with polycystic ovarian syndrome (PCOS) patients were also excluded. AMH reduction was the main endpoint. Egger's and Begg's tests were used to assess the risk of publication bias. RESULTS: Eighteen trials were included from the 833 examined. A statistically significant decline in serum AMH concentration was found after CT, persisting even after years, with an overall reduction of -1.97 (95% CI: -3.12, -0.82). No significant differences in ovarian reserve loss were found in the BRCA1/2 mutation carriers compared to wild-type patients. CONCLUSIONS: Although this study has some limitations, including publication bias, failure to stratify the results by some important factors and low to medium quality of the studies included, this metanalysis demonstrates that the level of AMH markedly falls after CT in BC patients, corresponding to a reduction in ovarian reserve. These findings should be routinely discussed during oncofertility counseling and used to guide fertility preservation choices in young women before starting treatment.

Prevalence of adenomyosis in endometrial cancer patients: a systematic review and meta-analysis.

INTRODUCTION: Several studies have assessed the histological co-existence of endometrial carcinoma (EC) and adenomyosis. However, the significance of this association is still unclear. OBJECTIVE: To assess the prevalence of adenomyosis in women with EC for a better understanding of the association between the two diseases. MATERIALS AND METHODS: A systematic review and meta-analysis was performed by searching electronics databases from their inception to March 2020, for all studies that allowed extraction of data about prevalence of adenomyosis in EC patients. Adenomyosis prevalence was calculated for each included study and as pooled estimate, with 95% confidence interval (CI). RESULTS: Eight retrospective cohort studies assessing 5573 EC patients were included in our analysis. Of total, 1322 were patients with adenomyosis, and 4251 were patients without adenomyosis. Pooled prevalence of adenomyosis in EC patients was 22.6% (95% CI 12.7-37.1%). CONCLUSION: Adenomyosis prevalence in EC patients was not different from that reported for other gynecological conditions. The supposed association between the two diseases appears unsupported.

Feasibility and safety of laparoscopic approach in obese patients with endometriosis: a multivariable regression analysis.

PURPOSE: To evaluate the feasibility and safety of laparoscopic excision of endometriotic lesions in obese women. METHODS: Retrospective analysis of prospectively collected data involving consecutive women scheduled for complete laparoscopic removal of macroscopic endometriotic lesions between January 2012 and November 2018. Operative time, laparotomic conversion rates, complication rates, and length of hospital stay were recorded. RESULTS: One thousand two hundred thirty women were enrolled and divided into two main groups, according to the World Health Organization classification of obesity, obese (body mass index ≥ 30 kg/m2) and non-obese (body mass index < 30 kg/m2). During the study period, 91 (7.4% of overall study cohort) obese women underwent surgery. At univariate analyses, significant differences between the two groups were found in terms of age, rates of severe endometriosis, American Society of Anesthesiologists physical status classification ≥ III, and different surgical procedures. Significant differences in terms of median operative time [125 (interquartile range (IQR) 85-165) in obese group vs 110 min (IQR 75-155) in non-obese group, P = 0.04] were observed. There were no significant differences between the obese and non-obese groups with respect to the other variables of interest. After adjusted multivariable regression models for potential confounders, difference in operating time (coefficient of 13.389; 95% CI 1.716, 25.060) was still found to be significant. CONCLUSION: In our tertiary care referral center, laparoscopic removal of endometriosis is feasible and safe, except for a slight increase of operative time and conversion rate.

Endometriosis of the Appendix: When to Predict and How to Manage-A Multivariate Analysis of 1935 Endometriosis Cases.

STUDY OBJECTIVE: To evaluate appendiceal endometriosis (AE) prevalence and risk factors in endometriotic patients submitted to surgery. DESIGN: A retrospective cohort study. SETTING: A tertiary level referral center, university hospital. PATIENTS: One thousand nine hundred thirty-five consecutive patients who underwent surgical removal for symptomatic endometriosis. INTERVENTIONS: Electronic medical records of patients submitted to surgery over a 12-year period were reviewed. We assessed any correlation between demographic, clinical, and surgical variables and AE. In our center, appendectomy was performed using a selective approach. Appendix removal was performed in case of gross abnormalities of the organ, such as enlargement, dilation, tortuosity, or discoloration of the organ or the presence of suspected endometriotic implants. MEASUREMENTS AND MAIN RESULTS: AE prevalence was 2.6% (50/1935), with only 1 false-positive case at gross intraoperative evaluation. In multivariate analysis using a stepwise logistic regression model, independent risk factors for AE were adenomyosis (adjusted odds ratio [aOR] = 2.48; 95% confidence interval [CI], 1.32-4.68), right endometrioma (aOR = 8.03; 95% CI, 4.08-15.80), right endometrioma ≥5 cm (aOR = 13.90; 95% CI, 6.63-29.15), bladder endometriosis (aOR = 2.05; 95% CI, 1.05-3.99), deep posterior pelvic endometriosis (aOR = 5.79; 95% CI, 2.82-11.90), left deep lateral pelvic endometriosis (aOR = 2.11; 95% CI, 1.10-4.02), and ileocecal involvement (aOR = 12.51; 95% CI, 2.07-75.75). CONCLUSION: Among patients with endometriosis submitted to surgery, AE was observed in 2.6%, and it was associated with adenomyosis, large right endometrioma, bladder endometriosis, deep posterior pelvic endometriosis, left deep lateral pelvic endometriosis, and ileocecal involvement.

Balloon Catheter for Cervical Priming before Operative Hysteroscopy in Young Women: A Pilot Study.

Aim: To investigate regarding the safety and effectiveness of a balloon catheter (Aqueduct-100 device) to dilate the uterine cervix before operative hysteroscopies. Secondary objectives were to evaluate the duration of the dilatation procedure and to investigate on physicians' satisfaction with the device. Methods: Fifty women younger than 40 years, wishing pregnancies and diagnosed with apparently benign intrauterine lesions and/or uterine anomalies, were enrolled into this study and submitted to cervical priming with Aqueduct-100 device before operative hysteroscopy. Results: Initial cervical dilatation was ≤4 mm in all but one patient. Adequate (10 mm) cervical dilatation was achieved in only one patient, the remaining forty-nine needed additional cervical dilatation. All women, however, presented with a ≥6 mm cervical width after balloon catheter removal. The mean time to final cervical dilatation was 8.5 minutes. No perioperative complications occurred. Physicians reported, in the majority of cases, satisfaction for the dilatation achieved (66%), the ease of balloon catheter insertion/use (82%), and for the ease of additional dilatation (96%). Conclusions: Aqueduct-100 device before operative hysteroscopy is safe and useful to increase the baseline cervical width and facilitate additional dilatation, with good effects on physician satisfaction and acceptable dilatation times.