A clinical decision support system for prediction of pregnancy outcome in pregnant women with systemic lupus erythematosus

Machine learning techniques for prediction in pregnancy complicated by autoimmune rheumatic diseases: Applications and challenges

Autoimmune rheumatic diseases are chronic conditions affecting multiple systems and often occurring in young women of childbearing age. The diseases and the physiological characteristics of pregnancy significantly impact maternal-fetal health and pregnancy outcomes. Currently, the integration of big data with healthcare has led to the increasing popularity of using machine learning (ML) to mine clinical data for studying pregnancy complications. In this review, we introduce the basics of ML and the recent advances and trends of ML in different prediction applications for common pregnancy complications by autoimmune rheumatic diseases. Finally, the challenges and future for enhancing the accuracy, reliability, and clinical applicability of ML in prediction have been discussed. This review will provide insights into the utilization of ML in identifying and assisting clinical decision-making for pregnancy complications, while also establishing a foundation for exploring comprehensive management strategies for pregnancy and enhancing maternal and child health.

Quantitative prediction of postpartum hemorrhage in cesarean section on machine learning

BACKGROUND

Cesarean section-induced postpartum hemorrhage (PPH) potentially causes anemia and hypovolemic shock in pregnant women. Hence, it is helpful for obstetricians and anesthesiologists to prepare pre-emptive prevention when predicting PPH occurrence in advance. However, current works on PPH prediction focus on whether PPH occurs rather than assessing PPH amount. To this end, this work studies quantitative PPH prediction with machine learning (ML).

METHODS

The study cohort in this paper was selected from individuals with PPH who were hospitalized at Shijiazhuang Obstetrics and Gynecology Hospital from 2020 to 2022. In this study cohort, we built a dataset with 6,144 subjects covering clinical parameters, anesthesia operation records, laboratory examination results, and other information in the electronic medical record system. Based on our built dataset, we exploit six different ML models, including logistic regression, linear regression, gradient boosting, XGBoost, multilayer perceptron, and random forest, to automatically predict the amount of bleeding during cesarean section. Eighty percent of the dataset was used as model training, and 20 % was used for verification. Those ML models are constantly verified and improved by root mean squared error(RMSE) and mean absolute error(MAE). Moreover, we also leverage the importance of permutation and partial dependence plot (PDP) to discuss their feasibility.

RESULT

The experiment results show that random forest obtains the highest accuracy for PPH amount prediction compared to other ML methods. Random forest reaches the mean absolute error of 21.7, less than 5.4 % prediction error. It also gains the root mean squared error of 33.75, less than 9.3 % prediction error. On the other hand, the experimental results also disclose indicators that contributed most to PPH prediction, including Ca, hemoglobin, white blood cells, platelets, Na, and K.

CONCLUSION

It effectively predicts the amount of PPH during a cesarean section by ML methods, especially random forest. With the above insight, ML predicting PPH amounts provides early warning for clinicians, thus reducing complications and improving cesarean sections' safety. Furthermore, the importance of ML and permutation, complemented by incorporating PDP, promises to provide clinicians with a transparent indication of individual risk prediction.

Machine learning-based clinical decision support systems for pregnancy care: A systematic review

BACKGROUND

Clinical decision support systems (CDSSs) can provide various functions and advantages to healthcare delivery. Quality healthcare during pregnancy and childbirth is of vital importance, and machine learning-based CDSSs have shown positive impact on pregnancy care.

OBJECTIVE

This paper aims to investigate what has been done in CDSSs in the context of pregnancy care using machine learning, and what aspects require attention from future researchers.

METHODS

We conducted a systematic review of existing literature following a structured process of literature search, paper selection and filtering, and data extraction and synthesis.

RESULTS

17 research papers were identified on the topic of CDSS development for different aspects of pregnancy care using various machine learning algorithms. We discovered an overall lack of explainability in the proposed models. We also observed a lack of experimentation, external validation and discussion around culture, ethnicity and race from the source data, with most studies using data from a single centre or country, and an overall lack of awareness of applicability and generalisability of the CDSSs regarding different populations. Finally, we found a gap between machine learning practices and CDSS implementation, and an overall lack of user testing.

CONCLUSION

Machine learning-based CDSSs are still under-explored in the context of pregnancy care. Despite the open problems that remain, the few studies that tested a CDSS for pregnancy care reported positive effects, reinforcing the potential of such systems to improve clinical practice. We encourage future researchers to take into consideration the aspects we identified in order for their work to translate into clinical use.

Studying Pregnancy Outcome Risk in Patients with Systemic Lupus Erythematosus Based on Cluster Analysis

Background. Pregnancy in systemic lupus erythematosus (SLE) patients is a challenge due to the potential maternal and fetal complications. Therefore, a multidisciplinary assessment of disease risk before and during pregnancy is essential to improve pregnancy outcomes. Objectives. Our purpose was to (i) define clusters of patients with similar history and laboratory features and determine the associative maternal and perinatal outcomes and (ii) evaluate the risk spectrum of maternal and perinatal outcomes of pregnancy in SLE patients, represented by our established risk-assessment chart. Methods. Medical records of 119 patients in China were analyzed retrospectively. Significant variables with $p<0.05$ were selected. The self-organizing map was used for clustering the data based on historical background and laboratory features. Results. Clustering was conducted using 21 maternal and perinatal features. Five clusters were recognized, and their prominent maternal manifestations were as follows: cluster 1 (including 27.73% of all patients): preeclampsia and lupus nephritis; cluster 2 (22.69%): oligohydramnios, uterus scar, and femoral head necrosis; cluster 3 (13.45%): upper respiratory tract infection; cluster 4 (15.97%): premature membrane rupture; and cluster 5 (20.17%): no problem. Conclusion. Pregnancy outcomes in SLE women fell into three categories, namely high risk, moderate risk, and low risk. Present manifestations, besides the medical records, are a potential assessment means for better management of pregnant SLE patients.

