Machine learning to further improve the decision which boar ejaculates to process into artificial insemination doses

Multimodal distribution and its impact on the accurate assessment of spermatozoa morphological data: Lessons from machine learning

Objective assessment of sperm morphology is an essential component for assessing ejaculate quality. Due to economic limitations, investigators often divert to conducting observational studies instead of experimental ones, which provide the strongest statistical power, yielding more heterogeneous data regardless of the number of data sources (barns/farms). Using such data inevitably leads to higher variances of estimates, which negatively impacts the statistical power of a study. In this article, we describe a statistical methodology called finite mixture modeling (FMM), which, based on the supplied data and assumed number of sub-classes, classifies the data into two or more homogeneous types of distributions and determines their fractional size relative to the entire cohort. The goal is to use statistical methods that will confound the variance of the sample. A figure from a previous publication was used to generate simulated data (n=1559) on the cytoplasmic droplet rate. We identified that a bi-modal distribution with two latent classes best described the simulated data. Post-hoc estimation showed that about 80 % of observations belonged to latent class 1, with 20 % in latent class 2. The FMM methodology identified a cutoff point of 8.7 %. Finally, when estimating the standard error for the total cohort, the FMM methodology yielded a 40 % reduction in the standard error compared to standard methodologies. In conclusion, here we show that FMM successfully confounded the variance of the data and, as such, yielded lower estimates of the variance than standard methodologies, increasing the statistical power of the cohort.

Artificial intelligence and porcine breeding

Livestock management is evolving into a new era, characterized by the analysis of vast quantities of data (Big Data) collected from both traditional breeding methods and new technologies such as sensors, automated monitoring system, and advanced analytics. Artificial intelligence (A-In), which refers to the capability of machines to mimic human intelligence, including subfields like machine learning and deep learning, is playing a pivotal role in this transformation. A wide array of A-In techniques, successfully employed in various industrial and scientific contexts, are now being integrated into mainstream livestock management practices. In the case of swine breeding, while traditional methods have yielded considerable success, the increasing amount of information requires the adoption of new technologies such as A-In to drive productivity, enhance animal welfare, and reduce environmental impact. Current findings suggest that these techniques have the potential to match or exceed the performance of traditional methods, often being more scalable in terms of efficiency and sustainability within the breeding industry. This review provides insights into the application of A-In in porcine breeding, from the perspectives of both sows (including welfare and reproductive management) and boars (including semen quality and health), and explores new approaches which are already being applied in other species.

Artificial insemination and optimization of the use of seminal doses in swine

The optimization of processes associated with artificial insemination (AI) is of great importance for the success of the pig industry. Over the last two decades, great reproductive performance has been achieved, making further significant progress limited. Optimizing the AI program, however, is essential to the pig industry's sustainability. Thus, the aim is not only to reduce the number of sperm cells used per estrous sow but also to improve some practical management in sow farms and boar studs to transform the high reproductive performance to a more efficient program. As productivity is mainly influenced by the number of inseminated sows, guaranteeing a constant breeding group and with healthy animals is paramount. In the AI studs, all management must ensure conditions to the health of the boars. Some strategies have been proposed and discussed to achieve these targets. A constant flow of high-quality and well-managed breeding groups, quality control of semen doses produced, more reliable technology in the laboratory routine, removal of less fertile boars, the use of intrauterine AI, the use of a single AI with control of estrus and ovulation (fixed-time AI), estrus detection based on artificial intelligence technologies, and optimization regarding the use of semen doses from high genetic-indexed boars are some strategies in which improvement is sought. In addition to these new approaches, we must revisit the processes used in boar studs, semen delivery network, and sow farm management for a more efficient AI program. This review discusses the challenges and opportunities in adopting some technologies to achieve satisfactory reproductive performance and efficiency.

Update on artificial insemination: Semen, techniques, and sow fertility

Over the years, reproductive efficiency in the swine industry has focused on reducing the sperm cell number required per sow. Recent advances have included the identification of subfertile boars, new studies in extended semen quality control, new catheters and cannulas for intrauterine artificial insemination (AI), and fixed-time AI under commercial use. Therefore, it is essential to link field demands with scientific studies. In this review, we intend to discuss the current status of porcine AI, pointing out challenges and opportunities to improve reproductive efficiency.

Machine Learning Approach for Muscovy Duck (Cairina moschata) Semen Quality Assessment

This study aimed to develop a comprehensive approach for assessing fresh ejaculate from Muscovy duck (Cairina moschata) drakes to fulfil the requirements of artificial insemination in farm practices. The approach combines sperm kinetics (CASA) with non-kinetic parameters, such as vitality, enzyme activities (alkaline phosphatase (AP), creatine kinase (CK), lactate dehydrogenase (LDH), and γ-glutamyl-transferase (GGT)), and total DNA methylation as training features for a set of machine learning (ML) models designed to enhance the predictive capacity of sperm parameters. Samples were classified based on their progressive motility and DNA methylation features, exhibiting significant differences in total and progressive motility, curvilinear velocity (VCL), velocity of the average path (VAP), linear velocity (VSL), amplitude of lateral head displacement (ALH), beat-cross frequency (BCF), and live normal sperm cells in favour of fast motility ones. Additionally, there were significant differences in enzyme activities for AP and CK, with correlations to LDH and GGT levels. Although motility showed no correlation with total DNA methylation, ALH, wobble of the curvilinear trajectory (WOB), and VCL were significantly different in the newly introduced classification for "suggested good quality", where both motility and methylation were high. The performance differences observed while training various ML classifiers using different feature subsets highlight the importance of DNA methylation for achieving more accurate sample quality classification, even though there is no correlation between motility and DNA methylation. The parameters ALH, VCL, triton extracted LDH, and VAP were top-ranking for "suggested good quality" predictions by the neural network and gradient boosting models. In conclusion, integrating non-kinetic parameters into machine-learning-based sample classification offers a promising approach for selecting kinetically and morphologically superior duck sperm samples that might otherwise be hindered by a predominance of lowly methylated cells.

