Skip to Content
MDPIMDPI

Advanced

HealthcareHealthcare
PDF
  • Review
  • Open Access

3 April 2024

Predicting Male Infertility Using Artificial Neural Networks: A Review of the Literature

,
,
and
1
Department of Psychology, University Sapienza of Rome, 00185 Rome, Italy
2
Santa Lucia Foundation, Scientific Institute for Research, Hospitalization and Health Care (IRCCS), 00179 Rome, Italy
*
Author to whom correspondence should be addressed.
Healthcare2024, 12(7), 781;https://doi.org/10.3390/healthcare12070781
This article belongs to the Special Issue Application Prospect of Artificial Intelligence in Healthcare
Version Notes
Review Reports
Order Reprints

Abstract

Male infertility is a relevant public health problem, but there is no systematic review of the different machine learning (ML) models and their accuracy so far. The present review aims to comprehensively investigate the use of ML algorithms in predicting male infertility, thus reporting the accuracy of the used models in the prediction of male infertility as a primary outcome. Particular attention will be paid to the use of artificial neural networks (ANNs). A comprehensive literature search was conducted in PubMed, Scopus, and Science Direct between 15 July and 23 October 2023, conducted under the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We performed a quality assessment of the included studies using the recommended tools suggested for the type of study design adopted. We also made a screening of the Risk of Bias (RoB) associated with the included studies. Thus, 43 relevant publications were included in this review, for a total of 40 different ML models detected. The studies included reported a good quality, even if RoB was not always good for all the types of studies. The included studies reported a median accuracy of 88% in predicting male infertility using ML models. We found only seven studies using ANN models for male infertility prediction, reporting a median accuracy of 84%.

1. Introduction

Demographers tend to interpret (define) infertility as childlessness in a population of women of reproductive age, whereas the epidemiological definition refers to trying for or time to pregnancy in a population of women exposed to the probability of conception [1]. Infertility is a universal health issue, covering about 8 to 12% of couples worldwide [2], and as a chronic condition, it can result in distress, stigma, and financial hardship, affecting people’s mental and psychosocial well-being [3].
Approximately 30% of infertility cases are attributed to male factors [4]. Recent meta-analysis studies show that male factors are present in 20–70 percent of infertility cases [5]. These findings are significantly broader than previously reported [4,5]. However, the wide range of male infertility in meta-analysis studies [6] may not reflect the prevalence of this complication in all parts of the world because of reasons, such as the lack of rigorous statistical methods that include bias, heterogeneity in data collection, and cultural constraints [6]. Indeed, the inability to conceive a child can be emotionally taxing, leading to feelings of inadequacy, and men may contemplate their masculinity, as societal standards frequently associate potency with manliness. The constant demand to excel and procreate can intensify these feelings [7]. As mentioned before, in almost all cases, it is not only the man who suffers from it, as couples also facing fertility issues often experience increased tension in their relationships. The strain comes from various factors, such as the emotional reaction to fertility treatments, the financial burden, and the pressure to conceive [8]. Furthermore, the necessity for timed sexual intercourses and a heightened focus on fertility can take away the spontaneous nature of such interactions, causing them to feel more like clinical procedures than intimate moments, which may lead to an environment unsuitable for conception [9].
Recent studies addressed the investigation of genetic abnormalities [10], such as chromosomal abnormalities and gene mutations [11], as factors affecting sperm production and function [10]. Furthermore, the role of hormonal imbalances, such as low levels of testosterone or high levels of prolactin, in disrupting the normal functioning of the reproductive system and sperm production has been investigated [12]. Indeed, lifestyle factors, including smoking, excessive alcohol consumption, drug use, and exposure to environmental toxins, can also contribute to male infertility [13,14]. Recently, there has been an increasing interest in chemical exposure as a cause of male infertility [14], which likely plays a substantial role in the sperm count trend.
Last but not least, risk factors exhibit a correlation with obesity, which is a known risk factor for diminished-quality sperm [15]. Diet is another important factor that is hard to decouple from chemical exposures since pesticide residues linger on much of the food we eat [16].
Given the complexity of the topic, it is essential to have predictive models that can provide accurate information on the factors that influence male infertility. Traditionally, the total motile sperm count (volume × concentration × motility) has been used as the most predictive factor in determining fertility compared to volume, concentration, and motility individually [17]. On the other hand, a recent study [18] reported that semen analysis (sperm quality) is a fundamental method for evaluating male fertility, including the analysis of sperm count, motility, morphology, and volume. Therefore, abnormalities in these parameters can indicate male infertility.
Since many factors have been investigated as potential risk factors related to male infertility [5,7,9,10,11,12,13,14,15,16,17], it is important to identify statistical models that can accurately predict how much each single risk factor impacts the onset of male infertility.
The introduction of artificial intelligence (AI) into the healthcare field promoted a significant change in the approach of medical practitioners to diagnosing, treating, and anticipating health ailments. Indeed, the use of AI, which is powered by machine learning algorithms, data, and computational capabilities, allows for the analysis of large datasets with impressive speed [18]. This transformative technology can predict illnesses, categorize patient information, and offer tailored treatment recommendations. Predictive analytics in medicine leverages AI to forecast patient outcomes, identify potential health risks, and optimize care pathways using algorithms such as the Random Forest (RF) Classifier [19,20]. Exploiting different types of information AI models can forecast the commencement and advancement of diseases, such as in the study of heart disease [21], cancer [22], and diabetes [23], in which AI models have already proven their efficiency in facilitating timely intervention and tailored healthcare strategies.
Machine learning (ML), artificial neural networks (ANNs), and deep learning have emerged as feeble tools, stagnating our approach to comprehending and addressing male reproductive health. These tools have been employed to scrutinize extensive datasets, thereby assisting in the identification of pivotal factors that influence fertility outcomes.
Currently, in the scientific literature, there are not many review documents that have specifically addressed the use of machine learning models for predicting male infertility. A recent review [24] addressed the topic in a non-systematic and narrative way. The authors [24] drew up examples of possible applications of ML algorithms in a hospital setting, providing indications on robots and mechanics in vogue for specialized analyses, such as Advancing Computer-Assisted Semen Analyses (CASA), in the investigation of preoperative images before surgical sperm retrieval, for intra-operative advancement to identify genetic material at a microscopic level and sperm identification and analysis. This review, although concise and well argued, has the limitation of being focused on assisted reproductive technologies without any specific reference to machine learning models. Another review [25] analyzed the role of ML models exclusively for sperm selection before fertilization practices without considering other factors. Two other reviews instead investigated azoospermia in patients with Klinefelter Syndrome [26] or, in an all-encompassing way [27], about both male and female fertility, raising important consideration about the need for data augmentation, feature extraction, explainability, and the need to revisit the meaning of an effective system for fertility analysis.
For this reason, it seems important to provide precise indications on what types of algorithms and their relevant clinical performances are currently used in the scientific panorama, being able to discriminate specific models and data in which they have been used. Among the multiple ML models currently available, artificial neural networks (ANNs) are widely used in medical settings to improve the delivery of care at a reduced cost [28]. These models are particularly interesting since they are inspired by the neural organization of the human brain, which is modelled as an interconnection of nodes. In the context of male infertility, ANNs have played a crucial role in predicting sperm concentration [29]. This predictive ability is invaluable in assessing sperm quality, an essential determinant of male fertility.
Even though we have access to several computational predictive models, challenges remain, such as understanding which of all these models is or has the potential to be the most accurate overall.
By conducting a review on the accuracy of ML, with a focus on ANNs, in predicting male infertility, this research will provide valuable insights into this emerging field. Therefore, this review does not aim to claim a technical contribution, but it is intended to be a reference guide. Therefore, the aim of this review is to fill this gap and contribute to the existing literature by critically evaluating the accuracy of ML in predicting male infertility and gaining insights about which model might be more efficient.

