Pregnancy Rates after Hysteroscopic Endometrial Polypectomy versus Endometrial Curettage Polypectomy: A Retrospective Study.

Background and Objectives: A relationship between endometrial polypectomy and in vitro fertilization (IVF) pregnancy outcomes has been reported; however, only a few studies have compared polyp removal techniques and pregnancy rates. We investigated whether different polypectomy techniques with endometrial curettage and hysteroscopic polypectomy for endometrial polyps affect subsequent pregnancy outcomes. Materials and Methods: Data from 434 patients who had undergone polypectomy for suspected endometrial polyps using transvaginal ultrasonography before embryo transfer in IVF at four institutions between January 2017 and December 2020 were retrospectively analyzed. Overall, there were 157 and 277 patients in the hysteroscopic (mean age: 35.0 years) and curettage (mean age: 37.3 years) groups, respectively. Single-blastocyst transfer cases were selected from both groups and age-matched to unify background factors. Results: In the single-blastocyst transfer cases, 148 (mean age: 35.0 years) and 196 (mean age: 35.9 years) were in the hysteroscopic and curettage groups, respectively, with the 148 cases matched by age. In these cases, the pregnancy rates for the first embryo transfer were 68.2% (odds ratio (OR): 2.14) and 51.4% (OR: 1.06) in the hysteroscopic and curettage groups, respectively; the resulting OR was 2.03. The pregnancy rates after up to the second transfer were 80.4% (OR: 4.10) and 68.2% (OR: 2.14) in the hysteroscopic and curettage groups, respectively, in which the OR was 1.91. The live birth rates were 66.2% (OR: 1.956) and 53.4% (OR: 1.15) in the hysteroscopic and curettage groups, respectively, in which the odds ratio was 1.71. These results show the effectiveness of hysteroscopic endometrial polypectomy compared to polypectomy with endometrial curettage. No significant difference was found regarding the miscarriage rates between the two groups. Conclusions: Hysteroscopic endometrial polypectomy resulted in a higher pregnancy rate in subsequent embryo transfer than polypectomy with endometrial curettage. Therefore, establishing a facility where polypectomy can be performed hysteroscopically is crucial.

Investigation of an efficient method of oocyte retrieval by dual stimulation for patients with cancer.

Purpose: To examine the optimal timing of second ovarian stimulation using the dual stimulation method for good ovarian responders with cancer undergoing oocyte retrieval for fertility preservation. Methods: A retrospective analysis was conducted using data from 69 patients with cancer who underwent oocyte retrieval for fertility preservation at four Japanese institutions during 2010-2021. Twenty-two patients underwent two oocyte retrievals for fertility preservation. We studied the relationship between the initial number of oocytes retrieved via dual stimulation and risk of ovarian enlargement as well as the appropriate waiting interval between the end of the first ovarian stimulation and beginning of the second ovarian stimulation. Results: The risk of ovarian enlargement was high when the initial number of oocytes retrieved via dual stimulation was ≥5. An 8-day waiting interval may be more effective for performing a second ovarian stimulation oocyte retrieval in these cases, although the difference was not significant. Conclusions: This study provides one policy for effectively managing ovarian enlargement and timing of second ovarian stimulation during oocyte retrieval via the dual stimulation method for patients with cancer undergoing fertility preservation. If more facilities implement this procedure, more oocytes may be obtained in a short period for fertility preservation purposes.

Relation between weight size and degree of over-fitting in neural network regression.

This paper investigates the relation between over-fitting and weight size in neural network regression. The over-fitting of a network to Gaussian noise is discussed. Using re-parametrization, a network function is represented as a bounded function g multiplied by a coefficient c. This is considered to bound the squared sum of the outputs of g at given inputs away from a positive constant delta(n), which restricts the weight size of a network and enables the probabilistic upper bound of the degree of over-fitting to be derived. This reveals that the order of the probabilistic upper bound can change depending on delta(n). By applying the bound to analyze the over-fitting behavior of one Gaussian unit, it is shown that the probability of obtaining an extremely small value for the width parameter in training is close to one when the sample size is large.

On the problem in model selection of neural network regression in overrealizable scenario.

In considering a statistical model selection of neural networks and radial basis functions under an overrealizable case, the problem of unidentifiability emerges. Because the model selection criterion is an unbiased estimator of the generalization error based on the training error, this article analyzes the expected training error and the expected generalization error of neural networks and radial basis functions in overrealizable cases and clarifies the difference from regular models, for which identifiability holds. As a special case of an overrealizable scenario, we assumed a gaussian noise sequence as training data. In the least-squares estimation under this assumption, we first formulated the problem, in which the calculation of the expected errors of unidentifiable networks is reduced to the calculation of the expectation of the supremum of the chi2 process. Under this formulation, we gave an upper bound of the expected training error and a lower bound of the expected generalization error, where the generalization is measured at a set of training inputs. Furthermore, we gave stochastic bounds on the training error and the generalization error. The obtained upper bound of the expected training error is smaller than in regular models, and the lower bound of the expected generalization error is larger than in regular models. The result tells us that the degree of overfitting in neural networks and radial basis functions is higher than in regular models. Correspondingly, it also tells us that the generalization capability is worse than in the case of regular models. The article may be enough to show a difference between neural networks and regular models in the context of the least-squares estimation in a simple situation. This is a first step in constructing a model selection criterion in an overrealizable case. Further important problems in this direction are also included in this article.