Exploration of Digital Health Technologies

Table 1

Artificial intelligence (AI) model for histological grading and prognostic identification of various epithelial dysplasia

Organ	Purpose	WSI scanning system	Magnification	Patches	AI model	Model performance	Reference
Oral cavity	Histological grading	Aperio	20×	299 × 299 pixels	VGG16	Testing AUC: 0.65	Araújo et al. [16], Brazil (2023)
	Prognostic identification	Nanozoomer	20×	512 × 512 pixels	ResNet50 and lightGBM	Testing AUC: 0.81 (95% CI, 0.73–0.90)	Cai et al. [3], China (2023)
		Aperio, Nanozoomer	20×, 40×	512 × 512 pixels	IDaRS	Testing AUC: 0.78	Bashir et al. [12], UK (2023)
		Aperio, Nanozoomer	10×	NA	NuClick (for cell detection)	Testing AUC: 0.76 (95% CI, 0.68–0.85)	Mahmood et al. [11], UK (2023)
Larynx	Histological grading	Nanozoomer	20×	224 × 224 pixels	DenseNet121	Testing AUC: 0.89 (95% CI, 0.81–0.95)	Lubrano et al. [2], France (2024)
Esophagus	Histological grading	Mirax Desk	20×	224 × 224 pixels	ResNet50	Testing AUC: 0.80	Beuque et al. [9], Netherlands (2021)
Esophagus	Histological grading	Aperio	40×	1,280 × 1,280 pixels	YOLOv5 and ResNet101	0.89 accuracy for 3-class, 0.96 accuracy for 2-class	Faghani et al. [5], USA (2022)
Stomach	Histological grading	Iscan Coreo	20×	320 × 320 pixels	ResNet50 and domain adaption	Testing AUC: 0.82	Shi et al. [4], China (2022)
Colorectum	Histological grading	AT2	20×	224 × 224 pixels	ResNet18	Testing AUC: 0.97 (95% CI, 0.95–0.99)	Kim et al. [15], USA (2023)
Cervix	Histological grading	PloidyScanner	20×	224 × 224 pixels	VGG16	Testing AUC: 0.76 (95% CI, 0.73–0.78)	Bao et al. [8], China (2020)

Return to article view

WSI: whole slide image; lightGBM: light gradient boosting machine; IDaRS: iterative draw-and-rank sampling; AUC: area under the receiver operating characteristic curve; 95% CI: lower and upper values of the 95% confidence interval; NA: not applicable; VGG16: visual geometry group 16; ResNet50: residual neural network 50; DenseNet121: dense convolutional network 121; YOLOv5: You Only Look Once version 5