Performance and comparison of artificial intelligence and human experts in the detection and classification of colonic polyps

Table 3 Results of AI/human in diagnosis of polyps

Author	Years	Model type	AI or human	TP	FP	FN	TN
Halligan [24]	2006	CAD	AI	330	55	270	415
Petrick [38]	2008	CAD system	AI	51	44	33	112
Tischendorf [39]	2010	Linear classifier、k-NN、SVM	AI	305	42	15	56
Ignjatovic [7]	2011	NA	AI	274	49	41	266
Gross [17]	2011	SVM	AI	358	28	20	283
Mang [40]	2012	CTC CAD system	AI	580	44	44	212
Mesejo [23]	2016	RF、RS、SVM	AI	52	5	3	16
Mori [41]	2018	SVM	AI	2146	142	164	1272
Renner [42]	2018	CNN	AI	110	42	7	97
Chen [21]	2018	DNN-CAD	AI	181	21	7	75
Shin [43]	2018	SVM	AI	188	8	7	163
Byrne [11]	2019	CNN	AI	65	7	1	33
Cristina [44]	2019	SVM	AI	174	15	18	118
Zachariah [35]	2020	CNN	AI	6443	434	294	3971
Shahidi [45]	2020	CNN	AI	409	168	49	18
Qadir [14]	2020	Faster R-CNN	AI	8171	1166	1854	1347
Halligan [24]	2006	CAD	Expert	239	27	361	443
Petrick [38]	2008	CAD system	Expert	38	23	46	133
Tischendorf [39]	2010	Linear classifier、k-NN、SVM	Expert	305	21	15	77
Ignjatovic [7]	2011	NA	Expert	172	82	38	128
Ignjatovic [7]	2011	NA	Novice	70	44	35	61
Gross [17]	2011	SVM	Expert	699	46	57	576
Gross [17]	2011	SVM	Novice	632	69	124	553
Mesejo [23]	2016	RF、RS、SVM	Expert	46.2	6.7	8.7	14.2
Mesejo [23]	2016	RF、RS、SVM	Novice	49.3	10	5.7	11
Renner [42]	2018	CNN	Expert	188	46	36	216
Chen [21]	2018	DNN-CAD	Expert	367	55	9	137
Chen [21]	2018	DNN-CAD	Novice	671	95	81	289
Cristina [44]	2019	SVM	Expert	187	5	5	128

TP, true positive; FP, false positive; FN, false negative; TN, true negative; CAD, computer-aided diagnosis; K-NN, k-nearest neighbor; RF, random forests; RS, random subspaces; SVM, support vector machine; CNN, convolutional neural network; DNN-CAD, computer-aided diagnosis with a deep neural network

ISSN: 1471-230X