2. Materials and Methods

We used the PICO framework [30] for formulating the research questions reported in Table 1. The literature search was undertaken between August 2023 and October 2023 by two independent researchers (VS, DDB) using three databases (PubMed, Scopus, and ScienceDirect). Therefore, a search string was created as similar as possible for every database. The search strategy was conducted through the combination of Mesh, Tiab, and synonym terms for the general string of “Man, infertility, prediction”. In Appendix A, we report specifications about Mesh, Tiab, synonyms (Table A1), and the specific strings search (Table A2) adapted for each database. This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist [31].
Table 1. PICO framework.
All results from Pubmed, Scopus, and Science Direct searches were aggregated in an Excel sheet, arranging the titles in alphabetical order. Since the aim of this review was to make a comprehensive panoramic of ML models used to predict male infertility, we did not introduce any new approaches or databases to claim a technical contribution.

2.1. Inclusion Criteria

After removing duplicates manually, the search results underwent title and abstract screening, applying criteria for inclusion and exclusion of search results as follows:
-
Inclusion: Studies published in a peer-reviewed journal; any year of publication; all study designs; human male population.
-
Exclusion: Any language other than English; grey literature, letter to the editor, or reviews; search results with content not directly relevant to the research question after; undergoing a title and abstract screening; studies with different target populations (female, animals); articles with ambiguity in the context of male infertility in humans and machine learning; paper with no direct access to the full text.
The reference lists of the included studies were screened for further relevant publications. For a paper in which animal and male models are mutually included, we considered only data relating to human male infertility.

2.2. Quality of the Studies and Risk of Bias Analysis

Two independent researchers analyzed the quality (VS) and the Risk of Bias (RoB) (DDB) of the included studies. The selected studies were grouped following the study design, and then a suitable tool for quality appraisal had to be chosen according to the study design. Retrospective and prospective studies were both assessed with the JBI Checklist for Case Series, which consists of a rapid assessment based on the identification of ten specific items of information regarding the participants (from enrolment to the final stage of the experiment) and the procedure [32]. The remaining studies were assessed according to the EQUATOR guidelines [33]. Indeed, cohort, case–control, and cross-sectional studies were assessed through the STROBE guidelines for reporting observational studies. This tool is made of 22 items, assessing in detail all the sections of the paper and also including sub-items to further distinguish specific criteria of scoring according to the study design (cohort, case–control, cross-sectional) for the participants, the statistical methods, and the results sections. Studies using a multivariable prediction model were assessed through the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) checklist [34]. Similar to the STROBE, this tool is made up of 22 items including sub-items, thus collecting a score made up of 35 investigated domains.
For each study, we collected a score based on the screening performed using the suitable tool. This scoring was then normalized to the total, expressed as a percentage, and we called it quality grade. This normalization was applied to all the studies included in this review and allowed us to make a direct comparison of studies with different designs. Since these tools do not provide a cut-off score for further classification of the study quality, according to a previous study [35], we considered it as good quality an assessed adherence ranging from 60 to 80%.
The RoB analysis was performed through the administration of the prediction model risk of bias assessment tool (PROBAST) [36]. This instrument is suitable for reviews of studies about clinical prediction models. The PROBAST checklist assesses the risk that arises from the methods used and the consequential applicability of the prediction model.