Assessment of sperm motility in livestock: Perspectives based on sperm swimming conditions in vivo

Evaluation of sperm motility is well-established in farm animals for quickly selecting ejaculates for semen processing into insemination doses and for evaluating the quality of preserved semen. Likewise, sperm motility is a fundamental parameter used by spermatologists in basic and applied science. Motility is commonly assessed using computer-assisted semen analysis (CASA). Recent increases in computational power, as well as utilization of mobile CASA systems and open-source CASA programs, broaden the possibilities for motility evaluation. Despite this technological progress, the potential of computer-generated motility data to assess male fertility remains challenging and may be limited. Relevance for fertility assessment could be improved if measurement conditions would more closely mimic the in vivo situation. Hence, this review is focused on the current trends of automated semen assessment in livestock and explores perspectives for future use with respect to the physiological and physical conditions encountered by sperm in the female reproductive tract. Validation of current CASA systems with more complex, microfluidic-based devices mimicking the female reproductive tract environment could improve the value of sperm kinematic data for assessing the fertilizing capacity of semen samples, not only for application in livestock but also for use in conducting assisted reproduction techniques in other species.

Contributions of artificial intelligence reported in Obstetrics and Gynecology journals: a systematic review

Background

The applications of artificial intelligence (AI) processes have grown significantly in all medical disciplines during the last decades. Two main types of AI have been applied in medicine: symbolic AI (eg, knowledge base and ontologies) and nonsymbolic AI (eg, machine learning and artificial neural networks). Consequently, AI has also been applied across most obstetrics and gynecology (OB/GYN) domains, including general obstetrics, gynecology surgery, fetal ultrasound, and assisted reproductive medicine, among others.

Objective

The aim of this study was to provide a systematic review to establish the actual contributions of AI reported in OB/GYN discipline journals.

Methods

The PubMed database was searched for citations indexed with “artificial intelligence” and at least one of the following medical subject heading (MeSH) terms between January 1, 2000, and April 30, 2020: “obstetrics”; “gynecology”; “reproductive techniques, assisted”; or “pregnancy.” All publications in OB/GYN core disciplines journals were considered. The selection of journals was based on disciplines defined in Web of Science. The publications were excluded if no AI process was used in the study. Review, editorial, and commentary articles were also excluded. The study analysis comprised (1) classification of publications into OB/GYN domains, (2) description of AI methods, (3) description of AI algorithms, (4) description of data sets, (5) description of AI contributions, and (6) description of the validation of the AI process.

Results

The PubMed search retrieved 579 citations and 66 publications met the selection criteria. All OB/GYN subdomains were covered: obstetrics (41%, 27/66), gynecology (3%, 2/66), assisted reproductive medicine (33%, 22/66), early pregnancy (2%, 1/66), and fetal medicine (21%, 14/66). Both machine learning methods (39/66) and knowledge base methods (25/66) were represented. Machine learning used imaging, numerical, and clinical data sets. Knowledge base methods used mostly omics data sets. The actual contributions of AI were method/algorithm development (53%, 35/66), hypothesis generation (42%, 28/66), or software development (3%, 2/66). Validation was performed on one data set (86%, 57/66) and no external validation was reported. We observed a general rising trend in publications related to AI in OB/GYN over the last two decades. Most of these publications (82%, 54/66) remain out of the scope of the usual OB/GYN journals.

Conclusions

In OB/GYN discipline journals, mostly preliminary work (eg, proof-of-concept algorithm or method) in AI applied to this discipline is reported and clinical validation remains an unmet prerequisite. Improvement driven by new AI research guidelines is expected. However, these guidelines are covering only a part of AI approaches (nonsymbolic) reported in this review; hence, updates need to be considered.

Rapid Multi-Hybridisation FISH Screening for Balanced Porcine Reciprocal Translocations Suggests a Much Higher Abnormality Rate Than Previously Appreciated

With demand rising, pigs are the world’s leading source of meat protein; however significant economic loss and environmental damage can be incurred if boars used for artificial insemination (AI) are hypoprolific (sub-fertile). Growing evidence suggests that semen analysis is an unreliable tool for diagnosing hypoprolificacy, with litter size and farrowing rate being more applicable. Once such data are available, however, any affected boar will have been in service for some time, with significant financial and environmental losses incurred. Reciprocal translocations (RTs) are the leading cause of porcine hypoprolificacy, reportedly present in 0.47% of AI boars. Traditional standard karyotyping, however, relies on animal specific expertise and does not detect more subtle (cryptic) translocations. Previously, we reported development of a multiple hybridisation fluorescence in situ hybridisation (FISH) strategy; here, we report on its use in 1641 AI boars. A total of 15 different RTs were identified in 69 boars, with four further animals XX/XY chimeric. Therefore, 4.5% had a chromosome abnormality (4.2% with an RT), a 0.88% incidence. Revisiting cases with both karyotype and FISH information, we reanalysed captured images, asking whether the translocation was detectable by karyotyping alone. The results suggest that chromosome translocations in boars may be significantly under-reported, thereby highlighting the need for pre-emptive screening by this method before a boar enters a breeding programme.