Early Prediction of Adverse Pregnancy Outcome in Women with Systemic Lupus Erythematosus, Antiphospholipid Syndrome, or Non-Criteria Obstetric Antiphospholipid Syndrome

A prospectively study of pregnant women with systemic lupus erythematosus (SLE), antiphospholipid syndrome, or non-criteria obstetric antiphospholipid syndrome was conducted to describe the characteristics of women followed in a referral unit and to derive a predictive tool for adverse pregnancy outcome (APO). Demographic characteristics, treatments, SLE activity, and flares were recorded. Laboratory data included a complete blood cell count, protein-to-creatinine urinary ratio (Pr/Cr ratio), complement, anti dsDNA, anti-SSA/Ro, anti-SSB/La, and antiphospholipid antibodies status. A stepwise regression was used to identify baseline characteristics available before pregnancy and during the 1st trimester that were most predictive of APO and to create the predictive model. A total of 217 pregnancies were included. One or more APO occurred in 45 (20.7%) women. A baseline model including non-Caucasian ethnicity (OR 2.78; 95% CI [1.16-6.62]), smoking (OR 4.43; 95% CI [1.74-11.29]), pregestational hypertension (OR 16.13; 95% CI [4.06-64.02]), and pregestational corticosteroids treatment OR 2.98; 95% CI [1.30-6.87]) yielded an AUC of 0.78 (95% CI, [0.70-0.86]). Among first-trimester parameters, only Pr/Cr ratio improved the model fit, but the predictive performance was not significantly improved (AUC of 0.78 vs. 0.81; p = 0.16). Better biomarkers need to be developed to efficiently stratify pregnant women with the most common autoimmune diseases.

Prediction models and associated factors on the fertility behaviors of the floating population in China

The floating population has been growing rapidly in China, and their fertility behaviors do affect urban management and development. Based on the data set of the China Migrants Dynamic Survey in 2016, the logistic regression model and multiple linear regression model were used to explore the related factors of fertility behaviors among the floating populace. The artificial neural network model, the naive Bayes model, and the logistic regression model were used for prediction. The findings showed that age, gender, ethnic, household registration, education level, occupation, duration of residence, scope of migration, housing, economic conditions, and health services all affected the reproductive behavior of the floating population. Among them, the improvement duration of post-migration residence and family economic conditions positively impacted their fertility behavior. Non-agricultural new industry workers with college degrees or above living in first-tier cities were less likely to have children and more likely to delay childbearing. Among the prediction models, both the artificial neural network model and logistic regression model had better prediction effects. Improving the employment and income of new industry workers, and introducing preferential housing policies might improve their probability of bearing children. The artificial neural network and logistic regression model could predict individual fertility behavior and provide a scientific basis for the urban population management.

Risk assessment score for adverse pregnancy outcome in systemic lupus erythematosus patients

Background

Systemic lupus erythematosus (SLE) is a chronic autoimmune multisystem dihe criteria for SLE according to thsease that mainly affects females of childbearing age. SLE still possesses risks during pregnancy that lead to poor maternal and fetal outcomes. The objectives of the study were to identify factors associated with unfavorable pregnancy outcomes and develop a predictive risk score for adverse pregnancy outcomes in patients with SLE.

Results

The main predictive factors associated with adverse pregnancy outcomes among lupus patients in multiple linear regression were an absence of remission for at least 6 months before conception, preexisting lupus nephritis, active disease at conception, C3 hypocomplementemia, and antiphospholipid antibody syndrome. Each predictor is assigned a weighted point score, and the sum of points represents the risk score. The area under the receiver operating characteristic curve (ROC) was 0.948 (95% confidence interval, 0.908–0.988), suggesting that the score had strong discriminatory power for adverse pregnancy outcomes.

Conclusions

In this study, a predictive model with a risk score classification for adverse pregnancy outcomes in SLE patients was developed. This could help rheumatologists identify high-risk pregnant patients for better disease monitoring and management, resulting in better maternal/fetal outcomes.

Predicting Parkinson’s Disease Progression: Evaluation of Ensemble Methods in Machine Learning

Parkinson’s disease (PD) is a complex neurodegenerative disease. Accurate diagnosis of this disease in the early stages is crucial for its initial treatment. This paper aims to present a comparative study on the methods developed by machine learning techniques in PD diagnosis. We rely on clustering and prediction learning approaches to perform the comparative study. Specifically, we use different clustering techniques for PD data clustering and support vector regression ensembles to predict Motor-UPDRS and Total-UPDRS. The results are then compared with the other prediction learning approaches, multiple linear regression, neurofuzzy, and support vector regression techniques. The comparative study is performed on a real-world PD dataset. The prediction results of data analysis on a PD real-world dataset revealed that expectation-maximization with the aid of SVR ensembles can provide better prediction accuracy in relation to decision trees, deep belief network, neurofuzzy, and support vector regression combined with other clustering techniques in the prediction of Motor-UPDRS and Total-UPDRS.

Using Machine Learning to Predict Complications in Pregnancy: A Systematic Review

Introduction: Artificial intelligence is widely used in medical field, and machine learning has been increasingly used in health care, prediction, and diagnosis and as a method of determining priority. Machine learning methods have been features of several tools in the fields of obstetrics and childcare. This present review aims to summarize the machine learning techniques to predict perinatal complications.

