Machine learning methods in sport injury prediction and prevention: a systematic review

Abstract Purpose Injuries are common in sports and can have significant physical, psychological and financial consequences. Machine learning (ML) methods could be used to improve injury prediction and allow proper approaches to injury prevention. The aim of our study was therefore to perform a systematic review of ML methods in sport injury prediction and prevention. Methods A search of the PubMed database was performed on March 24th 2020. Eligible articles included original studies investigating the role of ML for sport injury prediction and prevention. Two independent reviewers screened articles, assessed eligibility, risk of bias and extracted data. Methodological quality and risk of bias were determined by the Newcastle–Ottawa Scale. Study quality was evaluated using the GRADE working group methodology. Results Eleven out of 249 studies met inclusion/exclusion criteria. Different ML methods were used (tree-based ensemble methods (n = 9), Support Vector Machines (n = 4), Artificial Neural Networks (n = 2)). The classification methods were facilitated by preprocessing steps (n = 5) and optimized using over- and undersampling methods (n = 6), hyperparameter tuning (n = 4), feature selection (n = 3) and dimensionality reduction (n = 1). Injury predictive performance ranged from poor (Accuracy = 52%, AUC = 0.52) to strong (AUC = 0.87, f1-score = 85%). Conclusions Current ML methods can be used to identify athletes at high injury risk and be helpful to detect the most important injury risk factors. Methodological quality of the analyses was sufficient in general, but could be further improved. More effort should be put in the interpretation of the ML models.

Download Full-text

Machine learning techniques for mortality prediction in emergency departments: a systematic review

BMJ Open ◽

10.1136/bmjopen-2021-052663 ◽

2021 ◽

Vol 11 (11) ◽

pp. e052663

Author(s):

Amin Naemi ◽

Thomas Schmidt ◽

Marjan Mansourvar ◽

Mohammad Naghavi-Behzad ◽

Ali Ebrahimi ◽

...

Keyword(s):

Machine Learning ◽

Systematic Review ◽

Neural Networks ◽

Hospital Mortality ◽

Emergency Departments ◽

Missing Values ◽

Data Extraction ◽

Risk Of Bias ◽

Gradient Boosting ◽

Support Vector

ObjectivesThis systematic review aimed to assess the performance and clinical feasibility of machine learning (ML) algorithms in prediction of in-hospital mortality for medical patients using vital signs at emergency departments (EDs).DesignA systematic review was performed.SettingThe databases including Medline (PubMed), Scopus and Embase (Ovid) were searched between 2010 and 2021, to extract published articles in English, describing ML-based models utilising vital sign variables to predict in-hospital mortality for patients admitted at EDs. Critical appraisal and data extraction for systematic reviews of prediction modelling studies checklist was used for study planning and data extraction. The risk of bias for included papers was assessed using the prediction risk of bias assessment tool.ParticipantsAdmitted patients to the ED.Main outcome measureIn-hospital mortality.ResultsFifteen articles were included in the final review. We found that eight models including logistic regression, decision tree, K-nearest neighbours, support vector machine, gradient boosting, random forest, artificial neural networks and deep neural networks have been applied in this domain. Most studies failed to report essential main analysis steps such as data preprocessing and handling missing values. Fourteen included studies had a high risk of bias in the statistical analysis part, which could lead to poor performance in practice. Although the main aim of all studies was developing a predictive model for mortality, nine articles did not provide a time horizon for the prediction.ConclusionThis review provided an updated overview of the state-of-the-art and revealed research gaps; based on these, we provide eight recommendations for future studies to make the use of ML more feasible in practice. By following these recommendations, we expect to see more robust ML models applied in the future to help clinicians identify patient deterioration earlier.

Download Full-text

Physical Therapy to Prevent Osteopenia in Preterm Infants: A Systematic Review

Children ◽

10.3390/children8080664 ◽

2021 ◽

Vol 8 (8) ◽

pp. 664

Author(s):

Galaad Torró-Ferrero ◽

Francisco Javier Fernández-Rego ◽

Antonia Gómez-Conesa

Keyword(s):

