Machine learning in the clinical microbiology laboratory: has the time come for routine practice?
N Peiffer-Smadja,S Dellière,C Rodriguez,G Birgand,F-X Lescure,S Fourati,E Ruppé,N. Peiffer-Smadja,S. Dellière,C. Rodriguez,G. Birgand,F.-X. Lescure,S. Fourati,E. Ruppé
DOI: https://doi.org/10.1016/j.cmi.2020.02.006
IF: 13.31
2020-10-01
Clinical Microbiology and Infection
Abstract:<h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Background</h3><p>Machine learning (ML) allows the analysis of complex and large data sets and has the potential to improve healthcare. The clinical microbiology laboratory, at the interface of clinical practice and diagnostics, is of special interest for the development of ML systems.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Objectives</h3><p>This narrative review aims to explore the current use of ML In clinical microbiology.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Sources</h3><p>References for this review were identified through searches of MEDLINE/PubMed, EMBASE, Google Scholar, biorXiv, arXiV, ACM Digital Library and IEEE Xplore Digital Library up to November 2019.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Content</h3><p>We found 97 ML systems aiming to assist clinical microbiologists. Overall, 82 (85%) ML systems targeted bacterial infections, 11 (11%) parasitic infections, nine (9%) viral infections and three (3%) fungal infections. Forty (41%) ML systems focused on micro-organisms detection, identification and quantification, 36 (37%) evaluated antimicrobial susceptibility and 21 (22%) targeted the diagnosis, disease classification and prediction of clinical outcomes. The ML systems used very diverse data sources: 21 (22%) used genomic data of micro-organisms, 19 (20%) microbiota data obtained by metagenomic sequencing, 19 (20%) analysed microscopic images, 17 (18%) spectroscopy data, 8 (8%) targeted gene sequencing, 6 (6%) volatile organic compounds, 4 (4%) bacterial colonies photographs, 4 (4%) transcriptome data, 3 (3%) protein structure and 3 (3%) clinical data. Most systems used high-income countries data (n=71, 73%) but a significant number used low- and middle-income countries data (n=36, 37%). Performance measures were reported for the 97 ML systems but no article described the use in clinical practice nor reported impact on processes or clinical outcomes.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Implications</h3><p>In clinical microbiology, ML has been used with various data sources and diverse practical applications. The evaluation and implementation processes represent the main gap of existing ML systems, requiring a focus on their interpretability and potential integration into real-world settings.</p>
infectious diseases,microbiology