Objective: To identify the applicability and performance of machine learning methods used to identify pregnancy complications.

Methods: A total of 98 articles were obtained with the keywords “machine learning,” “deep learning,” “artificial intelligence,” and accordingly as they related to perinatal complications (“complications in pregnancy,” “pregnancy complications”) from three scientific databases: PubMed, Scopus, and Web of Science. These were managed on the Mendeley platform and classified using the PRISMA method.

Results: A total of 31 articles were selected after elimination according to inclusion and exclusion criteria. The features used to predict perinatal complications were primarily electronic medical records (48%), medical images (29%), and biological markers (19%), while 4% were based on other types of features, such as sensors and fetal heart rate. The main perinatal complications considered in the application of machine learning thus far are pre-eclampsia and prematurity. In the 31 studies, a total of sixteen complications were predicted. The main precision metric used is the AUC. The machine learning methods with the best results were the prediction of prematurity from medical images using the support vector machine technique, with an accuracy of 95.7%, and the prediction of neonatal mortality with the XGBoost technique, with 99.7% accuracy.

Conclusion: It is important to continue promoting this area of research and promote solutions with multicenter clinical applicability through machine learning to reduce perinatal complications. This systematic review contributes significantly to the specialized literature on artificial intelligence and women’s health.

Application of an artificial neural network for predicting the potential chemotherapy benefit of patients with gastric cancer after radical surgery

BACKGROUND

Artificial neural network models have a strong self-learning ability and can deal with complex biological information, but there is no artificial neural network model for predicting the benefits of adjuvant chemotherapy in patients with gastric cancer.

METHODS

The clinicopathological data of patients who underwent radical resection of gastric cancer from January 2010 to September 2014 were analyzed retrospectively. Patients who underwent surgery combined with adjuvant chemotherapy were randomly divided into a training cohort (70%) and a validation cohort (30%). An artificial neural network model (potential-CT-benefit-ANN) was established, and its ability to predict the potential benefit of chemotherapy was evaluated by the C-index. The prognostic prediction and stratification ability of potential-CT-benefit-ANN and the eighth American Joint Committee on Cancer staging system were compared by receiver operating characteristic curves and Kaplan-Meier curves.

RESULTS

In both the training and validation cohort, potential-CT-benefit-ANN shows good prediction accuracy for potential adjuvant chemotherapy benefit. The receiver operating characteristic curve showed that the prediction accuracy of potential-CT-benefit-ANN was better than that of the eighth American Joint Committee on Cancer staging system in all groups. The calibration plots showed that the predicted prognosis of potential-CT-benefit-ANN was highly consistent with the actual value. The survival curves showed that potential-CT-benefit-ANN could stratify prognosis well for all groups and performed significantly better than the eighth AJCC staging system.

CONCLUSION

The potential-CT-benefit-ANN model developed in this study can accurately predict the potential benefits of adjuvant chemotherapy in patients with stage II/III gastric cancer. The benefit score based on potential-CT-benefit-ANN can predict the long-term prognosis of patients with adjuvant chemotherapy and has good prognostic stratification ability.

Maternal knowledge-seeking behavior among pregnant women in Tanzania

BACKGROUND

Maternal mortality continues to be a global challenge with about 830 women dying of childbirth and pregnancy complications every day. Tanzania has a maternal mortality rate of 524 deaths per 100,000 live births.

OBJECTIVE

Knowing symptoms associated with antenatal risks among pregnant women may result in seeking care earlier or self-advocating for more immediate treatment in health facilities. This article sought to identify knowledge-seeking behaviors of pregnant women in Northern Tanzania, to determine the challenges met and how these should be addressed to enhance knowledge on pregnancy risks and when to seek care.

METHODS

Interview questions and questionnaires were the main data collection tools. Six gynecologists and four midwives were interviewed, while 168 pregnant women and 14 recent mothers participated in the questionnaires.

RESULTS

With the rise in mobile technology and Internet penetration in Tanzania, more women are seeking information through online sources. However, for women to trust these sources, medical experts have to be involved in developing the systems.

CONCLUSION

Through expert systems diagnosis of pregnancy complications and recommendations from experts can be made available to pregnant women in Tanzania. In addition, self-care education during pregnancy will save women money and reduce hospital loads in Tanzania.

Prediction of neonatal deaths in NICUs: development and validation of machine learning models

Background

Prediction of neonatal deaths in NICUs is important for benchmarking and evaluating healthcare services in NICUs. Application of machine learning techniques can improve physicians’ ability to predict the neonatal deaths. The aim of this study was to present a neonatal death risk prediction model using machine learning techniques.

Methods

This study was conducted in Tehran, Iran in two phases. Initially, important risk factors in neonatal death were identified and then several machine learning models including Artificial Neural Network (ANN), decision tree (Random Forest (RF), C5.0 and CHART tree), Support Vector Machine (SVM), Bayesian Network and Ensemble models were developed. Finally, we prospectively applied these models to predict neonatal death in a NICU and followed up the neonates to compare the outcomes of these neonates with real outcomes.

Results

17 factors were considered important in neonatal mortality prediction. The highest Area Under the Curve (AUC) was achieved for the SVM and Ensemble models with 0.98. The best precision and specificity were 0.98 and 0.94, respectively for the RF model. The highest accuracy, sensitivity and F-score were achieved for the SVM model with 0.94, 0.95 and 0.96, respectively. The best performance of models in prospective evaluation was for the ANN, C5.0 and CHAID tree models.