Systematic Review ◽

Preterm Infants ◽

Methodological Quality ◽

Bone Mineralization ◽

Exercise Program ◽

Risk Of Bias ◽

Quantitative Synthesis ◽

Comprehensive Search ◽

Design Characteristics ◽

Almost All

Background: During the last trimester of pregnancy, about 80% of the infant’s calcium is incorporated, and for this reason, preterm infants have less bone mineralization compared to those born at term. The aim of the present systematic review was to identify, evaluate and summarize the studies that deal with the effect of physiotherapy modalities in the prevention and treatment of osteopenia in preterm infants. Methods: A comprehensive search (09/2019–02/2021) using PubMed, Web of Science, SCOPUS, ProQuest, SciELO, Latindex, ScienceDirect, PEDro and ClinicalTrials.gov was carried out. The following data were extracted: The number of participants, characteristics of the participants, design, characteristics of the intervention, outcome measures, time of evaluation and results. A non-quantitative synthesis of the extracted data was performed. The methodological quality and risk of bias were assessed using a PEDro scale and ROB-2 scale, respectively. Results: A total of 16 studies were analyzed, presenting a methodological quality that ranged from 3 to 8 points, and all showed some concerns regarding their risk of bias. Almost all studies (15/16) used passive mobilizations with joint pressure to prevent osteopenia, but they differed in the intensity and frequency of application. Conclusions: A daily exercise program of passive mobilizations with joint pressure, improves bone mineralization in preterm infants admitted to neonatal units.

Download Full-text

Methodological quality and risk of bias of meta-analyses of pharmacy services: A systematic review

Research in Social and Administrative Pharmacy ◽

10.1016/j.sapharm.2020.12.011 ◽

2020 ◽

Author(s):

Aline F. Bonetti ◽

Fernanda S. Tonin ◽

Ana M. Della Rocca ◽

Rosa C. Lucchetta ◽

Fernando Fernandez‐Llimos ◽

...

Keyword(s):

Systematic Review ◽

Methodological Quality ◽

Risk Of Bias ◽

Pharmacy Services ◽

Meta Analyses

Download Full-text

RANDOMISED CONTROLLED TRIALS OF CHRONIC PAIN IN CHILDREN AND ADOLESCENTS: A SYSTEMATIC REVIEW

Archives of Disease in Childhood ◽

10.1136/archdischild-2015-310148.69 ◽

2015 ◽

Vol 101 (1) ◽

pp. e1.66-e1

Author(s):

Rym Boulkedid ◽

Armiya Yousouf Abdou ◽

Emilie Desselas ◽

Marlène Monegat ◽

Corinne Alberti ◽

...

Keyword(s):

Systematic Review ◽

Chronic Pain ◽

Randomized Controlled Trials ◽

Children And Adolescents ◽

Methodological Quality ◽

Risk Of Bias ◽

Controlled Trials ◽

Recurrent Pain ◽

Pharmacological Agents ◽

Randomized Controlled

BackgroundApproximately 15 to 30% of children and adolescents suffer from daily pain persistent over more than 3 months and there is evidence supporting that the prevalence of chronic pain is steadily increasing in this population. Chronic pain is known to have a negative impact on children's development and social behaviour, leading often to severe psychological distress and physical disability. We reviewed medical literature to assess the characteristics and quality of randomized controlled trials (RCTs) on pharmacological and non-pharmacological therapies in chronic and recurrent pain in the paediatric population.MethodsWe performed a systematic search of PubMed, Embase and the Cochrane Library up to March 2014. Bibliographies of relevant articles were also hand-searched. We included all RCTs that involved children and adolescents (age 0 to 18 years) and evaluated the use of a pharmacological agent or a non-pharmacological approach in the context of chronic or recurrent pain. The latter was defined as pain persisting for more than 3 months. Methodological quality was evaluated using the Cochrane Risk of Bias Tool. Two reviewers independently assessed studies for inclusion and evaluated methodological quality.ResultsA total of 52 randomized controlled trials were selected and included in the analysis. The majority were conducted in single hospital institutions, with no information on study funding. Median sample size was 45 (34–57) participants. Almost 50% of the RCTs included both adults and children with a median age at inclusion of 13 years. Non-pharmacological approaches were more commonly tested whereas evaluation of pharmacological agents concerned less than 30% of RCTs. Abdominal pain and headache were the most common types of chronic pain experienced among trial participants. Overall, the methodological quality was poor and did not parallel the number of RCTs that increased over the years. The risk of bias was high or unclear in 70% of the trials.ConclusionsThis is the first systematic review of RCTs conducted to evaluate pharmacological and non-pharmacological therapies in chronic and recurrent pain in children and adolescents. Although, management of pain in adults has significantly improved over the years due to the evaluation of numerous analgesic therapies, our results highlight the existing knowledge gap with regards to children and adolescents. Therapeutic strategies, in particular pharmacological agents, applied to relieve chronic or recurrent pain in children and adolescents are not evaluated through high quality RCTs. The need to improve analgesic therapy in children and adolescents with chronic pain is still unmet. We discuss possible research constraints and challenges related to this fact as well as adequate methodologies to circumvent them.

