Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

Radiology Radiology

Improving detection of impacted animal bones on lateral neck radiograph using a deep learning artificial intelligence algorithm.

In Insights into imaging

OBJECTIVE : We aimed to develop a deep learning artificial intelligence (AI) algorithm to detect impacted animal bones on lateral neck radiographs and to assess its effectiveness for improving the interpretation of lateral neck radiographs.

METHODS : Lateral neck radiographs were retrospectively collected for patients with animal bone impaction between January 2010 and March 2020. Radiographs were then separated into training, validation, and testing sets. A total of 1733 lateral neck radiographs were used to develop the deep learning algorithm. The testing set was assessed for the stand-alone deep learning AI algorithm and for human readers (radiologists, radiology residents, emergency physicians, ENT physicians) with and without the aid of the AI algorithm. Another radiograph cohort, collected from April 1, 2020, to June 30, 2020, was analyzed to simulate clinical application by comparing the deep learning AI algorithm with radiologists' reports.

RESULTS : In the testing set, the sensitivity, specificity, and accuracy of the AI model were 96%, 90%, and 93% respectively. Among the human readers, all physicians of different subspecialties achieved a higher accuracy with AI-assisted reading than without. In the simulation set, among the 20 cases positive for animal bones, the AI model accurately identified 3 more cases than the radiologists' reports.

CONCLUSION : Our deep learning AI model demonstrated a higher sensitivity for detection of animal bone impaction on lateral neck radiographs without an increased false positive rate. The application of this model in a clinical setting may effectively reduce time to diagnosis, accelerate workflow, and decrease the use of CT.

Chen Yueh-Sheng, Luo Sheng-Dean, Lee Chi-Hsun, Lin Jian-Feng, Lin Te-Yen, Ko Sheung-Fat, Yu Chiun-Chieh, Chiang Pi-Ling, Wang Cheng-Kang, Chiu I-Min, Huang Yii-Ting, Tai Yi-Fan, Chiang Po-Teng, Lin Wei-Che

2023-Mar-16

Animal bone impaction, Artificial intelligence, Lateral neck radiograph, Retrospective studies

General General

Statistics and Machine Learning in Mass Spectrometry-Based Metabolomics Analysis.

In Methods in molecular biology (Clifton, N.J.)

In this chapter, we review the cutting-edge statistical and machine learning methods for missing value imputation, normalization, and downstream analyses in mass spectrometry metabolomics studies, with illustration by example datasets. The missing peak recovery includes simple imputation by zero or limit of detection, regression-based or distribution-based imputation, and prediction by random forest. The batch effect can be removed by data-driven methods, internal standard-based, and quality control sample-based normalization. We also summarize different types of statistical analysis for metabolomics and clinical outcomes, such as inference on metabolic biomarkers, clustering of metabolomic profiles, metabolite module building, and integrative analysis with transcriptome.

Fan Sili, Wilson Christopher M, Fridley Brooke L, Li Qian

2023

Imputation, Integrative analysis, Mass spectrometry, Metabolomics, Normalization, Statistical and machine learning

General General

Bioinformatic and Statistical Analysis of Microbiome Data.

In Methods in molecular biology (Clifton, N.J.)

Since advances in next-generation sequencing (NGS) technique enabled to investigate uncultured microbiota and their genomes in unbiased manner, many microbiome researches have been reporting strong evidences for close links of microbiome to human health and disease. Bioinformatic and statistical analysis of NGS-based microbiome data are essential components in those microbiome researches to explore the complex composition of microbial community and understand the functions of community members in relation to host and environment. This chapter introduces bioinformatic analysis methods that generate taxonomy and functional feature count table along with phylogenetic tree from raw NGS microbiome data and then introduce statistical methods and machine learning approaches for analyzing the outputs of the bioinformatic analysis to infer the biodiversity of a microbial community and unravel host-microbiome association. Understanding the advantages and limitations of the analysis methods will help readers use the methods correctly in microbiome data analysis and may give a new opportunity to develop new analytic techniques for microbiome research.