Conclusion

Using the developed machine learning models can help physicians predict the neonatal deaths in NICUs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-021-01497-8.

Learning from imbalanced fetal outcomes of systemic lupus erythematosus in artificial neural networks

OBJECTIVE

To explore an effective algorithm based on artificial neural network to pick correctly the minority of pregnant women with SLE suffering fetal loss outcomes from the majority with live birth and train a well behaved model as a clinical decision assistant.

METHODS

We integrated the thoughts of comparative and focused study into the artificial neural network and presented an effective algorithm aiming at imbalanced learning in small dataset.

RESULTS

We collected 469 non-trivial pregnant patients with SLE, where 420 had live-birth outcomes and the other 49 patients ended in fetal loss. A well trained imbalanced-learning model had a high sensitivity of 19/21 ([Formula: see text]) for the identification of patients with fetal loss outcomes.

DISCUSSION

The misprediction of the two patients was explainable. Algorithm improvements in artificial neural network framework enhanced the identification in imbalanced learning problems and the external validation increased the reliability of algorithm.

CONCLUSION

The well-trained model was fully qualified to assist healthcare providers to make timely and accurate decisions.

Comprehensive evaluation of the combined extracts of Epimedii Folium and Ligustri Lucidi Fructus for PMOP in ovariectomized rats based on MLP-ANN methods

ETHNOPHARMACOLOGICAL RELEVANCE

Kidney deficiency is the main pathogenesis of osteoporosis based on the theory of "kidney governing bones" in traditional Chinese medicine (TCM). Osteoporosis is a systemic disease; kidney deficiency influences the growth, aging and reproduction of human body, reflecting in endocrine, nerve, immunity, metabolism and other functions. Multi-target drugs composed of natural non-toxic products from kidney-reinforcing herbs, are being investigated for the treatment of osteoporosis. Therefore, it is necessary and imperative to develop an objective and comprehensive method to evaluate and compare the effects of herbs with listed drugs.

AIM OF THE STUDY

This study was designed to evaluate and compare the therapeutic effects and the underlying molecular mechanism of the combined extracts of Epimedii Folium and Ligustri Lucidi Fructus (EL) with Raloxifene hydrochloride (RH) in ovariectomy (OVX)-induced postmenopausal osteoporosis (PMOP) rats based on the multi-layer perception (MLP)-artificial neural network (ANN) model.

MATERIALS AND METHODS

Female SD rats were subjected to either sham surgery (n = 8) or bilateral OVX (n = 48). One week after recovering from surgery, the OVX-induced rats were randomly divided into three groups: OVX model group (n = 32, every 8 rats were killed at the end of the 5th, 9th, 11th or 13th week after OVX), EL group (treated with EL 0.35 g/kg, n = 8), and RH group (treated with RH 6.25 mg/kg, n = 8). The rats in the treatment groups were administrated once a day for 12 weeks, then sacrificed. We observed bone mass and quality, bone remodeling, the function of estrogen and TGF-β1/Smads pathway in all rats.

RESULTS

Both EL and RH could increase bone mineral density, enhance bone strength, relieve bone micro-structure degeneration, re-balance bone remodeling, regulate estrogen dysfunction, and up-regulate TGF-β1 expression. The evaluation of the MLP-ANN model showed that EL and RH had markedly anti-PMOP effects, and there was no significant difference in the comprehensive evaluation of anti-osteoporosis between the two drugs. However, RH had better effects on bone mass and quality and TGF-β1/Smads pathway than EL; EL had better effects on estrogen function than RH.

CONCLUSION

Combined extracts of Epimedii Folium and Ligustri Lucidi Fructus (EL) exhibited bone-protective effects on PMOP. The MLP-ANN method evaluated the efficacy of drugs more comprehensively, which provided a new direction for the evaluation and comparison of drugs.

The impact of education and clinical decision support on the quality of positive antinuclear antibody referrals

To implement and evaluate an intervention using education and clinical decision support (CDS) to improve the quality of positive ANA referrals. We retrospectively reviewed "positive ANA" referrals from April 2017 to May 2019. Demographic data and referring provider's location were recorded. Final diagnoses were categorized into two groups: rheumatic disease (RD) or no RD. We compared pre- and post-intervention groups for each type of referral. The positive predictive value (PPV) of an ANA referral leading to an RD for each referral group was calculated. Our intervention consisted of an educational poster and CDS which included a hard-stop prompt embedded into the electronic ANA order. All internal subgroups received CDS; only the main campus primary care providers (IPCP) received the educational poster. The external (EXT) referral subgroup did not receive either intervention. We found a significant increase in the number of RDs diagnosed post-intervention (p = 0.007). The PPV for all referrals increased from 16% to 26% during this project. All groups demonstrated improvement in PPV except the EXT group, which showed no change. Subgroups which demonstrated significant increase in the diagnosis of RD included total internal (p = 0.0005), internal PCP (p = 0.002), and affiliated primary care providers (p = 0.0002). The IPCP subgroup additionally received the educational intervention and did not demonstrate significant improvement. Implementing an intervention with a CDS component helps improve the quality of positive ANA referrals to rheumatology. Key Points • Clinical decision support improves the quality of positive ANA referrals. • Incorporating clinical decision support within the ANA order of an EHR is an effective way to deliver information to impact ordering at the "point of care."