Download Full-text

Machine Learning-based Prediction Models for Diagnosis and Prognosis in Inflammatory Bowel Diseases: A Systematic Review

Journal of Crohn s and Colitis ◽

10.1093/ecco-jcc/jjab155 ◽

2021 ◽

Author(s):

Nghia H Nguyen ◽

Dominic Picetti ◽

Parambir S Dulai ◽

Vipul Jairath ◽

William J Sandborn ◽

...

Keyword(s):

Machine Learning ◽

Systematic Review ◽

Risk Prediction ◽

Statistical Models ◽

Prediction Models ◽

Risk Of Bias ◽

Learning Models ◽

Bowel Diseases ◽

Inflammatory Bowel ◽

Machine Learning Models

Abstract Background and Aims There is increasing interest in machine learning-based prediction models in inflammatory bowel diseases (IBD). We synthesized and critically appraised studies comparing machine learning vs. traditional statistical models, using routinely available clinical data for risk prediction in IBD. Methods Through a systematic review till January 1, 2021, we identified cohort studies that derived and/or validated machine learning models, based on routinely collected clinical data in patients with IBD, to predict the risk of harboring or developing adverse clinical outcomes, and reported its predictive performance against a traditional statistical model for the same outcome. We appraised the risk of bias in these studies using the Prediction model Risk of Bias ASsessment (PROBAST) tool. Results We included 13 studies on machine learning-based prediction models in IBD encompassing themes of predicting treatment response to biologics and thiopurines, predicting longitudinal disease activity and complications and outcomes in patients with acute severe ulcerative colitis. The most common machine learnings models used were tree-based algorithms, which are classification approaches achieved through supervised learning. Machine learning models outperformed traditional statistical models in risk prediction. However, most models were at high risk of bias, and only one was externally validated. Conclusions Machine learning-based prediction models based on routinely collected data generally perform better than traditional statistical models in risk prediction in IBD, though frequently have high risk of bias. Future studies examining these approaches are warranted, with special focus on external validation and clinical applicability.

Download Full-text

Racecadotril for acute diarrhoea in children: systematic review and meta-analyses

Archives of Disease in Childhood ◽

10.1136/archdischild-2015-309676 ◽

2015 ◽

Vol 101 (3) ◽

pp. 234-240 ◽

Cited By ~ 14

Author(s):

Morris Gordon ◽

Anthony Akobeng

Keyword(s):

Systematic Review ◽

Adverse Events ◽

Methodological Quality ◽

Data Extraction ◽

Meta Analysis ◽

Acute Diarrhoea ◽

Risk Of Bias ◽

Duration Of Symptoms ◽

Duration Of Illness ◽

Significant Difference

ObjectiveRacecadotril is an antisecretory agent that can prevent fluid/electrolyte depletion from the bowel as a result of acute diarrhoea without affecting intestinal motility. An up-to-date systematic review is indicated to summarise the evidence on racecadotril for the treatment of acute diarrhoea in children.DesignA Cochrane format systematic review of randomised controlled trials (RCTs). Data extraction and assessment of methodological quality were performed independently by two reviewers. Methodological quality was assessed using the Cochrane risk of bias tool.PatientsChildren with acute diarrhoea, as defined by the primary studies.InterventionsRCTs comparing racecadotril with placebo or other interventions.Main outcome measursDuration of illness, stool output/volume and adverse events.ResultsSeven RCTs were included, five comparing racecadotril with placebo or no intervention, one with pectin/kaolin and one with loperamide. Moderate to high risk of bias was present in all studies. There was no significant difference in efficacy or adverse events between racecadotril and loperamide. A meta-analysis of three studies with 642 participants showed significantly shorter duration of symptoms with racecadotril compared with placebo (mean difference −53.48 h, 95% CI −65.64 to −41.33). A meta-analysis of five studies with 949 participants showed no significant difference in adverse events between racecadotril and placebo (risk ratio 0.99, 95% CI 0.73 to 1.34).ConclusionsThere is some evidence that racecadotril is more effective than placebo or no intervention in reducing the duration of illness and stool output in children with acute diarrhoea. However, the overall quality of the evidence is limited due to sparse data, heterogeneity and risk of bias. Racecadotril appears to be safe and well tolerated.