3. Results

From a total of 254 results in the above-mentioned database searches, 39 duplicates were removed manually, and 215 results remained for title and abstract screening. Of them, 142 records had to be ruled out after applying the exclusion criteria on titles and abstracts, and the remaining 73 records were sought for retrieval. It was not possible to get full access to 1 of them, and, thus, the full-text screening of 72 studies was performed according to the inclusion and exclusion criteria. From the full-text reading, 29 papers were excluded for a final 43 eligible studies included in this review, as reported in the flowchart in Figure 1.
Figure 1. PRISMA flowchart diagram of the review. ** Reasons for the 142 records exclusion were: (1) used animals or females as participants; (2) did not use ML; (3) the format of the paper was review, letter to the editor, or others; (4) the aim of the study was out of scope concerning the prediction of male infertility.
Since the studies are very different from each other, it was not possible to carry out a meta-analysis. Therefore, to answer the PICO questions formulated in the previous section, we organized and schematically summarized them in the upcoming tables. Indeed, the studies were divided based on the general topic they dealt with; therefore, the specific variables considered in each model included sperm retrieval (Table 2, four studies [37,38,39,40]); sperm quality, further divided into the investigation of sperm quality and morphology (Table 3a, seventeen studies [41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57]) and quality of sperm and environmental factors (Table 3b, four studies [58,59,60,61]); non-obstructive azoospermia (Table 4, three studies [62,63,64]); IVF outcome (Table 5, three studies [65,66,67]); environmental and medical factors (Table 6, twelve studies [68,69,70,71,72,73,74,75,76,77,78,79]) as a function of male infertility, according to a recent review addressing the broad causes and risk factors of male infertility [80]. For the sake of brevity, we decided to shorten the content in the tables as much as possible. Therefore, many acronyms normally used in AI and ML papers have been adopted. Indeed, to avoid repetition in the caption of the table, we decided to put a detailed glossary of these terms in the Abbreviations section.
Table 2. In this table, studies in which ML models were employed to investigate the relationship between male infertility and the potential sperm retrieval as a risk factor for male infertility are grouped. Quality grade is the normalized scoring obtained through the administered chosen tool according to the criteria exposed in Section 2.2. Accuracy of the model refers to the tested accuracy of the model in the prediction of male infertility. Abbreviation: BMI stands for body mass index; LH is luteinizing hormone; FSH is follicle-stimulating hormone; ICSI is intracytoplasmic sperm injection; TESE is testicular sperm extraction; T is testosterone; AUC is the area under the curve; NOA is non-obstructive azoospermia.
Table 3. (a) In this table, studies in which ML models were employed for investigating sperm quality e/o morphology as a function of male infertility are grouped. Quality grade is the normalized scoring obtained through the administered chosen tool according to the criteria exposed in Section 2.2. Accuracy of the model refers to the tested accuracy of the model in the prediction of male infertility. Abbreviation: CASA is Computer Assisted Sperm Analysis, DFI is DNA fragmentation index; GC-MS is gas chromatography-mass spectrometry; AAA is Artificial Algae Algorithm; SMA is sperm morphology analysis; MAE is mean absolute error. (b) In this table, studies dealing with the sperm quality analysis are grouped, thus exploring the environmental factors that may be associated with predisposition with male infertility. Quality grade is the normalized scoring obtained through the administered chosen tool according to the criteria exposed in Section 2.2. Accuracy of the model refers to the tested accuracy of the model in the prediction of male infertility.
Table 4. In this table, studies in which ML models were used for the diagnosis or investigation of non-obstructive azoospermia (NOA) are grouped. Quality grade is the normalized scoring obtained through the administered chosen tool according to the criteria exposed in Section 2.2. Accuracy of the model refers to the tested accuracy of the model in the prediction of male infertility.
Table 5. In this table, studies in which ML models were used for the prediction of in vitro fertilization (IVF) outcomes are grouped. Quality grade is the normalized scoring obtained through the administered chosen tool according to the criteria exposed in Section 2.2. Accuracy of the model refers to the tested accuracy of the model in the prediction of male infertility.
Table 6. In this table, studies in which ML models were used for the prediction of male infertility as a function of environmental and medical factors are grouped. Quality grade is the normalized scoring obtained through the administered chosen tool according to the criteria exposed in Section 2.2. Accuracy of the model refers to the tested accuracy of the model in the prediction of male infertility. Abbreviation: MAE is mean absolute error, TIF is Tamura Image Features, and AUC is the area under the curve.
We found that 86% of the included studies in this review used an ML algorithm to predict male infertility, while the remaining 14% were studies about the development of an ML algorithm [37,38,39] or for the purpose of predicting IVF outcome [20,42].
In these studies, about forty different AI and ML methods were used, most of the time in combination with other methods (mean = 3.1, IQR = 1–9). Two studies [37,43] reported the generic term ML, while in eight studies [42,43,45,46,47,51,59,61], algorithms available on the industrial market were used. Of the 40 models analyzed, most are models with 73% supervised learning, 20% are unsupervised, 3% are semi-supervised, and the remaining 5% are models that can be programmed in both modes, as in the case of ANNs.
Unsupervised ML models were always used in conjunction with those supervised to create classifications or compare ML algorithms’ performance. Found as the most used models, support vector machine (SVM, 11%), Random Forest (RF, 10%), linear regression (LR), artificial neural networks (ANNs), and multi-layer perceptron (MLP) were employed with a frequency of 7%. eXtreme Gradient Boosting (XGBoost), Least Absolute Shrinkage and Selection Operator (LASSO—GFLASSO), Convolution Neural Network (CNN), and K-Nearest Neighbor (KNN) had a frequency of 4%. The remaining 31 models were used in percentages ranging from 3 to 1%. Despite the variety of these studies, from the different data sources processed, and the methods used, it seems very clear that ML algorithms can process a huge amount of data, as in the studies [39,43,68,79] in which the sample size exceeded one thousand units, and multiple variables were combined for the prediction. However, this does not imply that a lower number of inputs corresponds to lower model efficiency [52,55,56].
Finally, since the aim of this review was to understand not only which models were used but also what their actual level of prediction accuracy was, we analyzed the results of the reported accuracy, considering as a primary outcome measure the tested accuracy of the model in the prediction of male infertility. In six studies [43,47,55,56,57,76], the performance of each model was not evaluated as accuracy, but an error estimation of the model prediction was made; in two studies [74,75], there was no information on the matter. In the remaining 35 studies, there was an average accuracy of 88% (IQR = 80–94%), with 26 studies declaring an accuracy greater than 80%. Considering the average of the accuracy values reported in the individual studies, we found that Boruta, DNN, and KNN had an accuracy of 99%, AdaBoost 97%, DMTL, and DTLA of 94%,
FSNN, LDFA, and QDFA were at 93%, U-NET and NNET architectures 92%, Bayesian models (NB—BKMR) 91%, and all other models reported an accuracy between 89 and 71%. Studies in which ANNs were used reported a mean accuracy level of 84% (IQR = 81–94%), showing a promising role for the adoption of this model for male infertility prediction.

3.1. Retrospective and Prospective Studies

In this review, 63% of the included papers were retrospective or prospective studies and, therefore, screened with the JBI critical appraisal quality tool. They reported, on average, a good quality (mean = 90%, IQR = 80–100%), while a more in-depth analysis performed with the RoB tool (PROBAST) revealed a lot of uncertainty or high concern (Figure 2).
Figure 2. Risk of Bias (RoB) analysis of retrospective and prospective studies performed using PROBAST [37,38,39,40,41,42,43,44,47,48,54,55,56,57,58,59,60,61,63,64,71,72,74,75,76,77,79].
Most of the studies reported an overall unclear applicability, mainly due to missing information about inclusion and exclusion criteria in the selection of data sources (60%). This is mainly due to the use of datasets when authors do not report specific information about the selection of records but merely information about data extracted for their analysis. Further, 40% of the included studies in this category also reported a high concern due to the applicability of the chosen analysis because no information is available about the distinction of predictors in the outcomes of the study. This does not allow us to infer the causality of results, which are, most of the time, summarized in terms of probability, association, and rates. Despite these considerations, the results are properly reported in most of the studies (93%), even if, most of the time, the results are mainly related to the ML algorithm, i.e., its validation or comparison with other models, and no specific outcomes are related with respect to the data sources. For this reason, the overall risk of bias is unclear or with high concern.

3.2. Observational Studies

Furthermore, 25% of the included papers were clinical research with an observational design and were assessed according to the STROBE guidelines. They reported, on average, a good quality (mean = 86%, IQR = 77–91%), which mainly reflects the RoB analysis. Only one study [66] reported high concern for both the overall Risk of Bias and applicability score. This study [66] combined ML algorithms with data mining in its method. Indeed, the authors used mixed data to predict the increment in the rate of implantation, also computing the classification of sperm quality. The results are also not properly reported in terms of prediction of the determined outcomes and, instead, as a function of the goodness of performance computed for each algorithm. As for the previous group of papers, these studies are also missing specific information about the enrollment and selection of participants, which may lead to an unclear RoB due to participant selection. On the other hand, the methods are well described, so 73% of these studies have a low concern in this regard (Figure 3).
Figure 3. Risk of Bias (RoB) analysis of observational studies performed using PROBAST [45,46,49,52,53,65,66,68,69,70,78].

