Artificial intelligence for characterization of diminutive colorectal polyps: A feasibility study comparing two computer-aided diagnosis systems

Table 1 Baseline characteristics of patients, n (%)

		Patients, n = 30
Gender, female		13 (43.3)
Age in years, mean (SD) [range]		65.8 (8.4) [50-78]
Indication colonoscopy
	Bowel cancer screening program	15 (50.0)
	Surveillance	10 (33.3)
	Symptoms	5 (16.7)
Family history positive for CRC		5 (16.7)
BBPS, mean (SD)		6.6 (1.4)
Number of colorectal polyps per patient1
	1 colorectal polyp	15 (50.0)
	2 colorectal polyps	9 (30.0)
	3 colorectal polyps	6 (20.0)

¹A maximum of three colorectal polyps were included per patient.

BBPS: Boston bowel preparation scale; CRC: colorectal cancer.

Table 2 Baseline characteristics for colorectal polyps, n (%)

		Colorectal polyps, n = 51
Location
	Cecum	7 (13.7)
	Ascending colon	8 (15.7)
	Transverse colon	15 (29.4)
	Descending colon	5 (9.8)
	Sigmoid	10 (19.6)
	Rectum	6 (11.8)
Size, mean (SD) [range]		2.8 (1.0) [2-5]
Morphology
	Sessile (Paris Is)	45 (88.2)
	Flat-elevated (Paris IIa)	6 (11.8)
Histopathology
	Tubular adenoma, LGD	32 (62.7)
	Tubulovillous adenoma, LGD	1 (2.0)
	Sessile serrated lesion, no dysplasia	6 (11.8)
	Hyperplastic polyp, no dysplasia	12 (23.5)
Resection technique – cold snare		51 (100.0)

LGD: Low grade dysplasia.

Table 3 Diagnostic performances of artificial intelligence for ColoRectal polyps in different image enhancement modes

	AI4CRP (n = 51)
	BLI, % (95%CI)	HDWL, % (95%CI)	LCI, % (95%CI)	Multimodal imaging, % (95%CI)
Sensitivity	82.1 (0.66-0.92)	59.0 (0.42-0.74)	76.9 (0.60-0.88)	71.8 (0.55-0.84)
Specificity	75.0 (0.43-0.93)	91.7 (0.60-1.00)	83.3 (0.51-0.97)	91.7 (0.60-1.00)
PPV	91.4 (0.76-0.98)	95.8 (0.77-1.00)	93.8 (0.78-0.99)	96.6 (0.80-1.00)
NPV	56.3 (0.31-0.79)	40.7 (0.23-0.61)	52.6 (0.29-0.75)	50.0 (0.29-0.71)
Diagnostic accuracy	80.4 (0.66-0.90)	66.7 (0.52-0.79)	78.4 (0.64-0.88)	76.5 (0.62-0.87)

AI4CRP: Artificial intelligence for ColoRectal polyps; BLI: Blue light imaging; CI: Confidence interval; HDWL: High definition white light; LCI: Linked color imaging; NPV: Negative predictive value; PPV: Positive predictive value.

Table 4 Diagnostic performance of artificial intelligence for ColoRectal polyps, self-critical artificial intelligence for ColoRectal polyps, CAD EYE, and the endoscopist

	AI4CRP1, % (95%CI), n = 51	Self-critical AI4CRP1, % (95%CI), n = 37	CAD EYE, % (95%CI), n = 49	Endoscopist alone2, % (95%CI), n = 47	AI-assisted endoscopist2,3, % (95%CI), n = 49
Sensitivity	82.1 (0.66-0.92)	89.7 (0.72-0.97)	74.2 (0.55-0.87)	97.4 (0.85-1.00)	97.4 (0.85-1.00)
Specificity	75.0 (0.43-0.93)	87.5 (0.47-0.99)	100.0 (0.78-1.00)	77.8 (0.40-0.96)	90.9 (0.57-1.00)
PPV	91.4 (0.76-0.98)	96.3 (0.79-1.00)	100.0 (0.82-1.00)	94.9 (0.81-0.99)	97.4 (0.85-1.00)
NPV	56.3 (0.31-0.79)	70.0 (0.35-0.92)	69.2 (0.48-0.85)	87.5 (0.47-0.99)	90.9 (0.57-1.00)
Diagnostic accuracy	80.4 (0.66-0.90)	89.2 (0.74-0.96)	83.7 (0.70-0.92)	93.6 (0.81-0.98)	95.9 (0.85-0.99)

¹AI4CRP and self-critical AI4CRP both in BLI mode.

²Optical diagnosis by the endoscopist only taking into account diagnoses made with high confidence.

³Optical diagnosis by the endoscopist after reviewing predictions of both AI4CRP and CAD EYE.

AI4CRP: Artificial intelligence for ColoRectal polyps; AI: Artificial intelligence; BLI: Blue light imaging; CI: Confidence interval; NPV: Negative predictive value; PPV: Positive predictive value.