Machine Learning Takes Laboratory Automation to the Next Level

Artificial intelligence in the clinical laboratory

Laboratory medicine has become a highly automated medical discipline. Nowadays, artificial intelligence (AI) applied to laboratory medicine is also gaining more and more attention, which can optimize the entire laboratory workflow and even revolutionize laboratory medicine in the future. However, only a few commercially available AI models are currently approved for use in clinical laboratories and have drawbacks such as high cost, lack of accuracy, and the need for manual review of model results. Furthermore, there are a limited number of literature reviews that comprehensively address the research status, challenges, and future opportunities of AI applications in laboratory medicine. Our article begins with a brief introduction to AI and some of its subsets, then reviews some AI models that are currently being used in clinical laboratories or that have been described in emerging studies, and explains the existing challenges associated with their application and possible solutions, finally provides insights into the future opportunities of the field. We highlight the current status of implementation and potential applications of AI models in different stages of the clinical testing process.

Miniaturization and Automation Protocol of a Urinary Organic Acid Liquid-Liquid Extraction Method on GC-MS

The aim of this study was to improve the extraction method for urinary organic acids by miniaturizing and automating the process. Currently, manual extraction methods are commonly used, which can be time-consuming and lead to variations in test results. To address these issues, we reassessed and miniaturized the in-house extraction method, reducing the number of steps and the sample-to-solvent volumes required. The evaluated miniaturized method was translated into an automated extraction procedure on a MicroLab (ML) Star (Hamilton Technologies) liquid handler. This was then validated using samples obtained from the ERNDIM External Quality Assurance program. The organic acid extraction method was successfully miniaturized and automated using the Autosampler robot. The linear range for most of the thirteen standard analytes fell between 0 to 300 mg/L in spiked synthetic urine, with low (50 mg/L), medium (100 mg/L), and high (500 mg/L) levels. The correlation coefficient (r) for most analytes was >0.99, indicating a strong relationship between the measured values. Furthermore, the automated extraction method demonstrated acceptable precision, as most organic acids had coefficients of variation (CVs) below 20%. In conclusion, the automated extraction method provided comparable or even superior results compared to the current in-house method. It has the potential to reduce solvent volumes used during extraction, increase sample throughput, and minimize variability and random errors in routine diagnostic settings.

Clinlabomics: leveraging clinical laboratory data by data mining strategies

The recent global focus on big data in medicine has been associated with the rise of artificial intelligence (AI) in diagnosis and decision-making following recent advances in computer technology. Up to now, AI has been applied to various aspects of medicine, including disease diagnosis, surveillance, treatment, predicting future risk, targeted interventions and understanding of the disease. There have been plenty of successful examples in medicine of using big data, such as radiology and pathology, ophthalmology cardiology and surgery. Combining medicine and AI has become a powerful tool to change health care, and even to change the nature of disease screening in clinical diagnosis. As all we know, clinical laboratories produce large amounts of testing data every day and the clinical laboratory data combined with AI may establish a new diagnosis and treatment has attracted wide attention. At present, a new concept of radiomics has been created for imaging data combined with AI, but a new definition of clinical laboratory data combined with AI has lacked so that many studies in this field cannot be accurately classified. Therefore, we propose a new concept of clinical laboratory omics (Clinlabomics) by combining clinical laboratory medicine and AI. Clinlabomics can use high-throughput methods to extract large amounts of feature data from blood, body fluids, secretions, excreta, and cast clinical laboratory test data. Then using the data statistics, machine learning, and other methods to read more undiscovered information. In this review, we have summarized the application of clinical laboratory data combined with AI in medical fields. Undeniable, the application of Clinlabomics is a method that can assist many fields of medicine but still requires further validation in a multi-center environment and laboratory.