3.3. Multivariable Prediction Model Studies

The remaining 12% of papers were studies on multivariable prediction models, assessed using the TRIPOD tool. They report an average adherence of 67%. In this small group of studies, no one reported high concern in any of the overall RoB and applicability scores. Since these studies properly set their aim as the diagnostic/prognostic prediction of male infertility, PROBAST assessment is probably more in line with these categories of studies than the others reported above (Figure 4).
Figure 4. Risk of Bias (RoB) analysis of multivariate prognostic and diagnostic studies performed using PROBAST [50,51,62,67,73].

4. Discussion

During the last few years, researchers have been exploring the use of machine learning (ML), artificial neural networks (ANN), and deep learning (DL) techniques to predict male infertility. This review covers a diverse array of studies and presents a comprehensive overview of the published literature in the field of male infertility prediction using machine learning. Advancements in the prediction of male infertility have been witnessed through the integration of machine learning (ML), artificial neural networks (ANN), and deep learning techniques. Noteworthy studies, such as the one proposed by Chen A. [81] based on region-based convolutional neural networks (R-CNNs), exemplify the potential of deep learning architectures in the assessment of sperm morphology.
Some studies [58,62,63,66] are actually comparative analyses, encompassing the utilization of diverse AI techniques, including ANNs, DT, and SVM, for the detection of male fertility. These methodologies demonstrated differing degrees of accuracy, emphasizing the significance of choosing the suitable algorithm for specific tasks. ANNs, in association with the LR, have exhibited promise in the prediction of biochemical parameters associated with male infertility [63]. The adaptability of ANNs permits one to consider intricate relationships within the data, thereby facilitating precise predictions. Furthermore, deep learning techniques have been investigated for the prediction of sperm fertility. Machine learning frameworks, such as the one proposed for the automatic prediction of human semen parameters [60], offer potential applications in assisted reproductive technologies. In addition, the application of machine learning-based spectrophotometry for quantifying sperm concentration presents a rapid and cost-effective technique [52].
Certain studies have the limitation of being vaguely defined [54,58,59,66,70,71], thereby posing challenges in assessing the robustness of predictive models. It is crucial to clearly delineate the limitations of the study for appropriate interpretation of the results.
This is the first review that attempted to collect all the studies addressing the use of machine learning to predict male infertility, including multiple approaches and different data sources to build the model. This meant that even if our intent was ambitious, it encompasses some limitations. Indeed, during the screening of the studies, we found incomplete or insufficient information about datasets or the adopted method, which impeded a punctual analysis of the accuracy of the ML models used in the studies included in the present review.
Machine learning (ML) models, despite their effectiveness, may lack interpretability, thereby making it difficult to comprehend the reasoning behind the predictions. It is imperative to ensure the transparency and interpretability of models to instill trust in their clinical applications. Indeed, in the included studies, we found huge differences in the input information used to build models. As we reported in Table 1, Table 2, Table 3, Table 4, Table 5 and Table 6, the data sources used for building each model were very different, including demographic data, RNA, quality of the embryo, sperm retrieval, hormonal data, and others. This variety, in addition to the fact that, in many studies, multiple variables were used as a combination matrix, limited possible direct comparisons among the models, also leading to different results. This also means that it is not possible to make a precise statement about what variables are strictly correlated to the pathogenesis of male infertility, since each model made a unique combination of the set variables to be included for prediction. Further reviews may face this problem, formulating a specific question that could be addressed with a meta-analysis.
Another limitation of our study is that a mathematical discussion of the ML algorithms used in the reviewed studies is missing. Although an interesting point of view, this aspect was excluded because it was out of the scope of this review and because it would need considerable work, most suitable for a book or an article collection.
Despite the focus of this review being precisely to analyze ANN models, we found that only a few studies used this model for this purpose. However, even if they are a small number, it seems that these studies have a lower RoB, like those of Ma et al. [40] and Girela et al. [60], who obtained low concern regarding the overall applicability of the methods.
Finally, considering that the ANN was associated with a high level of accuracy in the prediction of male infertility, we can hypothesize that it has potential, so its use may increase in the future.

5. Conclusions

Models based on artificial neural networks (ANNs) have demonstrated potential in predicting male fertility, with a few achieving remarkable accuracy. One ANN model achieved a maximum accuracy of 95% in predicting male fertility utilizing explainable AI [77]. Deep learning techniques, a subset of ML, have been employed to evaluate sperm motility and morphology [44,45,68]. Unlike traditional methods, CNNs excel at processing visual data, making them particularly well suited for tasks involving embryo morphology and have also demonstrated effectiveness in assessing spermatozoid motility [44,45,47,53]. Certain investigations might employ a combination of ML, ANN, and DL techniques for a comprehensive analysis, harnessing the strengths of each.
Choosing one model based only on the highest accuracy reported might be an inappropriate decision since it is crucial to consider the nature of the data, the complexity of the prediction task, and the interpretability of the results when selecting the most suitable model. We conclude that the use of ML models for the identification of risk factors and prediction of male infertility has the potential to assist personalized medicine in achieving its goal of creating optimally tailored diagnostic, preventive, and therapeutic measures. This review had the ambitious intent to collect all the studies in which ML models were used to predict male infertility to support those who are approaching the use of machine learning in the medical field for the first time in choosing the most suitable model, based on the accuracy levels already established by previous research. Future studies on male infertility prediction may consider other computational methods, such as the exploitation of the graph contrastive clustering techniques [82] that are based on the use of deep neural networks for representation learning before clustering. This approach, for example, has the advantage of bringing similar or positive sample pairs closer and pushing dissimilar or negative sample pairs further away, thus going beyond the clustering methods of unsupervised machine learning algorithms.
Finally, considering that claiming a technical contribution was out of the scope of this review, we think that further research should address this point by introducing new approaches or databases.

Author Contributions

Conceptualization, M.I., D.D.B. and V.S.A.; methodology, V.S.A. and D.D.B.; software, D.D.B.; validation, D.D.B.; formal analysis, D.D.B.; investigation, M.I., D.D.B. and V.S.A.; resources, M.I. and D.D.B.; data curation, V.S.A. and D.D.B.; writing—original draft preparation, D.D.B.; writing—review and editing, M.I., G.A., D.D.B. and V.S.A.; visualization, D.D.B.; supervision, M.I. and G.A.; project administration, M.I.; funding acquisition, M.I. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Santa Lucia Foundation and the Italian Ministry of Health, name of grant: “NEURO-METAVERSE: Application in Neurorehabilitation and Neuroscience of Metaverse Technologies as Virtual Reality and Artificial Intelligence”.

