scholarly journals The study of sperm head vacuoles using deep learning algorithm and its correlation with protamine mRNA ratio

2020 ◽  
Author(s):  
Fatemeh Ghasemian ◽  
Mohammad Hadi Bahadori ◽  
Seyedeh Zahra Hosseini Kolkooh ◽  
Maryam Esmaeili
Keyword(s):  
Deep Learning ◽  
Assisted Reproduction ◽  
Learning Algorithm ◽  
Sperm Head ◽  
Sperm Parameters ◽  
Large Vacuole ◽  
Deep Learning Algorithm ◽  
Mrna Ratio ◽  
Grade Iii ◽  
Protamine 1

Abstract BackgroundAs regards the routine semen analysis is not sufficient to assess male fertility status, is it necessary to use other morphological sperm examination that may be more relevant in regard to the promotion of assisted reproduction outcomes? This study was designed for examination of sperm vacuole characteristics, its association with other sperm parameters and protamine 1 to protamine 2 ratio, and predict assisted pregnancy outcomes. Methods 98 Semen samples from subfertile men were classified based on Vanderzwalmen's criteria as follows: grade I, no vacuoles; grade II, ≤ 2small vacuoles; grade III, ≥ 1 large vacuole; grade IV, large vacuole with other abnormalities. The location, frequency and size of vacuoles were assessed using high magnification, a deep learning algorithm, and scanning electron microscope methods. The chromatin integrity (toluidine blue staining), condensation status (aniline blue), viability and acrosome integrity (triple staining), and protamination status (CMA3 staining) was evaluated for vacuolated samples. Protamine 1 and protamine 2 gene expression was analyzed by quantitative real-time PCR. The assisted reproduction outcomes were also followed for each cycle. Results The results show a significant correlation between the vacuole size (III and IV) and abnormal sperm chromatin condensation (p<0.05), and protamine-deficient (p<0.05). The percentage of reacted acrosomes was significantly higher in spermatozoa with grades III and IV compared with normal group (p<0.05). A high protamine mRNA ratio ( prm-2 was underexpressed) was observed in the vacuolated spermatozoa with grade IV (p<0.01). The sperm head vacuole was negatively associated with the fertilization rate (p<0.01) under IVF cycles. This association was also significantly observed in pregnancy and live birth rate in the groups with grade III and IV (P<0.05). Conclusions The results of our study highlight the importance of follow up of more sperm parameters such as sperm head vacuole characteristics, because may reflect protamine-deficient and poor IVF/ICSI outcomes.

scholarly journals The study of sperm head vacuoles using deep learning algorithm and its correlation with protamine mRNA ratio

2020 ◽  
Author(s):  
Fatemeh Ghasemian ◽  
Mohammad Hadi Bahadori ◽  
Seyedeh Zahra Hosseini Kolkooh ◽  
Maryam Esmaeili
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Sperm Head ◽  
Deep Learning Algorithm ◽  
Mrna Ratio

Abstract The authors have requested that this preprint be withdrawn due to author disagreement.

scholarly journals Study on Radar Echo-Filling in an Occlusion Area by a Deep Learning Algorithm

2021 ◽  
Vol 13 (9) ◽  
pp. 1779
Author(s):  
Xiaoyan Yin ◽  
Zhiqun Hu ◽  
Jiafeng Zheng ◽  
Boyong Li ◽  
Yuanyuan Zuo
Keyword(s):  
Deep Learning ◽  
Loss Function ◽  
Learning Algorithm ◽  
Weather Radar ◽  
Loss Functions ◽  
Training Dataset ◽  
Echo Intensity ◽  
Common Mean ◽  
Deep Learning Algorithm ◽  
Radar Beam

Radar beam blockage is an important error source that affects the quality of weather radar data. An echo-filling network (EFnet) is proposed based on a deep learning algorithm to correct the echo intensity under the occlusion area in the Nanjing S-band new-generation weather radar (CINRAD/SA). The training dataset is constructed by the labels, which are the echo intensity at the 0.5° elevation in the unblocked area, and by the input features, which are the intensity in the cube including multiple elevations and gates corresponding to the location of bottom labels. Two loss functions are applied to compile the network: one is the common mean square error (MSE), and the other is a self-defined loss function that increases the weight of strong echoes. Considering that the radar beam broadens with distance and height, the 0.5° elevation scan is divided into six range bands every 25 km to train different models. The models are evaluated by three indicators: explained variance (EVar), mean absolute error (MAE), and correlation coefficient (CC). Two cases are demonstrated to compare the effect of the echo-filling model by different loss functions. The results suggest that EFnet can effectively correct the echo reflectivity and improve the data quality in the occlusion area, and there are better results for strong echoes when the self-defined loss function is used.

Sign in / Sign up

Export Citation Format

Share Document