Immunotherapies application in active stage of systemic lupus erythematosus in pregnancy: A case report and review of literature

BACKGROUND

Pregnancy in the setting of systemic lupus erythematosus can worsen the condition from the stable to active stage, with quality of life and fertility desire being particular concerns. Pregnancy in the active stage of systemic lupus erythematosus (ASLE), although rare and complicated to manage, can be treated favorably with immunotherapies ifs used properly. Here we report such a success case.

CASE SUMMARY

A 31-year-old primigravida patient, diagnosed with SLE seven years ago, was induced ASLE after a cold at 21 + weeks. The patient’s vital signs on presentation were normal. Her laboratory exam was remarkable for significant proteinuria, liver and renal dysfunction, and low C3 and C4 levels. Infectious work-up was negative. The patient was diagnosed with ASLE. She was given immunosuppressive agents (methylprednisolone, gamma globulin and azathioprine etc.) and plasma adsorption therapy, monitoring blood pressure every 8 h, fetal heart rate twice a day, and liver and renal function at least twice a week. Successful maternal and fetal outcomes are presented here.

CONCLUSION

Child-bearing in ASLE has become more promising, even for this difficult case of ASLE with multiple organ damage. Thorough antepartum counseling, cautious maternal-fetal monitoring, and multi-organ function monitoring by multidisciplinary specialties are keys to favorable pregnancy outcomes.

Comparison of Multivariable Logistic Regression and Other Machine Learning Algorithms for Prognostic Prediction Studies in Pregnancy Care: Systematic Review and Meta-Analysis

Background

Predictions in pregnancy care are complex because of interactions among multiple factors. Hence, pregnancy outcomes are not easily predicted by a single predictor using only one algorithm or modeling method.

Objective

This study aims to review and compare the predictive performances between logistic regression (LR) and other machine learning algorithms for developing or validating a multivariable prognostic prediction model for pregnancy care to inform clinicians’ decision making.

Methods

Research articles from MEDLINE, Scopus, Web of Science, and Google Scholar were reviewed following several guidelines for a prognostic prediction study, including a risk of bias (ROB) assessment. We report the results based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Studies were primarily framed as PICOTS (population, index, comparator, outcomes, timing, and setting): Population: men or women in procreative management, pregnant women, and fetuses or newborns; Index: multivariable prognostic prediction models using non-LR algorithms for risk classification to inform clinicians’ decision making; Comparator: the models applying an LR; Outcomes: pregnancy-related outcomes of procreation or pregnancy outcomes for pregnant women and fetuses or newborns; Timing: pre-, inter-, and peripregnancy periods (predictors), at the pregnancy, delivery, and either puerperal or neonatal period (outcome), and either short- or long-term prognoses (time interval); and Setting: primary care or hospital. The results were synthesized by reporting study characteristics and ROBs and by random effects modeling of the difference of the logit area under the receiver operating characteristic curve of each non-LR model compared with the LR model for the same pregnancy outcomes. We also reported between-study heterogeneity by using τ² and I².

Results

Of the 2093 records, we included 142 studies for the systematic review and 62 studies for a meta-analysis. Most prediction models used LR (92/142, 64.8%) and artificial neural networks (20/142, 14.1%) among non-LR algorithms. Only 16.9% (24/142) of studies had a low ROB. A total of 2 non-LR algorithms from low ROB studies significantly outperformed LR. The first algorithm was a random forest for preterm delivery (logit AUROC 2.51, 95% CI 1.49-3.53; I²=86%; τ²=0.77) and pre-eclampsia (logit AUROC 1.2, 95% CI 0.72-1.67; I²=75%; τ²=0.09). The second algorithm was gradient boosting for cesarean section (logit AUROC 2.26, 95% CI 1.39-3.13; I²=75%; τ²=0.43) and gestational diabetes (logit AUROC 1.03, 95% CI 0.69-1.37; I²=83%; τ²=0.07).

Conclusions

Prediction models with the best performances across studies were not necessarily those that used LR but also used random forest and gradient boosting that also performed well. We recommend a reanalysis of existing LR models for several pregnancy outcomes by comparing them with those algorithms that apply standard guidelines.

Trial Registration

PROSPERO (International Prospective Register of Systematic Reviews) CRD42019136106; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=136106

Validation of the usefulness of artificial neural networks for risk prediction of adverse drug reactions used for individual patients in clinical practice

Artificial neural networks are the main tools for data mining and were inspired by the human brain and nervous system. Studies have demonstrated their usefulness in medicine. However, no studies have used artificial neural networks for the prediction of adverse drug reactions. We aimed to validate the usefulness of artificial neural networks for the prediction of adverse drug reactions and focused on vancomycin -induced nephrotoxicity. For constructing an artificial neural network, a multilayer perceptron algorithm was employed. A 10-fold cross validation method was adopted for evaluating the resultant artificial neural network. In total, 1141 patients who received vancomycin at Hokkaido University Hospital from November 2011 to February 2019 were enrolled. Among these patients, 179 (15.7%) developed vancomycin -induced nephrotoxicity. The top three risk factors of vancomycin -induced nephrotoxicity which are relatively important in the artificial neural networks were average vancomycin trough concentration ≥ 13.0 mg/L and concomitant use of piperacillin–tazobactam and vasopressor drugs. The predictive accuracy of the artificial neural network was 86.3% and that of the multiple logistic regression model (conventional statistical method) was 85.1%. Moreover, area under the receiver operating characteristic curve (AUROC) of the artificial neural network was 0.83. In the 10-fold cross-validation, the accuracy obtained was 86.0% and AUROC was 0.82. The artificial neural network model predicting the vancomycin -induced nephrotoxicity showed good predictive performance. This appears to be the first report of the usefulness of artificial neural networks for an adverse drug reactions risk prediction model.