Institutional Review Board Statement

Not applicable.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

Glossary of all terms used in this paper.
AbbreviationsDescription
AAAArtificial Algae Algorithm
AdaBoostAdaptive Boosting
ANNArtificial Neural Network
ANNsArtificial Neural Networks
BKMRBayesian Kernel Machine Regression
BorutaBoruta is a features selection algorithm
BPNNBackpropagation Neural Networks
CatBoost It is an open-source gradient-boosting library
CNNConvolutional Neural Network
Data miningIt is a model designed to extract the rules from large quantities of data?
DLDeep Learning
DMTLDeep Multi-Task Learning
DNNDeep Neural Network
DPPCGDeep Learning Computational Modeling Alternative
DTDecision Tree
DTLADesign Thinking Learning Approach
FSNNFeedback System Neural Network
GBT Gradient Boosting Technique
GFLASSOLeast Absolute Shrinkage and Selection Operator
GFLASSOGraph-Guided Fused Least Absolute Shrinkage and Selection Operator
GLMGeneralized Linear Model
GPGaussian Process
INKSimple instance-based learner that uses the class of the nearest k training instances for the class of the test instances
J48Also known as the c4.5 algorithm, it examines the data categorically and continuously
KNNK-Nearest Neighbors
KstarInstance-Based Classifier Algorithm
LDALinear Discriminant Analysis
LDFALinear Discriminant Function Analysis
LRLinear Regression
MLMachine Learning
MLPMulti-Layer Perceptron
MvLRMMultivariate Logistic Regression Model
NBNaïve Bayes
NNETNon-Linear Method of Neural Computation

Appendix A

In this section specifications, about search strategy are reported.
Table A1. Mesh, Tiab, and synonyms were used for the literature search.
Table A2. String used for literature search in each database.