Microbiology 2.0–A “behind the scenes” consideration for artificial intelligence applications for interpretive culture plate reading in routine diagnostic laboratories

Laboratory automation with Artificial Intelligence (AI) features have now emerged into routine diagnostic clinical use to interpret growth on agar plates. Applications are currently limited to urine samples and infection control screens, yet some of the details around the development of algorithms remain entrenched with AI development specialists and are not well understood by laboratorians. The generation of algorithms is not a trivial task and is a highly structured process, with several considerations needed to develop the appropriate data for specific intended uses. Understanding these considerations highlights the limitations of any algorithm created and informs better design practices so that algorithm objectives can be thoroughly tested prior to routine use.

Total Laboratory Automation for Rapid Detection and Identification of Microorganisms and Their Antimicrobial Resistance Profiles

At a time when diagnostic bacteriological testing procedures have become more complex and their associated costs are steadily increasing, the expected benefits of Total laboratory automation (TLA) cannot just be a simple transposition of the traditional manual procedures used to process clinical specimens. In contrast, automation should drive a fundamental change in the laboratory workflow and prompt users to reconsider all the approaches currently used in the diagnostic work-up including the accurate identification of pathogens and the antimicrobial susceptibility testing methods. This review describes the impact of TLA in the laboratory efficiency improvement, as well as a new fully automated solution for AST by disk diffusion testing, and summarizes the evidence that implementing these methods can impact clinical outcomes.

Total Laboratory Automation and Three Shifts Reduce Turnaround Time of Cerebrospinal Fluid Culture Results in the Chinese Clinical Microbiology Laboratory

Background

Total laboratory automation (TLA) has the potential to reduce specimen processing time, optimize workflow, and decrease turnaround time (TAT). The purpose of this research is to investigate whether the TAT of our laboratory has changed since the adoption of TLA, as well as to optimize laboratory workflow, improve laboratory testing efficiency, and provide better services of clinical diagnosis and treatment.

Materials and Methods

Laboratory data was extracted from our laboratory information system in two 6-month periods: pre-TLA (July to December 2019) and post-TLA (July to December 2020), respectively.

Results

The median TAT for positive cultures decreased significantly from pre-TLA to post-TLA (65.93 vs 63.53, P<0.001). For different types of cultures, The TAT of CSF changed the most (86.76 vs 64.30, P=0.007), followed by sputum (64.38 vs 61.41, P<0.001), urine (52.10 vs 49,57, P<0.001), blood (68.49 vs 66.60, P<0.001). For Ascites and Pleural fluid, there was no significant difference (P>0.05). Further analysis found that the incidence of broth growth only for pre-TLA was 12.4% (14/133), while for post-TLA, it was 3.4% (4/119). The difference was statistically significant (P=0.01). The common isolates from CSF samples were Cryptococcus neoformans, coagulase-negative Staphylococcus, Acinetobacter baumannii, and Klebsiella pneumonia.

Conclusion

Using TLA and setting up three shifts shortened the TAT of our clinical microbiology laboratory, especially for CSF samples.

Validation of HER2 Status in Whole Genome Sequencing Data of Breast Cancers with the Ploidy-Corrected Copy Number Approach

BACKGROUND AND OBJECTIVE

Human epidermal growth factor receptor 2 (HER2) protein overexpression is one of the most significant biomarkers for breast cancer diagnostics, treatment prediction, and prognostics. The high accessibility of HER2 inhibitors in routine clinical practice directly translates into the diagnostic need for precise and robust marker identification. Even though multigene next-generation sequencing methodologies have slowly taken over the field of single-biomarker molecular tests, the copy number alterations such as amplification of the HER2-coding ERBB2 gene are hard to validate on next-generation sequencing platforms as they are characterized by chromosomal structural heterogeneity, polysomy, and genomic context of ploidy. In our study, we tested the approach of using whole genome sequencing instead of next-generation sequencing panels to determine HER2 status in the clinical set-up.

METHODS

We used a large dataset of 876 patients with breast cancer whole genomes with curated clinical data and an additional set of 551 patients' external genomic data. We used the decision-tree-based algorithm for optimization of the diagnostic tool for HER2 status assessment by whole genome sequencing.