Development and validation of an artificial neural network prognostic model after gastrectomy for gastric carcinoma: An international multicenter cohort study

BACKGROUND

Recently, artificial neural network (ANN) methods have also been adopted to deal with the complex multidimensional nonlinear relationship between clinicopathologic variables and survival for patients with gastric cancer. Using a multinational cohort, this study aimed to develop and validate an ANN-based survival prediction model for patients with gastric cancer.

METHODS

Patients with gastric cancer who underwent gastrectomy in a Chinese center, a Japanese center, and recorded in the Surveillance, Epidemiology, and End Results database, respectively, were included in this study. Multilayer perceptron neural network was used to develop the prediction model. Time-dependent receiver operating characteristic (ROC) curves, area under the curves (AUCs), and decision curve analysis (DCA) were used to compare the ANN model with previous prediction models.

RESULTS

An ANN model with nine input nodes, nine hidden nodes, and two output nodes was constructed. These three cohort's data showed that the AUC of the model was 0.795, 0.836, and 0.850 for 5-year survival prediction, respectively. In the calibration curve analysis, the ANN-predicted survival had a high consistency with the actual survival. Comparison of the DCA and time-dependent ROC between the ANN model and previous prediction models showed that the ANN model had good and stable prediction capability compared to the previous models in all cohorts.

CONCLUSIONS

The ANN model has significantly better discriminative capability and allows an individualized survival prediction. This model has good versatility in Eastern and Western data and has high clinical application value.

A Clinical Decision Support System for Predicting the Early Complications of One-Anastomosis Gastric Bypass Surgery

BACKGROUND/OBJECTIVE

One of the most effective treatments for patients with obesity, albeit with some complications, is obesity surgery. The aim of this study was to develop a clinical decision support system (CDSS) to predict the early complications of one-anastomosis gastric bypass (OAGB) surgery.

SUBJECTS/METHODS

This study was conducted in Tehran, Iran on patients who underwent OAGB surgery in 2011-2014 in five hospitals. Initially, variables affecting the OAGB early complications were identified using the literature review. Patients' data were extracted from an existing database of obesity surgery. Then, different artificial neural networks (ANNs) (multilayer perceptron (MLP) network) were developed and evaluated for prediction of 10-day, 1-month, and 3-month complications.

RESULTS

Factors including age, BMI, smoking status, intra-operative complications, comorbidities, laboratory tests, sonography results, and endoscopy results were considered important factors for predicting early complications of OAGB. A CDSS was developed with these variables. The accuracy, specificity, and sensitivity of the 10-day prediction system in the test data were 98.4%, 98.6%, and 98.3%, respectively. These figures for 1-month system were 96%, 93%, and 98.4% and for the 3-month system were 89.3%, 86.6%, and 91.5%, respectively.

CONCLUSIONS

Using the CDSS designed, we could accurately predict the early complications of OAGB surgery.

Maternal and Fetal Outcomes in Iranian Patients with Systemic Lupus Erythematosus: A Five-Year Retrospective Study of 60 Pregnancies

BACKGROUND

As a multisystem autoimmune disease, Systemic Lupus Erythematosus (SLE) mainly affects women during reproductive age. This retrospective study was designed to investigate the fetal and maternal outcomes of Iranian women with SLE.

METHODS

Clinical and laboratory records of 60 pregnancies in 55 SLE patients who attended Hafez hospital, a tertiary referral center for high risk pregnancies and SLE patients affiliated with Shiraz University of Medical Science, were reviewed during April 2012 and March 2016.

RESULTS

The mean age of the patients was 29.28±4.6 years and mean disease duration was 5.09±4.2 years. Live birth rate was 83.3% after exclusion of elective abortions. There were 50 live births, 3 neonatal deaths, 3 spontaneous abortions and 7 stillbirths. 9 (15%) women developed preeclampsia and there was 1 (1.6%) case of HELLP syndrome. Lupus flares occurred in 27 (45%) patients during pregnancy. Preterm delivery occurred in 11.6% of pregnancies. Skin and joints were the most frequently affected organs. Patients with previous lupus nephritis (n=18) were associated with a higher risk of maternal complication, but fetal outcomes were similar in both groups. Cesarean rate was about 66%, mostly related to fetal indications (50%).

CONCLUSION

Pregnancies in most women with pre-existing SLE can now be managed with successful results although presence of previous lupus nephritis is still a major risk factor for adverse maternal outcomes. In our study, fetal outcome was not different between patients with lupus nephritis compared with the patients without nephritis who were under treatments. Hence, to achieve favorable long-term results, we recommend regular multispecialty treatment approaches and progestational counseling for women with SLE.

Enabling pregnant women and their physicians to make informed medication decisions using artificial intelligence

The role of artificial intelligence (AI) in healthcare for pregnant women. To assess the role of AI in women’s health, discover gaps, and discuss the future of AI in maternal health. A systematic review of English articles using EMBASE, PubMed, and SCOPUS. Search terms included pregnancy and AI. Research articles and book chapters were included, while conference papers, editorials and notes were excluded from the review. Included papers focused on pregnancy and AI methods, and pertained to pharmacologic interventions. We identified 376 distinct studies from our queries. A final set of 31 papers were included for the review. Included papers represented a variety of pregnancy concerns and multidisciplinary applications of AI. Few studies relate to pregnancy, AI, and pharmacologics and therefore, we review carefully those studies. External validation of models and techniques described in the studies is limited, impeding on generalizability of the studies. Our review describes how AI has been applied to address maternal health, throughout the pregnancy process: preconception, prenatal, perinatal, and postnatal health concerns. However, there is a lack of research applying AI methods to understand how pharmacologic treatments affect pregnancy. We identify three areas where AI methods could be used to improve our understanding of pharmacological effects of pregnancy, including: (a) obtaining sound and reliable data from clinical records (15 studies), (b) designing optimized animal experiments to validate specific hypotheses (1 study) to (c) implementing decision support systems that inform decision-making (11 studies). The largest literature gap that we identified is with regards to using AI methods to optimize translational studies between animals and humans for pregnancy-related drug exposures.

