Irrespective of tubal infertility, for fresh IVF/ICSI cycles the rate of EP is positively associated with ovarian stimulation; for thawed IVF/ICSI cycles, blastocyst transfer or transfer with fewer embryos reduces the EP rate. In IUI cycles, EP is associated with sperm source.

In order to explore the relationship between serum progesterone (P) level on the day of human chorionic gonadotrophin (HCG) administration and cumulative live birth rate in patients with different ovarian response during in vitro fertilization (IVF), we carried out this retrospective cohort study including a total of 4,651 patients undergoing their first IVF cycles from January 2011 to December 2012. All patients with a final live birth outcome (4,332 patients) were divided into three groups according to ovarian response: poor ovarian responder (≤5 oocytes, 785 patients), intermediate ovarian responder (6–19 oocytes, 3065 patients) and high ovarian responder (≥20 oocytes, 482 patients). The thresholds for serum P elevation were 1.60 ng/ml, 2.24 ng/ml, and 2.50 ng/ml for poor, intermediate, and high ovarian responders, respectively. Cumulative live birth rate per oocyte retrieval cycle was calculated in each group. The relationship between serum P level and cumulative live birth rate was evaluated by both univariate and multivariate logistic regression analysis. Cumulative live birth rate per oocyte retrieval cycle was inversely associated with serum P level in patients with different ovarian response. For all responders, patients with elevated P level had significantly higher number of oocytes retrieved, but lower high quality embryo rate, and lower cumulative live birth rate compared with patients with normal serum P level. In addition, serum P level adversely affected cumulative live birth rate by both univariate and multivariate logistic regression analysis, independent of ovarian response. Serum P elevation on the day of HCG administration adversely affects cumulative live birth rate per oocyte retrieval cycle in patients with different ovarian response.

In order to explore the relationship between endometrial thickness on the day of embryo transfer and pregnancy outcomes in frozen-thawed embryo transfer (FET) cycles, we retrospectively analyzed data from 2997 patients undergoing their first FET cycles from January 2010 to December 2012. All patients were divided into three groups (Group A, ≤8 mm; Group B, 9-13 mm; Group C, ≥14 mm) according to the endometrial thickness on embryo transfer day. Compared with patients in the other two groups, patients with thin endometrial thickness in Group A had significantly lower clinical pregnancy rate (33.4%, 41.3% and 45.4%, p < 0.01) and live birth rate (23.8%, 32.2% and 34.0%, p < 0.01). After adjusting for age, body mass index (BMI), baseline follicle stimulating hormone (FSH) FET protocol and number of embryos transferred, the associations between medium endometrial thickness (Group B) and clinical pregnancy rate [adjusted odds ratio (aOR): 1.39; 95% confidence interval (CI): 1.10-1.77, p < 0.01] and live birth rate (aOR: 1.50; 95% CI: 1.16-1.95, p < 0.01) were significant. We conclude that for patients undergoing FET, endometrial thickness on the embryo transfer day significantly affects IVF outcomes in cleavage embryo transfer cycles independent of other factors.

With success of Deep Belief Networks (DBNs) in computer vision, DBN has attracted great attention in hyperspectral classification. Many deep learning based algorithms have been focused on deep feature extraction for classification improvement. Multi-features, such as texture feature, are widely utilized in classification process to enhance classification accuracy greatly. In this paper, a novel hyperspectral classification framework based on an optimal DBN and a novel texture feature enhancement (TFE) is proposed. Through band grouping, sample band selection and guided filtering, the texture features of hyperspectral data are improved. After TFE, the optimal DBN is employed on the hyperspectral reconstructed data for feature extraction and classification. Experimental results demonstrate that the proposed classification framework outperforms some state-of-the-art classification algorithms, and it can achieve outstanding hyperspectral classification performance. Furthermore, our proposed TFE method can play a significant role in improving classification accuracy.

No abstract