RESULTS

The most efficient approach to assess HER2 status in whole genome sequencing data was the ploidy-corrected copy number, utilizing ERBB2 copy number and mean tumor ploidy. The classifier achieved sensitivity of 91.18% and specificity of 98.69% on the internal validation dataset and 89.86% and 96.06% on the external data, which is similar to other next-generation sequencing methods, currently tested in the clinic.

CONCLUSIONS

We provide evidence that the HER2 status may be reliably determined by whole genome sequencing and is applicable across different laboratory protocols and pipelines. We suggest using the ploidy-corrected copy number for diagnostic purposes.

Artificial Intelligence in Differential Diagnostics of Meningitis: A Nationwide Study

Differential diagnosis between bacterial and viral meningitis is crucial. In our study, to differentiate bacterial vs. viral meningitis, three machine learning (ML) algorithms (multiple logistic regression (MLR), random forest (RF), and naïve-Bayes (NB)) were applied for the two age groups (0-14 and >14 years) of patients with meningitis by both conventional (culture) and molecular (PCR) methods. Cerebrospinal fluid (CSF) neutrophils, CSF lymphocytes, neutrophil-to-lymphocyte ratio (NLR), blood albumin, blood C-reactive protein (CRP), glucose, blood soluble urokinase-type plasminogen activator receptor (suPAR), and CSF lymphocytes-to-blood CRP ratio (LCR) were used as predictors for the ML algorithms. The performance of the ML algorithms was evaluated through a cross-validation procedure, and optimal predictions of the type of meningitis were above 95% for viral and 78% for bacterial meningitis. Overall, MLR and RF yielded the best performance when using CSF neutrophils, CSF lymphocytes, NLR, albumin, glucose, gender, and CRP. Also, our results reconfirm the high diagnostic accuracy of NLR in the differential diagnosis between bacterial and viral meningitis.

Laboratory Automation in the Microbiology Laboratory: an Ongoing Journey, Not a Tale?

Clinical chemistry laboratories implemented fully automated devices decades before microbiologists started their subtle approaches to follow. Meanwhile several papers have been published about reduced time to reports, faster workflows, and increased sensitivity as results of lab automation. While the journey of automating microbiology workflows step by step was fascinating and beneficial, monetary aspects were uncommon in most publications. In this issue of the Journal of Clinical Microbiology, K. Culbreath, H. Piwonka, J. Korver, and M. Noorbakhsh (J Clin Microbiol 59:e01969-20, https://doi.org/10.1128/JCM.01969-20) calculate the benefits of total lab automation in terms of cost savings and lab efficiency in a "tale of four laboratories." The authors here provide facts and solid calculations about the benefits achieved in four different-sized labs after implementation of full laboratory automation.

Recent evolutions of machine learning applications in clinical laboratory medicine

Machine learning (ML) is gaining increased interest in clinical laboratory medicine, mainly triggered by the decreased cost of generating and storing data using laboratory automation and computational power, and the widespread accessibility of open source tools. Nevertheless, only a handful of ML-based products are currently commercially available for routine clinical laboratory practice. In this review, we start with an introduction to ML by providing an overview of the ML landscape, its general workflow, and the most commonly used algorithms for clinical laboratory applications. Furthermore, we aim to illustrate recent evolutions (2018 to mid-2020) of the techniques used in the clinical laboratory setting and discuss the associated challenges and opportunities. In the field of clinical chemistry, the reviewed applications of ML algorithms include quality review of lab results, automated urine sediment analysis, disease or outcome prediction from routine laboratory parameters, and interpretation of complex biochemical data. In the hematology subdiscipline, we discuss the concepts of automated blood film reporting and malaria diagnosis. At last, we handle a broad range of clinical microbiology applications, such as the reduction of diagnostic workload by laboratory automation, the detection and identification of clinically relevant microorganisms, and the detection of antimicrobial resistance.