Clinical decision support system to predict chronic kidney disease: A fuzzy expert system approach

BACKGROUND AND OBJECTIVES

Diagnosis and early intervention of chronic kidney disease are essential to prevent loss of kidney function and a large amount of financial resources. To this end, we developed a fuzzy logic-based expert system for diagnosis and prediction of chronic kidney disease and evaluate its robustness against noisy data.

METHODS

At first, we identified the diagnostic parameters and risk factors through a literature review and a survey of 18 nephrologists. Depending on the features selected, a set of fuzzy rules for the prediction of chronic kidney disease was determined by reviewing the literature, guidelines and consulting with nephrologists. Fuzzy expert system was developed using MATLAB software and Mamdani Inference System. Finally, the fuzzy expert system was evaluated using data extracted from 216 randomly selected medical records of patients with and without chronic kidney disease. We added noisy data to our dataset and compare the performance of the system on original and noisy datasets.

RESULTS

We selected 16 parameters for the prediction of chronic kidney disease. The accuracy, sensitivity, and specificity of the final system were 92.13 %, 95.37 %, and 88.88 %, respectively. The area under the curve was 0.92 and the Kappa coefficient was 0.84, indicating a very high correlation between the system diagnosis and the final diagnosis recorded in the medical records. The performance of the system on noisy input variables indicated that in the worse scenario, the accuracy, sensitivity, and specificity of the system decreased only by 4.43 %, 7.48 %, and 5.41 %, respectively.

CONCLUSION

Considering the desirable performance of the proposed expert system, the system can be useful in the prediction of chronic kidney disease.

Predictive models for patients with lung carcinomas to identify EGFR mutation status via an artificial neural network based on multiple clinical information

PURPOSE

Epidermal growth factor receptor (EGFR) mutation testing has several limitations. Therefore, we built predictive models to determine the EGFR mutation status of patients and guide therapeutic decision-making.

METHODS

We collected data from 320 patients with lung carcinoma, including sex, age, smoking history, serum tumour marker levels, maximum standardized uptake value, pathological results, computed tomography images, and EGFR mutation status. Artificial neural network (ANN) models based on multiple clinical characteristics were proposed to predict EGFR mutation status.

RESULTS

A training set (n = 200) was used to develop predictive models of the EGFR mutation status (Model 1: area under the receiver operating characteristic curve [AUROC] = 0.910, 95% CI 0.861-0.945; Model 2: AUROC = 0.859, 95% CI 0.803-0.904; Model 3: AUROC = 0.711, 95% CI 0.643-0.773). A testing set (n = 50) and temporal validation data set (n = 70) were used to evaluate the generalisation performance of the established models (testing set: Model 1, AUROC = 0.845, 95% CI 0.715-0.932; Model 2, AUROC = 0.882, 95% CI 0.759-0.956; Model 3, AUROC = 0.817, 95% CI 0.682-0.912; temporal validation dataset: Model 1, AUROC = 0.909, 95% CI 0.816-0.964; Model 2, AUROC = 0.855, 95% CI 0.751-0.928; Model 3, AUROC = 0.831, 95% CI 0.723-0.910). The predictive abilities of the three ANN models were superior to that of a previous logistic regression model (P < 0.001, 0.027, and 0.050, respectively).

CONCLUSIONS

ANN models provide a non-invasive and readily available method for EGFR mutation status prediction.

Artificial Intelligence in Clinical Decision Support: a Focused Literature Survey

OBJECTIVES

This survey analyses the latest literature contributions to clinical decision support systems (DSSs) on a two-year period (2017-2018), focusing on the approaches that adopt Artificial Intelligence (AI) techniques in a broad sense. The goal is to analyse the distribution of data-driven AI approaches with respect to "classical" knowledge-based ones, and to consider the issues raised and their possible solutions.

METHODS

We included PubMed and Web of ScienceTM publications, focusing on contributions describing clinical DSSs that adopted one or more AI methodologies.

RESULTS

We selected 75 papers, 49 of which describe approaches in the data-driven AI area, 20 present purely knowledge-based DSSs, and 6 adopt hybrid approaches relying on both formalized knowledge and data.

CONCLUSIONS

Recent studies in the clinical DSS area demonstrate a prevalence of data-driven AI, which can be adopted autonomously in purely data-driven systems, or in cooperation with domain knowledge in hybrid systems. Such hybrid approaches, able to conjugate all available knowledge sources through proper knowledge integration steps, represent an interesting example of synergy between the two AI categories. This synergy can lead to the resolution of some existing issues, such as the need for transparency and explainability, nowadays recognized as central themes to be addressed by both AI and medical informatics research.

Prediction of fetal loss in Chinese pregnant patients with systemic lupus erythematosus: a retrospective cohort study

OBJECTIVE

To develop a predictive model for fetal loss in women with systemic lupus erythematosus (SLE).

DESIGN

A retrospective cohort study.

SETTING

Data were collected in a tertiary medical centre, located in Shanghai, China, from September 2011 to May 2017.

