RESUMEN
Background: CTG remains the only non-invasive tool available to the maternity team for continuous monitoring of fetal well-being during labour. Despite widespread use and investment in staff training, difficulty with CTG interpretation continues to be identified as a problem in cases of fetal hypoxia, which often results in permanent brain injury. Given the recent advances in AI, it is hoped that its application to CTG will offer a better, less subjective and more reliable method of CTG interpretation. Objectives: This mini-review examines the literature and discusses the impediments to the success of AI application to CTG thus far. Prior randomised control trials (RCTs) of CTG decision support systems are reviewed from technical and clinical perspectives. A selection of novel engineering approaches, not yet validated in RCTs, are also reviewed. The review presents the key challenges that need to be addressed in order to develop a robust AI tool to identify fetal distress in a timely manner so that appropriate intervention can be made. Results: The decision support systems used in three RCTs were reviewed, summarising the algorithms, the outcomes of the trials and the limitations. Preliminary work suggests that the inclusion of clinical data can improve the performance of AI-assisted CTG. Combined with newer approaches to the classification of traces, this offers promise for rewarding future development.
RESUMEN
INTRODUCTION: We present the first study to critically appraise the quality of reporting of the data analysis step in metabolomics studies since the publication of minimum reporting guidelines in 2007. OBJECTIVES: The aim of this study was to assess the standard of reporting of the data analysis step in metabolomics biomarker discovery studies and to investigate whether the level of detail supplied allows basic understanding of the steps employed and/or reuse of the protocol. For the purposes of this review we define the data analysis step to include the data pretreatment step and the actual data analysis step, which covers algorithm selection, univariate analysis and multivariate analysis. METHOD: We reviewed the literature to identify metabolomic studies of biomarker discovery that were published between January 2008 and December 2014. Studies were examined for completeness in reporting the various steps of the data pretreatment phase and data analysis phase and also for clarity of the workflow of these sections. RESULTS: We analysed 27 papers, published anytime in 2008 until the end of 2014 in the area or biomarker discovery in serum metabolomics. The results of this review showed that the data analysis step in metabolomics biomarker discovery studies is plagued by unclear and incomplete reporting. Major omissions and lack of logical flow render the data analysis' workflows in these studies impossible to follow and therefore replicate or even imitate. CONCLUSIONS: While we await the holy grail of computational reproducibility in data analysis to become standard, we propose that, at a minimum, the data analysis section of metabolomics studies should be readable and interpretable without omissions such that a data analysis workflow diagram could be extrapolated from the study and therefore the data analysis protocol could be reused by the reader. That inconsistent and patchy reporting obfuscates reproducibility is a given. However even basic understanding and reuses of protocols are hampered by the low level of detail supplied in the data analysis sections of the studies that we reviewed.