Download Full-text

Rapid diagnostic test for antenatal syphilis screening in low-income and middle-income countries: a systematic review and meta-analysis

BMJ Open ◽

10.1136/bmjopen-2017-018132 ◽

2018 ◽

Vol 8 (2) ◽

pp. e018132 ◽

Cited By ~ 5

Author(s):

Carmen Phang Romero Casas ◽

Marrissa Martyn-St James ◽

Jean Hamilton ◽

Daniel S Marinho ◽

Rodolfo Castro ◽

...

Keyword(s):

Systematic Review ◽

High Risk ◽

Diagnostic Accuracy ◽

Pregnant Women ◽

Sensitivity And Specificity ◽

Methodological Quality ◽

Meta Analysis ◽

Risk Of Bias ◽

Antenatal Clinics

ObjectivesTo undertake a systematic review and meta-analysis to evaluate the test performance including sensitivity and specificity of rapid immunochromatographic syphilis (ICS) point-of-care (POC) tests at antenatal clinics compared with reference standard tests (non-treponemal (TP) and TP tests) for active syphilis in pregnant women.MethodsFive electronic databases were searched (PubMed, EMBASE, CRD, Cochrane Library and LILACS) to March 2016 for diagnostic accuracy studies of ICS test and standard reference tests for syphilis in pregnant women. Methodological quality was assessed using QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies). A bivariate meta-analysis was undertaken to generate pooled estimates of diagnostic parameters. Results were presented using a coupled forest plot of sensitivity and specificity and a scatter plot.ResultsThe methodological quality of the five included studies with regards to risk of bias and applicability concern judgements was either low or unclear. One study was judged as high risk of bias for patient selection due to exclusion of pregnant women with a previous history of syphilis, and one study was judged at high risk of bias for study flow and timing as not all patients were included in the analysis. Five studies contributed to the meta-analysis, providing a pooled sensitivity and specificity for ICS of 0.85 (95% CrI: 0.73 to 0.92) and 0.98 (95% CrI: 0.95 to 0.99), respectively.ConclusionsThis review and meta-analysis observed that rapid ICS POC tests have a high sensitivity and specificity when performed in pregnant women at antenatal clinics. However, the methodological quality of the existing evidence base should be taken into consideration when interpreting these results.PROSPERO registration numberCRD42016036335.

Download Full-text

Psychometric properties of the Jefferson Scale of Empathy: a COSMIN systematic review protocol

Systematic Reviews ◽

10.1186/s13643-019-1240-0 ◽

2019 ◽

Vol 8 (1) ◽

Cited By ~ 1

Author(s):

Brett Williams ◽

Bronwyn Beovich

Keyword(s):

Systematic Review ◽

Psychometric Properties ◽

Methodological Quality ◽

Grey Literature ◽

Risk Of Bias ◽

Measurement Properties ◽

Future Research ◽

Health Measurement ◽

Jefferson Scale Of Empathy

Abstract Background Empathy is an important characteristic to possess for healthcare professionals. It has been found to improve communication between professionals and patients and to improve clinical health outcomes. The Jefferson Scale of Empathy (JSE) was developed to measure this quality and has been used extensively, and psychometrically appraised, with a variety of cohorts and in different cultural environments. However, no study has been undertaken to systematically examine the methodological quality of studies which have assessed psychometric factors of the JSE. This systematic review will examine the quality of published papers that have reported on psychometric factors of the JSE. Methods A systematic review of studies which report on the psychometric properties of the JSE will be conducted. We will use a predefined search strategy to identify studies meeting the following eligibility criteria: original data is reported on for at least one of the psychometric measurement properties described in the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) Risk of Bias checklist, examines the JSE in a healthcare cohort (using the student, physician or health profession versions of the JSE), and is published from January 2001 and in the English language. Conference abstracts, editorials and grey literature will be excluded. Six electronic databases (Medline, EMBASE, PsychInfo, PubMed, Web of Science and CINAHL) will be systematically searched for articles meeting these criteria and studies will be assessed for eligibility by two review authors. The methodological quality of included papers will be examined using the COSMIN Risk of Bias checklist. Discussion A narrative description of the findings will be presented along with summary tables. Recommendations for use of the JSE with various cohorts and circumstances will be offered which may inform future research in this field. Systematic review registration PROSPERO CRD42018111412