PARTICIPANTS

338 pregnancies with SLE were analysed retrospectively. Cases of multiple pregnancy and those in which artificial abortion was performed for personal reasons were excluded.

PRIMARY OUTCOME MEASURES

Fetal loss was the primary outcome. A stepwise regression to identify the predictors related to the fetal loss and coefficient B of each variable was used to develop a predictive model and make a corresponding risk classification. The Hosmer-Lemeshow test, Omnibus test and area under the receiver-operating characteristic curve (AUC) were used to assess the goodness-of-fit and discrimination of the predictive model. A 10-fold cross validation was used to assess the model for overfitting.

RESULTS

Unplanned pregnancies (OR 2.84, 95% CI 1.12 to 7.22), C₃ hypocomplementemia (OR 5.46, 95% CI 2.30 to 12.97) and 24 hour-urinary protein level (0.3≤protein<1.0 g/24 hours: OR 2.10, 95% CI 0.63 to 6.95; protein≥1.0 g/24 hours: OR 5.89, 95% CI 2.30 to 15.06) were selected by the stepwise regression. The Hosmer-Lemeshow test resulted in p=0.325; the Omnibus test resulted in p<0.001 and the AUC was 0.829 (95% CI 0.744 to 0.91) in the regression model. The corresponding risk score classification was divided into low risk (0-3) and high risk groups (>3), with a sensitivity of 60.5%, a specificity of 93.3%, positive likelihood ratio of 9.03 and negative likelihood ratio of 0.42.

CONCLUSIONS

A predictive model for fetal loss in women with SLE was developed using the timing of conception, C₃ complement and 24 hour-urinary protein level. This model may help clinicians in identifying women with high risk pregnancies, thereby carrying out monitoring or/and interventions for improving fetal outcomes.

Diagnosis of mesothelioma with deep learning

Malignant mesothelioma (MM) is a rare but aggressive cancer. The definitive diagnosis of MM is critical for effective treatment and has important medicolegal significance. However, the definitive diagnosis of MM is challenging due to its composite epithelial/mesenchymal pattern. The aim of the current study was to develop a deep learning method to automatically diagnose MM. A retrospective analysis of 324 participants with or without MM was performed. Significant features were selected using a genetic algorithm (GA) or a ReliefF algorithm performed in MATLAB software. Subsequently, the current study constructed and trained several models based on a backpropagation (BP) algorithm, extreme learning machine algorithm and stacked sparse autoencoder (SSAE) to diagnose MM. A confusion matrix, F-measure and a receiver operating characteristic (ROC) curve were used to evaluate the performance of each model. A total of 34 potential variables were analyzed, while the GA and ReliefF algorithms selected 19 and 5 effective features, respectively. The selected features were used as the inputs of the three models. SSAE and GA+SSAE demonstrated the highest performance in terms of classification accuracy, specificity, F-measure and the area under the ROC curve. Overall, the GA+SSAE model was the preferred model since it required a shorter CPU time and fewer variables. Therefore, the SSAE with GA feature selection was selected as the most accurate model for the diagnosis of MM. The deep learning methods developed based on the GA+SSAE model may assist physicians with the diagnosis of MM.

Predicting postoperative facial swelling following impacted mandibular third molars extraction by using artificial neural networks evaluation

Patients' postoperative facial swelling following third molars extraction may have both biological impacts and social impacts. The purpose of this study was to evaluate the accuracy of artificial neural networks in the prediction of the postoperative facial swelling following the impacted mandibular third molars extraction. The improved conjugate grads BP algorithm combining with adaptive BP algorithm and conjugate gradient BP algorithm together was used. In this neural networks model, the functional projective relationship was established among patient's personal factors, anatomy factors of third molars and factors of surgical procedure to facial swelling following impacted mandibular third molars extraction. This neural networks model was trained and tested based on the data from 400 patients, in which 300 patients were made as the training samples, and another100 patients were assigned as the test samples. The improved conjugate grads BP algorithm was able to not only avoid the problem of local minimum effectively, but also improve the networks training speed greatly. 5-fold cross-validation was used to get a better sense of the predictive accuracy of the neural network and early stopping was used to improve generalization. The accuracy of this model was 98.00% for the prediction of facial swelling following impacted mandibular third molars extraction. This artificial intelligence model is approved as an accurate method for prediction of the facial swelling following impacted mandibular third molars extraction.

Evaluation of machine learning algorithms for improved risk assessment for Down's syndrome

Prenatal screening generates a great amount of data that is used for predicting risk of various disorders. Prenatal risk assessment is based on multiple clinical variables and overall performance is defined by how well the risk algorithm is optimized for the population in question. This article evaluates machine learning algorithms to improve performance of first trimester screening of Down syndrome. Machine learning algorithms pose an adaptive alternative to develop better risk assessment models using the existing clinical variables. Two real-world data sets were used to experiment with multiple classification algorithms. Implemented models were tested with a third, real-world, data set and performance was compared to a predicate method, a commercial risk assessment software. Best performing deep neural network model gave an area under the curve of 0.96 and detection rate of 78% with 1% false positive rate with the test data. Support vector machine model gave area under the curve of 0.95 and detection rate of 61% with 1% false positive rate with the same test data. When compared with the predicate method, the best support vector machine model was slightly inferior, but an optimized deep neural network model was able to give higher detection rates with same false positive rate or similar detection rate but with markedly lower false positive rate. This finding could further improve the first trimester screening for Down syndrome, by using existing clinical variables and a large training data derived from a specific population.