Kim Youngchul

2023

16s rRNA sequencing, Alpha diversity, Beta diversity, Metagenomics, Microbiome, Microbiome-wide association, Phylogeny tree

General General

Construction of anoikis-related lncRNAs risk model: Predicts prognosis and immunotherapy response for gastric adenocarcinoma patients.

In Frontiers in pharmacology

Background: Anoikis acts as a programmed cell death that is activated during carcinogenesis to remove undetected cells isolated from ECM. Further anoikis based risk stratification is expected to provide a deeper understanding of stomach adenocarcinoma (STAD) carcinogenesis. Methods: The information of STAD patients were acquired from TCGA dataset. Anoikis-related genes were obtained from the Molecular Signatures Database and Pearson correlation analysis was performed to identify the anoikis-related lncRNAs (ARLs). We performed machine learning algorithms, including Univariate Cox regression and Least Absolute Shrinkage and Selection Operator (Lasso) analyses on the ARLs to build the OS-score and OS-signature. Clinical subgroup analysis, tumor mutation burden (TMB) detection, drug susceptibility analysis, immune infiltration and pathway enrichment analysis were further performed to comprehensive explore the clinical significance. Results: We established a STAD prognostic model based on five ARLs and its prognostic value was verified. Survival analysis showed that the overall survival of high-risk score patients was significantly shorter than that of low-risk score patients. The column diagrams show satisfactory discrimination and calibration. The calibration curve verifies the good agreement between the prediction of the line graph and the actual observation. TIDE analysis and drug sensitivity analysis showed significant differences between different risk groups. Conclusion: The novel prognostic model based on anoikis-related lncRNAs we identified could be used for prognosis prediction and precise therapy in gastric adenocarcinoma.

Li Qinglin, Zhang Huangjie, Hu Jinguo, Zhang Lizhuo, Zhao Aiguang, Feng He

2023

anoikis, gastric adenocarcinoma, immunotherapy, lncRNAs, prognosis prediction

oncology Oncology

Profiling Cellular Ecosystems at Single-Cell Resolution and at Scale with EcoTyper.

In Methods in molecular biology (Clifton, N.J.)

Tissues are composed of diverse cell types and cellular states that organize into distinct ecosystems with specialized functions. EcoTyper is a collection of machine learning tools for the large-scale delineation of cellular ecosystems and their constituent cell states from bulk, single-cell, and spatially resolved gene expression data. In this chapter, we provide a primer on EcoTyper and demonstrate its use for the discovery and recovery of cell states and ecosystems from healthy and diseased tissue specimens.

Steen Chloé B, Luca Bogdan A, Alizadeh Ash A, Gentles Andrew J, Newman Aaron M

2023

Cell states, Ecosystems, Ecotypes, Single-cell RNA sequencing, Spatial transcriptomics, Tissue heterogeneity, Transcriptomics, Tumor microenvironment

General General

Statistical and Machine Learning Methods for Discovering Prognostic Biomarkers for Survival Outcomes.

In Methods in molecular biology (Clifton, N.J.)

Discovering molecular biomarkers for predicting patient survival outcomes is an essential step toward improving prognosis and therapeutic decision-making in the treatment of severe diseases such as cancer. Due to the high-dimensionality nature of omics datasets, statistical methods such as the least absolute shrinkage and selection operator (Lasso) have been widely applied for cancer biomarker discovery. Due to their scalability and demonstrated prediction performance, machine learning methods such as XGBoost and neural network models have also been gaining popularity in the community recently. However, compared to more traditional survival methods such as Kaplan-Meier and Cox regression methods, high-dimensional methods for survival outcomes are still less well known to biomedical researchers. In this chapter, we will discuss the key analytical procedures in employing these methods for identifying biomarkers associated with survival data. We will also identify important considerations that emerged from the analysis of actual omics data. Some typical instances of misapplication and misinterpretation of machine learning methods will also be discussed. Using lung cancer and head and neck cancer datasets as demonstrations, we provide step-by-step instructions and sample R codes for prioritizing prognostic biomarkers.

Yao Sijie, Wang Xuefeng

2023

Cox regression, Elastic net, Gradient boosting, Lasso, Machine learning, Survival analysis