References

  1. Gurunath, S.; Pandian, Z.; Anderson, R.A.; Bhattacharya, S. Defining infertility—A systematic review of prevalence studies. Hum. Reprod. Update 2011, 17, 575–588. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  2. Boivin, J.; Bunting, L.; Collins, J.A.; Nygren, K.G. International estimates of infertility prevalence and treatment-seeking: Potential need and demand for infertility medical care. Hum. Reprod. 2007, 22, 1506–1512. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  3. Xie, Y.; Ren, Y.; Niu, C.; Zheng, Y.; Yu, P.; Li, L. The impact of stigma on mental health and quality of life of infertile women: A systematic review. Front. Psychol. 2023, 13, 1093459. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  4. Sharlip, I.D.; Jarow, J.P.; Belker, A.M.; Lipshultz, L.I.; Sigman, M.; Thomas, A.J.; Schlegel, P.N.; Howards, S.S.; Nehra, A.; Damewood, M.D.; et al. Best practice policies for male infertility. Fertil. Steril. 2002, 77, 873–882. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  5. Babakhanzadeh, E.; Nazari, M.; Ghasemifar, S.; Khodadadian, A. Some of the factors involved in male infertility: A prospective review. Int. J. Gen. Med. 2020, ume 13, 29–41. [Google Scholar] [CrossRef] [Scilit]
  6. Dolan, A.; Lomas, T.; Ghobara, T.; Hartshorne, G. ‘It’s like taking a bit of masculinity away from you’: Towards a Theoretical Understanding of Men’s Experiences of Infertility. Sociol. Health Illn. 2017, 39, 878–892. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  7. Kooli, R.; Sallem, A.; Chebil, D.; Boussabbeh, M.; Ben Mohamed, B.; Ajina, T.; Boughzela, I.; Mougou, S.; Mehdi, M. Factors associated with anxiety and depression in men undergoing fertility investigations: A cross-sectional study. BMC Psychol. 2023, 11, 299. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  8. Cousineau, T.M.; Domar, A.D. Psychological impact of infertility. Best Pract. Res. Clin. Obstet. Gynaecol. 2007, 21, 293–308. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  9. Skakkebæk, N.E.; Lindahl-Jacobsen, R.; Levine, H.; Andersson, A.M.; Jørgensen, N.; Main, K.M.; Lidegaard, Ø.; Priskorn, L.; Holmboe, S.A.; Bräuner, E.V.; et al. Environmental factors in declining human fertility. Nat. Rev. Endocrinol. 2022, 18, 139–157. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  10. Azhar, M.; Altaf, S.; Uddin, I.; Cheng, J.; Wu, L.; Tong, X.; Qin, W.; Bao, J. Towards post-meiotic sperm production: Genetic insight into human infertility from mouse models. Int. J. Biol. Sci. 2021, 17, 2487. [Google Scholar] [CrossRef] [Scilit]
  11. Yahaya, T.O.; Oladele, E.O.; Anyebe, D.; Obi, C.; Bunza, M.D.A.; Sulaiman, R.; Liman, U.U. Chromosomal abnormalities predisposing to infertility, testing, and management: A narrative review. Bull. Natl. Res. Cent. 2021, 45, 65. [Google Scholar] [CrossRef] [Scilit]
  12. O’Connor, J.C.; Chapin, R.E. Critical evaluation of observed adverse effects of endocrine active substances on reproduction and development, the immune system, and the nervous system. Pure Appl. Chem. 2003, 75, 2099–2123. [Google Scholar] [CrossRef] [Scilit]
  13. Ediningtyas, A.N.; Laqif, A.; Melinawati, E.; Respati, S.H. Relationships between Smoking, Age, Body Mass Index, and Abstinence with DNA Fragmentation Index in Male Infertile Patients. J. Matern. Child Health 2022, 7, 398–407. [Google Scholar] [CrossRef] [Scilit]
  14. Kumar, N.; Singh, A.K. Impact of environmental factors on human semen quality and male fertility: A narrative review. Environ. Sci. Eur. 2022, 34, 6. [Google Scholar] [CrossRef] [Scilit]
  15. Leisegang, K.; Sengupta, P.; Agarwal, A.; Henkel, R. Obesity and male infertility: Mechanisms and management. Andrologia 2021, 53, e13617. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  16. Skoracka, K.; Eder, P.; Łykowska-Szuber, L.; Dobrowolska, A.; Krela-Kaźmierczak, I. Diet and nutritional factors in male (in) fertility—Underestimated factors. J. Clin. Med. 2020, 9, 1400. [Google Scholar] [CrossRef] [Scilit]
  17. Bonde, J.P.E.; Ernst, E.; Jensen, T.K.; Hjollund, N.H.; Kolstad, H.; Henriksen, T.B.; Scheike, T.; Giwercman, A.; Olsen, J.; Skakkebæk, N.E. Relation between semen quality and fertility: A population-based study of 430 first-pregnancy planners. Lancet 1998, 352, 1172–1177. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  18. Patel, A.S.; Leong, J.Y.; Ramasamy, R. Prediction of male infertility by the World Health Organization laboratory manual for assessment of semen analysis: A systematic review. Arab. J. Urol. 2018, 16, 96–102. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  19. Zhang, L.; Zhang, L. Artificial intelligence for remote sensing data analysis: A review of challenges and opportunities. IEEE Geosci. Remote Sens. Mag. 2022, 10, 270–294. [Google Scholar] [CrossRef] [Scilit]
  20. Feng, J.; Phillips, R.V.; Malenica, I.; Bishara, A.; Hubbard, A.E.; Celi, L.A.; Pirracchio, R. Clinical artificial intelligence quality improvement: Towards continual monitoring and updating of AI algorithms in healthcare. npj Digit. Med. 2022, 5, 66. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  21. Dalal, S.; Goel, P.; Onyema, E.M.; Alharbi, A.; Mahmoud, A.; Algarni, M.A.; Awal, H. Application of machine learning for cardiovascular disease risk prediction. Comput. Intell. Neurosci. 2023, 2023, 9418666. [Google Scholar] [CrossRef] [Scilit]
  22. Liu, C.J.; Zhang, L.; Sun, Y.; Geng, L.; Wang, R.; Shi, K.M.; Wan, J.X. Application of CT and MRI images based on an artificial intelligence algorithm for predicting lymph node metastasis in breast cancer patients: A meta-analysis. BMC Cancer 2023, 23, 1134. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  23. Chou, C.Y.; Hsu, D.Y.; Chou, C.H. Predicting the onset of diabetes with machine learning methods. J. Pers. Med. 2023, 13, 406. [Google Scholar] [CrossRef] [Scilit]
  24. Diaz, P.; Dullea, A.; Chu, K.Y.; Zizzo, J.; Loloi, J.; Reddy, R.; Campbell, K.; Li, P.S.; Ramasamy, R. Future of male infertility evaluation and treatment: Brief review of emerging technology. Urology 2022, 169, 9–16. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  25. You, J.B.; McCallum, C.; Wang, Y.; Riordon, J.; Nosrati, R.; Sinton, D. Machine learning for sperm selection. Nat. Rev. Urol. 2021, 18, 387–403. [Google Scholar] [CrossRef] [Scilit]
  26. Krenz, H.; Sansone, A.; Fujarski, M.; Krallmann, C.; Zitzmann, M.; Dugas, M.; Kliesch, S.; Varghese, J.; Tüttelmann, F.; Gromoll, J. Machine learning based prediction models in male reproductive health: Development of a proof-of-concept model for Klinefelter Syndrome in azoospermic patients. Andrology 2022, 10, 534–544. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  27. GhoshRoy, D.; Alvi, P.A.; Santosh, K.C. AI Tools for Assessing Human Fertility Using Risk Factors: A State-of-the-Art Review. J. Med. Syst. 2023, 47, 91. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  28. Shahid, N.; Rappon, T.; Berta, W. Applications of artificial neural networks in health care organizational decision-making: A scoping review. PLoS ONE 2019, 14, e0212356. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  29. Naseem, S.; Mahmood, T.; Saba, T.; Alamri, F.S.; Bahaj, S.A.; Ateeq, H.; Farooq, U. DeepFert: An Intelligent Fertility Rate Prediction Approach for Men Based on Deep Learning Neural Networks. IEEE Access 2023, 11, 75006–75022. [Google Scholar] [CrossRef] [Scilit]
