Table 1 Publications of studies that met the inclusion criteria, task(s) performed by each model, external validation dataset details and algorithm details for each study
Publication Author and Date | Publication Title | Model Task | External Validation Dataset Details | Algorithm Details | |||
|---|---|---|---|---|---|---|---|
Number of samples | Public dataset, or restricted dataset and name of public dataset(s) if used | Number of centres that restricted datasets were taken from and country | Algorithm name | Code availability | |||
Bilaloglu et al. 2019a17 | Efficient pan-cancer whole-slide image classification and outlier detection using convolutional neural networks | •Subtyping •Classification of malignant versus non-malignant tissue | 340 | Restricted | One academic medical centre in the US | PathCNN | |
Borras Ferris et al. 202432 | A full pipeline to analyse lung histopathology images | •Subtyping •Classification of malignant versus non-malignant tissue | 1036 | Public TCGA | Unavailable | ||
Cao et al. 202318 | E2EFP-MIL: End-to-end and high-generalizability weakly supervised deep convolutional network for lung cancer classification from whole slide image | •Subtyping | 1583 | Restricted | Three hospitals in China | E2EFP-MIL | |
Chen et al. 202219 | A whole-slide image (WSI)-based immunohistochemical feature prediction system improves the subtyping of lung cancer | •Subtyping •Classification of malignant versus non-malignant tissue •Biomarker identification | 299 | Restricted | One academic medical centre in China | WIFPS | |
Coudray et al. 201820 | Classification and mutation prediction from non–small cell lung cancer histopathology images using deep learning | •Subtyping •Classification of malignant versus non-malignant tissue | 340 | Restricted | One academic medical centre in the US | ||
Gertych et al. 201933 | Convolutional neural networks can accurately distinguish four histologic growth pattens of lung adenocarcinoma in digital slides | •Classification of malignant versus non-malignant tissue •Classification of tumour growth pattern | 27 | Public TCGA | https://github.com/zhaoxuanma/Deeplearning-digital-pathology | ||
Hari et al. 2021a21 | Examining batch effect in histopathology as a distributionally robust optimisation problem | •Subtyping •Classification of malignant versus non-malignant tissue | Unclear | Public TCGA | Four academic medical centres, one cancer centre and one hospital in the US | ERM model | Unavailable |
Kanavati et al. 202034 | Weakly-supervised learning for lung carcinoma classification using deep learning | •Classification of malignant versus non-malignant tissue | 1670 Public: 1170 Restricted: 500 | Mixed TCGA, TCIA | One hospital in Japan | Unavailable | |
Kanavati et al. 202122 | A deep learning model for the classification of indeterminate lung carcinoma in biopsy whole slide images | •Subtyping •Classification of malignant versus non-malignant tissue | 1405 Public: 905 Restricted: 500 | Mixed TCGA | One hospital in Japan | Unavailable | |
Le Page et al. 202123 | Using a convolutional neural network for classification of squamous and non-squamous non-small cell lung cancer based on diagnostic histopathology HES images | •Subtyping | 65 | Restricted | One academic medical centre in France | Available from author upon reasonable request | |
Lu et al. 202124 | Data Efficient and Weakly Supervised Computational Pathology on Whole Slide Images | •Subtyping | 241 | Restricted | One hospital in the US | CLAM | |
Mukashyaka et al. 202425 | SAMPLER: unsupervised representations for rapid analysis of whole slide tissue images | •Subtyping | Unclear | Public TCGA, CPTAC | SAMPLER | https://figshare.com/articles/software/SAMPLER_basic_code_example/23713404 | |
Noorbakhsh et al. 202026 | Deep learning-based cross-classifications reveal conserved spatial behaviours within tumour histological images | •Subtyping •Classification of malignant versus non-malignant tissue | 2115 | Public CPTAC | |||
Quiros et al. 202427 | Mapping the landscape of histomorphological cancer phenotypes using self-supervised learning on unannotated pathology slides | •Subtyping | 138 | Restricted | One academic medical centre in the US | PRL | https://github.com/AdalbertoCq/Histomorphological-Phenotype-Learning |
Sakamoto et al. 202235 | A collaborative workflow between pathologists and deep learning for the evaluation of tumour cellularity in lung adenocarcinoma | •Classification of malignant versus non-malignant tissue •Prediction of tumour cellularity | 125 | Restricted | Two hospitals in Japan | Available from author upon reasonable request | |
Sharma et al. 202431 | Optimizing Knowledge Transfer in Sequential Models: Leveraging Residual Connections in Flow Transfer Learning for Lung Cancer Classification | •Classification of malignant versus non-malignant tissue •Subtyping | 566 | Public TCGA | Unavailable | ||
Swiderska-Chadaj et al. 202036 | A deep learning approach to assess the predominant tumour growth pattern in whole-slide images of lung adenocarcinoma | •Classification of malignant versus non-malignant tissue •Classification of tumour growth pattern | 20 | Restricted | One academic medical centre in the Netherlands | DenseNet | Unavailable |
Vorontsov et al. 202416 | A foundation model for clinical-grade computational pathology and rare cancers detection | •Classification of malignant versus non-malignant tissue •Biomarker identification | Unclear | Restricted | Unknown | Virchow | |
Wang et al. 201937 | ConvPath: A software tool for lung adenocarcinoma digital pathological image analysis aided by a convolutional neural network | •Classification of cell types | 130 | Restricted | Unknown | ConvPath | |
Wang et al. 202328 | Deep Learning of Cell Spatial Organizations Identifies Clinically Relevant Insights in Tissue Images | •Subtyping | 496 | Restricted | Unknown | Ceograph | |
Yang et al. 202129 | Deep learning-based six-type classifier for lung cancer and mimics from histopathological whole slide images: a retrospective study | •Subtyping •Classification of malignant versus non-malignant tissue | 634 Public: 422 Restricted: 212 | Mixed TCGA | One hospital in China | Available from author upon reasonable request | |
Yu et al. 202030 | Classifying non-small cell lung cancer histopathology types and transcriptomic subtypes using convolutional neural networks | •Subtyping •Classification of malignant versus non-malignant tissue | 125 | Public ICGC | |||
