From: Data-driven decision-making for precision diagnosis of digestive diseases
First author, year | Disease | n | Data source and specific task | ML method | Diagnostic performance | Refs. |
---|---|---|---|---|---|---|
Loomba, 2017 | NAFLD | 86 | Gut metagenomics data of stool; classification based on a fecal metagenomic signature | RF/SVM/clustering | AUC: 0.936 | [153] |
Yang, 2020 | CRC | 534 | Fecal metagenomics data; classification based on fecal microbiomics biomarkers | Clustering/RF | AUC: 0.811–0.930 | [154] |
Bang, 2019 | CRC | 404 | Gut microbiome data from 16 S rRNA sequencing; classification based on gut microbiome | SVM/KNN/LogitBoost | Accuracy: 96.84% | [155] |
Dai, 2018 | CRC | 526 | Gut metagenomics data; classification based on seven CRC-enriched bacterial markers | PCA/SVM | AUC: 0.820–0.84 | [156] |
Abbas, 2019 | IBD | 973 | Gut metagenomics data of biopsy samples from QIITA database; classification based selected features by NBBD | RF | AUC: 0.760–0.800 | [157] |
Syama, 2023 | CRC/IBD | 1849 | Gut metagenomics data sets of CRC and IBD; classification based on gut metagenomics data by boosting GraphSAGE | GCN | AUC: 0.900–0.930 | [158] |
Lee, 2022 | IBD/CRC/LC | 644 | Gut metagenomics data sets; classification based on metagenome features | RF/SVM/PCR/LASSO/XGBoost/ | AUC: 0.840–0.980 | [159] |
Forbes, 2018 | UC | 102 | Gut metagenomics data; classification based on abundant taxonomic biomarkers of gut microbiota | Naive Bayes/RF/PCA | AUC: 0.900–0.930 | [160] |
Liang, 2020 | CRC | 1012 | Fecal metagenomics data; classification based on combining several gut microbial gene markers with FIT | LR | Sensitivity: 93.8% Specificity: 81.2% | [161] |
Hollister, 2019 | IBS | 45 | Fecal metagenomics and metabolomics data; classification based on fecal metagenomic and metabolic markers | RF/LASSO/SVM/Naive Bayes | AUC: 0.930 | [162] |