  30. Miller, S.A.; Forrest, J.L. Enhancing your practice through evidence-based decision making: PICO, learning how to ask good questions. J. Evid. Based Dent. Pract. 2001, 1, 136–141. [Google Scholar] [CrossRef] [Scilit]
  31. Moher, D.; Liberati, A.; Tetzlaff, J.; Altman, D.G.; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Ann. Intern. Med. 2009, 151, 264–269. [Google Scholar] [CrossRef] [PubMed]
  32. Munn, Z.; Moola, S.; Riitano, D.; Lisy, K. The development of a critical appraisal tool for use in systematic reviews addressing questions of prevalence. Int. J. Health Policy Manag. 2014, 3, 123. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  33. Altman, D.G.; Simera, I.; Hoey, J.; Moher, D.; Schulz, K. EQUATOR: Reporting guidelines for health research. Lancet 2008, 371, 1149–1150. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  34. Moons, K.G.M.; Altman, D.G.; Reitsma, J.B.; Ioannidis, J.P.A.; Macaskill, P.; Steyerberg, E.W.; Vickers, A.J.; Ransohoff, D.F.; Collins, G.S. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): Explanation and elaboration. Ann. Intern. Med. 2015, 162, W1–W73. [Google Scholar] [CrossRef] [Scilit]
  35. Liu, C.; Qi, Y.; Liu, X.; Chen, M.; Xiong, Y.; Huang, S.; Zou, K.; Tan, J.; Sun, X. The reporting of prognostic prediction models for obstetric care was poor: A cross-sectional survey of 10-year publications. BMC Med. Res. Methodol. 2023, 23, 9. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  36. Wolff, R.F.; Moons, K.G.; Riley, R.; Whiting, P.F.; Westwood, M.; Collins, G.S.; Reitsma, J.B.; Kleijnen, J.; Mallett, S.; PROBAST Group. PROBAST: A tool to assess the risk of bias and applicability of prediction model studies. Ann. Intern. Med. 2019, 170, 51–58. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  37. Bachelot, G.; Dhombres, F.; Sermondade, N.; Hamid, R.H.; Berthaut, I.; Frydman, V.; Prades, M.; Kolanska, K.; Selleret, L.; Mathieu-D’argent, E.; et al. A Machine Learning Approach for the Prediction of Testicular Sperm Extraction in Nonobstructive Azoospermia: Algorithm Development and Validation Study. J. Med. Internet Res. 2023, 25, e44047. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  38. Zeadna, A.; Khateeb, N.; Rokach, L.; Lior, Y.; Har-Vardi, I.; Harlev, A.; Huleihel, M.; Lunenfeld, E.; Levitas, E. Prediction of sperm extraction in non-obstructive azoospermia patients: A machine-learning perspective. Hum. Reprod. 2020, 35, 1505–1514. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  39. Ramasamy, R.; Padilla, W.O.; Osterberg, E.C.; Srivastava, A.; Reifsnyder, J.E.; Niederberger, C.; Schlegel, P.N. A comparison of models for predicting sperm retrieval before microdissection testicular sperm extraction in men with nonobstructive azoospermia. J. Urol. 2013, 189, 638–642. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  40. Ma, Y.; Chen, B.; Wang, H.; Hu, K.; Huang, Y. Prediction of sperm retrieval in men with non-obstructive azoospermia using artificial neural networks: Leptin is a good assistant diagnostic marker. Hum. Reprod. 2011, 26, 294–298. [Google Scholar] [CrossRef] [Scilit]
  41. Guo, Y.; Li, J.; Hao, F.; Yang, Y.; Yang, H.; Chang, Q.; Kong, P.; Liu, W.; Jiao, X.; Teng, X. A new perspective on semen quality of aged male: The characteristics of metabolomics and proteomics. Front. Endocrinol. 2023, 13, 1058250. [Google Scholar] [CrossRef] [Scilit]
  42. Yuzkat, M.; Ilhan, H.O.; Aydin, N. Detection of sperm cells by single-stage and two-stage deep object detectors. Biomed. Signal Process. Control 2023, 83, 104630. [Google Scholar] [CrossRef] [Scilit]
  43. Huang, H.H.; Hsieh, S.J.; Chen, M.S.; Jhou, M.J.; Liu, T.C.; Shen, H.L.; Yang, C.T.; Hung, C.C.; Yu, Y.Y.; Lu, C.J. Machine Learning Predictive Models for Evaluating Risk Factors Affecting Sperm Count: Predictions Based on Health Screening Indicators. J. Clin. Med. 2023, 12, 1220. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  44. Lee, R.; Witherspoon, L.; Robinson, M.; Lee, J.H.; Duffy, S.P.; Flannigan, R.; Ma, H. Automated rare sperm identification from low-magnification microscopy images of dissociated microsurgical testicular sperm extraction samples using deep learning. Fertil. Steril. 2022, 118, 90–99. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  45. Fan, X.; Xu, Z.; Zhang, H.; Liu, D.; Yang, Q.; Tao, Q.; Wen, M.; Kang, X.; Zhang, Z.; Lu, H. Fully automatic resolution of untargeted GC-MS data with deep learning assistance. Talanta 2022, 244, 123415. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  46. Miahi, E.; Mirroshandel, S.A.; Nasr, A. Genetic Neural Architecture Search for automatic assessment of human sperm images. Expert Syst. Appl. 2022, 188, 115937. [Google Scholar] [CrossRef] [Scilit]
  47. Otti, S.; Amiriparian, S.; Gerczuk, M.; Schuller, B.W. motilitAI: A machine learning framework for automatic prediction of human sperm motility. Iscience 2022, 25, 104644. [Google Scholar]
  48. Ilhan, H.O.; Serbes, G. Sperm morphology analysis by using the fusion of two-stage fine-tuned deep networks. Biomed. Signal Process. Control 2022, 71, 103246. [Google Scholar] [CrossRef] [Scilit]
  49. Abbasi, A.; Miahi, E.; Mirroshandel, S.A. Effect of deep transfer and multi-task learning on sperm abnormality detection. Comput. Biol. Med. 2021, 128, 104121. [Google Scholar] [CrossRef] [Scilit]
  50. Yüzkat, M.; Ilhan, H.O.; Aydin, N. Multi-model CNN fusion for sperm morphology analysis. Comput. Biol. Med. 2021, 137, 104790. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  51. Yibre, A.M.; Koçer, B. Semen quality predictive model using feed forwarded neural network trained by learning-based artificial algae algorithm. Eng. Sci. Technol. Int. J. 2021, 24, 310–318. [Google Scholar] [CrossRef] [Scilit]
  52. Lesani, A.; Kazemnejad, S.; Zand, M.M.; Azadi, M.; Jafari, H.; Mofrad, M.R.; Nosrati, R. Quantification of human sperm concentration using machine learning-based spectrophotometry. Comput. Biol. Med. 2020, 127, 104061. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  53. Javadi, S.; Mirroshandel, S.A. A novel deep learning method for automatic assessment of human sperm images. Comput. Biol. Med. 2019, 109, 182–194. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  54. Movahed, R.A.; Mohammadi, E.; Orooji, M. Automatic segmentation of Sperm’s parts in microscopic images of human semen smears using concatenated learning approaches. Comput. Biol. Med. 2019, 109, 242–253. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  55. Vickram, A.S.; Raja, D.; Srinivas, M.S.; Rao, K.A.; Jayaraman, G.; Sridharan, T.B. Prediction of Zn concentration in human seminal plasma of Normospermia samples by Artificial Neural Networks (ANN). J. Assist. Reprod. Genet. 2013, 30, 453–459. [Google Scholar] [PubMed]
  56. Steigerwald, P.; Krause, W. Estimation of sperm morphology using a new GASA system. Andrologia 1998, 30, 23–27. [Google Scholar] [CrossRef] [Scilit]
