Hysteroscopy in the Netherlands and Flanders: A survey amongst practicing gynaecologists

Deep learning model for classifying endometrial lesions

BACKGROUND

Hysteroscopy is a commonly used technique for diagnosing endometrial lesions. It is essential to develop an objective model to aid clinicians in lesion diagnosis, as each type of lesion has a distinct treatment, and judgments of hysteroscopists are relatively subjective. This study constructs a convolutional neural network model that can automatically classify endometrial lesions using hysteroscopic images as input.

METHODS

All histopathologically confirmed endometrial lesion images were obtained from the Shengjing Hospital of China Medical University, including endometrial hyperplasia without atypia, atypical hyperplasia, endometrial cancer, endometrial polyps, and submucous myomas. The study included 1851 images from 454 patients. After the images were preprocessed (histogram equalization, addition of noise, rotations, and flips), a training set of 6478 images was input into a tuned VGGNet-16 model; 250 images were used as the test set to evaluate the model's performance. Thereafter, we compared the model's results with the diagnosis of gynecologists.

RESULTS

The overall accuracy of the VGGNet-16 model in classifying endometrial lesions is 80.8%. Its sensitivity to endometrial hyperplasia without atypia, atypical hyperplasia, endometrial cancer, endometrial polyp, and submucous myoma is 84.0%, 68.0%, 78.0%, 94.0%, and 80.0%, respectively; for these diagnoses, the model's specificity is 92.5%, 95.5%, 96.5%, 95.0%, and 96.5%, respectively. When classifying lesions as benign or as premalignant/malignant, the VGGNet-16 model's accuracy, sensitivity, and specificity are 90.8%, 83.0%, and 96.0%, respectively. The diagnostic performance of the VGGNet-16 model is slightly better than that of the three gynecologists in both classification tasks. With the aid of the model, the overall accuracy of the diagnosis of endometrial lesions by gynecologists can be improved.

CONCLUSIONS

The VGGNet-16 model performs well in classifying endometrial lesions from hysteroscopic images and can provide objective diagnostic evidence for hysteroscopists.

Effectiveness of Hysteroscopic Techniques for Endometrial Polyp Removal: The Italian Multicenter Trial

STUDY OBJECTIVE

To compare the effectiveness and safety of different techniques of hysteroscopic polypectomy.

DESIGN

Multicenter, prospective observational trial (Canadian Task Force classification II-2).

SETTING

Nineteen Italian gynecologic departments (university-affiliated or public hospitals).

PATIENTS

Consecutive patients suffering from endometrial polyps (EPs).

INTERVENTIONS

Hysteroscopic polypectomy, as performed through different techniques.

MEASUREMENTS AND MAIN RESULTS

Included in the study were 1404 patients (with 1825 EPs). The setting was an ambulatory care unit in 40.38% of the cases (567 women), of whom 97.7% (554) did not require analgesia/anesthesia. In the remaining 59.62% of women (837 women), the procedures were performed in an operating room under mild sedation, local or general anesthesia. Minor complications occurred in 32 patients (2.27%), without significant differences between the techniques used (p = ns). Uterine perforation occurred in 14 cases, all performed in the operating room with some kind of anesthesia, only 1 with a vaginoscopic technique and the remaining during blind dilatation (odds ratio [OR], 19.98; 95% confidence interval [CI], 1.19-335.79; p = .04). An incomplete removal of EPs was documented in 39 patients. Logistic regression analysis showed that a higher risk of residual EPs was associated with the use of a fiber-based 3.5-mm hysteroscope (OR, 6.78; 95% CI, 2.97-15.52; p <.001), the outpatient setting (OR, 2.17; 95% CI, 1.14-4.14; p = .019), and EPs located at the tubal corner (OR, 1.98; 95% CI, 1.03-2.79; p = .039). No association between incomplete EP removal and EP size or number was recorded (p = ns), as well as with the other variables evaluated.

CONCLUSION

Outpatient polypectomy was associated with a minimal but significantly higher risk of residual EPs in comparison with inpatient polypectomy. Conversely, inpatient polypectomy was associated with a considerably higher risk of uterine perforation and penetration in comparison with office hysteroscopy. Because of lower intraoperative risks and higher cost-effectiveness, office hysteroscopy may be considered, whenever possible, as the gold standard technique for removing EPs.