Download Full-text

A Systematic Review of Machine Learning Techniques in Hematopoietic Stem Cell Transplantation (HSCT)

Sensors ◽

10.3390/s20216100 ◽

2020 ◽

Vol 20 (21) ◽

pp. 6100

Author(s):

Vibhuti Gupta ◽

Thomas M. Braun ◽

Mosharaf Chowdhury ◽

Muneesh Tewari ◽

Sung Won Choi

Keyword(s):

Machine Learning ◽

Systematic Review ◽

Cell Transplantation ◽

Hematopoietic Cell Transplantation ◽

Hematopoietic Cell ◽

Machine Learning Techniques ◽

Support Vector ◽

Hematopoietic Stem ◽

Curative Therapy ◽

Learning Techniques

Machine learning techniques are widely used nowadays in the healthcare domain for the diagnosis, prognosis, and treatment of diseases. These techniques have applications in the field of hematopoietic cell transplantation (HCT), which is a potentially curative therapy for hematological malignancies. Herein, a systematic review of the application of machine learning (ML) techniques in the HCT setting was conducted. We examined the type of data streams included, specific ML techniques used, and type of clinical outcomes measured. A systematic review of English articles using PubMed, Scopus, Web of Science, and IEEE Xplore databases was performed. Search terms included “hematopoietic cell transplantation (HCT),” “autologous HCT,” “allogeneic HCT,” “machine learning,” and “artificial intelligence.” Only full-text studies reported between January 2015 and July 2020 were included. Data were extracted by two authors using predefined data fields. Following PRISMA guidelines, a total of 242 studies were identified, of which 27 studies met the inclusion criteria. These studies were sub-categorized into three broad topics and the type of ML techniques used included ensemble learning (63%), regression (44%), Bayesian learning (30%), and support vector machine (30%). The majority of studies examined models to predict HCT outcomes (e.g., survival, relapse, graft-versus-host disease). Clinical and genetic data were the most commonly used predictors in the modeling process. Overall, this review provided a systematic review of ML techniques applied in the context of HCT. The evidence is not sufficiently robust to determine the optimal ML technique to use in the HCT setting and/or what minimal data variables are required.

Download Full-text

Performance in the Deep Squat Test and musculoskeletal injuries: a systematic review

Fisioterapia em Movimento ◽

10.1590/1980-5918.031.ao26 ◽

2018 ◽

Vol 31 (0) ◽

Author(s):

Priscila dos Santos Bunn ◽

Glória de Paula Silva ◽

Elirez Bezerra da Silva

Keyword(s):

Systematic Review ◽

Injury Risk ◽

Musculoskeletal Injuries ◽

Sports Teams ◽

Injury Prediction ◽

Bias Risk ◽

Newcastle Ottawa Scale ◽

Deep Squat

Abstract Introduction: The Deep Squat Test has been applied in pre-season evaluations of sports teams and in military courses to predict the risk of musculoskeletal injuries. Objective: To evaluate the association of DS performance and the risk of musculoskeletal injuries. Methods: In this systematic review, a search without language or time filters was carried out in MEDLINE, SciELO, SCOPUS, SPORTDiscuss, CINAHL and BVS databases with the following title words: injury prediction, injury risk and deep squat in December 2016. Participants' profile, sample size, classification of musculoskeletal injuries, follow-up time, study design and results were extracted from the studies. Bias risk analysis was performed with the Newcastle-Ottawa Scale. Results: Five studies were included, using different analyzes, whose results varied. Odds ratio ranged from 1.21 to 2.59 (95% CI = 1.01 - 3.28); relative risk was 1.68 (95% CI = 1.50 - 1.87), sensitivity from 3 to 24%, specificity from 90 to 99%, PPV from 42 to 63%, NPV from 72 to 75% and AUC from 51 to 58%. Conclusion: The DS can be a test whose presence of movement dysfunctions is a predictor of the risk of musculoskeletal injuries in individuals who practice physical exercises. However, due to the methodological limitations presented, caution is suggested when interpreting such results. PROSPERO registration: CRD4201706922.

Download Full-text