  57. Vickram, A.S.; Kamini, A.R.; Das, R.; Pathy, M.R.; Parameswari, R.; Archana, K.; Sridharan, T.B. Validation of artificial neural network models for predicting biochemical markers associated with male infertility. Syst. Biol. Reprod. Med. 2016, 62, 258–265. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  58. GhoshRoy, D.; Alvi, P.A.; Santosh, K.C. Unboxing Industry-Standard AI Models for Male Fertility Prediction with SHAP. Healthcare 2023, 11, 929. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  59. Ito, Y.; Unagami, M.; Yamabe, F.; Mitsui, Y.; Nakajima, K.; Nagao, K.; Kobayashi, H. A method for utilizing automated machine learning for histopathological classification of testis based on Johnsen scores. Sci. Rep. 2021, 11, 9962. [Google Scholar] [CrossRef] [Scilit]
  60. Girela, J.L.; Gil, D.; Johnsson, M.; Gomez-Torres, M.J.; De Juan, J. Semen parameters can be predicted from environmental factors and lifestyle using artificial intelligence methods. Biol. Reprod. 2013, 88, 99. [Google Scholar] [CrossRef] [Scilit]
  61. Gil, D.; Girela, J.L.; De Juan, J.; Gomez-Torres, M.J.; Johnsson, M. Predicting seminal quality with artificial intelligence methods. Expert Syst. Appl. 2012, 39, 12564–12573. [Google Scholar] [CrossRef] [Scilit]
  62. Zhou, R.; Liang, J.; Chen, Q.; Tian, H.; Yang, C.; Liu, C. A 3-Gene Random Forest Model to Diagnose Non-obstructive Azoospermia Based on Transcription Factor-Related Henes. Reprod. Sci. 2023, 30, 233–246. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  63. Peng, F.; Muhuitijiang, B.; Zhou, J.; Liang, H.; Zhang, Y.; Zhou, R. An artificial neural network model to diagnose non-obstructive azoospermia based on RNA-binding protein-related genes. Aging 2023, 15, 3120. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  64. Samli, M.M.; Dogan, I. An artificial neural network for predicting the presence of spermatozoa in the testes of men with nonobstructive azoospermia. J. Urol. 2004, 171, 2354–2357. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  65. He, J.; Zhao, Y.; Zhou, Z.; Zhang, M. Machine learning and integrative analysis identify the common pathogenesis of azoospermia complicated with COVID-19. Front. Immunol. 2023, 14, 1114870. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  66. Mirroshandel, S.A.; Ghasemian, F.; Monji-Azad, S. Applying data mining techniques for increasing implantation rate by selecting best sperms for intra-cytoplasmic sperm injection treatment. Comput. Methods Programs Biomed. 2016, 137, 215–229. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  67. Wald, M.; Sparks, A.E.; Sandlow, J.; Van-Voorhis, B.; Syrop, C.H.; Niederberger, C.S. Computational models for prediction of IVF/ICSI outcomes with surgically retrieved spermatozoa. Reprod. Biomed. Online 2005, 11, 325–331. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  68. Guo, Y.; Chen, Q.; Zhan, Y.; Zhou, W.; Zhang, H.; Zhou, N.; Yang, H.; Chen, Y.; Wang, Y.; Sun, L.; et al. Semen damage contributed over 50% to air-pollutant-induced infertility: A prospective cohort study of 3940 men in China. Sci. Total Environ. 2023, 885, 163532. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  69. Zhao, M.; Yu, W.X.; Liu, S.J.; Deng, Y.J.; Zhao, Z.W.; Guo, J.; Gao, Q.H. Identification and immuno-infiltration analysis of cuproptosis regulators in human spermatogenic dysfunction. Front. Genet. 2023, 14, 1115669. [Google Scholar] [CrossRef] [Scilit]
  70. Tang, Q.; Su, Q.; Wei, L.; Wang, K.; Jiang, T. Identifying potential biomarkers for non-obstructive azoospermia using WGCNA and machine learning algorithms. Front. Endocrinol. 2023, 14, 1108616. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  71. Ory, J.; Tradewell, M.B.; Blankstein, U.; Lima, T.F.; Nackeeran, S.; Gonzalez, D.C.; Nwefo, E.; Moryousef, J.; Madhusoodanan, V.; Lau, S.; et al. Artificial intelligence based machine learning models predict sperm parameter upgrading after varicocele repair: A multi-institutional analysis. World J. Men’s Health 2022, 40, 618. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  72. Gunderson, S.J.; Molina, L.C.P.; Spies, N.; Balestrini, P.A.; Buffone, M.G.; Jungheim, E.S.; Riley, J.; Santi, C.M. Machine-learning algorithm incorporating capacitated sperm intracellular pH predicts conventional in vitro fertilization success in normospermic patients. Fertil. Steril. 2021, 115, 930–939. [Google Scholar] [CrossRef] [Scilit]
  73. Xu, F.; Guo, G.; Zhu, F.; Tan, X.; Fan, L. Protein deep profile and model predictions for identifying the causal genes of male infertility based on deep learning. Inf. Fusion 2021, 75, 70–89. [Google Scholar] [CrossRef] [Scilit]
  74. Wang, Y.; Liu, K.; Han, Q.; Yang, H.; Zhou, N.; Sun, L.; Zou, P.; Ling, X.; Ao, L.; Cui, Z.; et al. An exposomic approach with 138 chemical and non-chemical exposures to predict 32 biomarkers of male reproductive damages: A case study of college students in Chongqing, China. Sci. Total Environ. 2021, 767, 144380. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  75. Karthikeyan, V.; Vickram, A.S.; Manian, R. Estimation of varicocele associated human ARG2 and NOS1 proteins and computational analysis on the effect of its nsSNPs. Int. J. Biol. Macromol. 2020, 164, 735–747. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  76. Hicks, S.A.; Andersen, J.M.; Witczak, O.; Thambawita, V.; Halvorsen, P.; Hammer, H.L.; Haugen, T.B.; Riegler, M.A. Machine learning-based analysis of sperm videos and participant data for male fertility prediction. Sci. Rep. 2019, 9, 16770. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  77. Akinsal, E.C.; Haznedar, B.; Baydilli, N.; Kalinli, A.; Ozturk, A. Artificial neural network for the prediction of chromosomal abnormalities in azoospermic males. Urol. J. 2018, 15, 122. [Google Scholar] [PubMed]
  78. Ho, N.R.Y.; Huang, N.; Conrad, D.F. Improved detection of disease-associated variation by sex-specific characterization and prediction of genes required for fertility. Andrology 2015, 3, 1140–1149. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  79. Powell, C.R.; Desai, R.A.; Makhlouf, A.A.; Sigman, M.; Jarow, J.P.; Ross, L.S.; Niederberger, C.S. Computational models for detection of endocrinopathy in subfertile males. Int. J. Impot. Res. 2008, 20, 79–84. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  80. Okonofua, F.E.; Ntoimo, L.F.C.; Omonkhua, A.; Ayodeji, O.; Olafusi, C.; Unuabonah, E.; Ohenhen, V. Causes and risk factors for Male infertility: A scoping review of published studies. Int. J. Gen. Med. 2022, ume 15, 5985–5997. [Google Scholar] [CrossRef] [Scilit]
  81. Chen, A.; Li, C.; Rahaman, M.; Yao, Y.; Chen, H.; Yang, H.; Zhao, P.; Hu, W.; Liu, W.; Zou, S.; et al. A Comprehensive Comparative Study of Deep Learning Methods for Noisy Sperm Image Classification: From Convolutional Neural Network to Visual Transformer. Intell. Med. 2023, in press. [Google Scholar] [CrossRef] [Scilit]
  82. Li, G.; Zhao, B.; Su, X.; Yang, Y.; Hu, P.; Zhou, X.; Hu, L. Discovering consensus regions for interpretable identification of rna n6-methyladenosine modification sites via graph contrastive clustering. IEEE J. Biomed. Health Inform. 2024, 1–10. [Google Scholar] [CrossRef] [Scilit] [PubMed]
Disclaimer/Publisher’s Note: The statements, opinions, and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions, or products referred to in the content.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
The Effectiveness of the Chronic Disease Self-Management Program in Improving Patients’ Self-Efficacy and Health-Related Behaviors: A Quasi-Experimental Study
Previous
“I’m Torn”: Qualitative Analysis of Dental Practitioner-Perceived Barriers, Facilitators, and Solutions to HPV Vaccine Promotion
Next