The Performance of Artificial Intelligence in Cervical Colposcopy: A Retrospective Data Analysis

Table 1

Clinical results of the enrolled medical records.

	Total (N)	Histopathology diagnosis
		Normal, n (%)		CIN1, n (%)		CIN2, n (%)		CIN3, n (%)		Cancer, n (%)
Cytology
NILM	132	89	67.4	38	28.8	1	0.8	1	0.8	3	2.3
ASC-US	98	61	62.2	29	29.6	2	2.0	2	2.0	4	4.1
AGC	6	2	33.3	1	16.7	0	0.0	0	0.0	3	50.0
LSIL	32	10	31.3	16	50.0	2	6.3	2	6.3	2	6.3
ASC-H	9	3	33.3	5	55.6	1	11.1	0	0.0	0	0.0
HSIL	17	2	11.8	1	5.9	2	11.8	4	23.5	8	47.1
SCC	52	0	0.0	0	0.0	0	0.0	1	1.9	51	98.1
HPV test
Negative	113	72	63.7	33	29.2	3	2.7	1	0.9	4	3.5
HPV 16/18 positive	129	43	33.3	23	17.8	4	3.1	5	3.9	54	41.9
Other high-risk subtypes positive	99	48	48.5	33	33.3	1	1.0	4	4.0	13	13.1
Colposcopy findings by the human colposcopists
Normal	163	107	65.6	46	28.2	4	2.5	2	1.2	4	2.5
LSIL	89	51	57.3	29	32.6	2	2.2	3	3.4	4	4.5
HSIL	34	8	23.5	14	41.2	1	2.9	2	5.9	9	26.5
Cancer	60	1	1.7	1	1.7	1	1.7	3	5.0	54	90.0
AI system findings
Negative	152	121	79.6	27	17.8	2	1.3	0	0.0	2	1.3
Positive	194	46	23.7	63	32.5	6	3.1	10	5.2	69	35.6
AI-assisted finding
Negative	101	85	84.2	13	12.9	1	1.0	0	0.0	2	2.0
Positive	245	82	33.5	77	31.4	7	2.9	10	4.1	69	28.2
Total	346	167	48.3	90	26.0	8	2.3	10	2.9	71	20.5

A positive finding by the AI system indicated the finding of the images was classified as positive by the AI system alone. A positive finding by the AI-assisted finding indicated the finding of the images was classified as positive either of the AI system or human colposcopists or both. AI, artificial intelligence; ASC-US, atypical squamous cells of undetermined significance; ASC-H, atypical squamous cells that cannot exclude high-grade squamous intraepithelial lesion; AGC, atypical glandular cells; CIN, cervical intraepithelial neoplasia; HPV, human papillomavirus; HSIL, high-grade squamous intraepithelial lesion; LSIL, low-grade squamous intraepithelial lesion; NILM, negative for intraepithelial lesion or malignancy; SCC, squamous cell carcinoma.