Dataset Information

Assessing artificial intelligence enabled liquid-based cytology for triaging HPV-positive women: a population-based cross-sectional study.

ABSTRACT:

Introduction

Cytology-based triaging is commonly used to manage the care of women with positive human papillomavirus (HPV) results, but it suffers from subjectivity and a lack of sensitivity and reproducibility. The diagnostic performance of an artificial intelligence-enabled liquid-based cytology (AI-LBC) triage approach remains unclear. Here, we compared the clinical performance of AI-LBC, human cytologists and HPV16/18 genotyping at triaging HPV-positive women.

Material and methods

HPV-positive women were triaged using AI-LBC, human cytologists and HPV16/18 genotyping. Histologically confirmed cervical intraepithelial neoplasia grade 2/3 or higher (CIN2+/CIN3+) were accepted as thresholds for clinical performance assessments.

Results

Of the 3514 women included, 13.9% (n = 489) were HPV-positive. The sensitivity of AI-LBC was comparable to that of cytologists (86.49% vs 83.78%, P = 0.744) but substantially higher than HPV16/18 typing at detecting CIN2+ (86.49% vs 54.05%, P = 0.002). While the specificity of AI-LBC was significantly lower than HPV16/18 typing (51.33% vs 87.17%, P < 0.001), it was significantly higher than cytologists at detecting CIN2+ (51.33% vs 40.93%, P < 0.001). AI-LBC reduced referrals to colposcopy by approximately 10%, compared with cytologists (51.53% vs 60.94%, P = 0.003). Similar patterns were also observed for CIN3+.

Conclusions

AI-LBC has equivalent sensitivity and higher specificity compared with cytologists, with more efficient colposcopy referrals for HPV-positive women. AI-LBC could be particularly useful in regions where experienced cytologists are few in number. Further investigations are needed to determine triaging performance through prospective designs.

SUBMITTER: Xue P

PROVIDER: S-EPMC10377999 | biostudies-literature | 2023 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Assessing artificial intelligence enabled liquid-based cytology for triaging HPV-positive women: a population-based cross-sectional study.

Xue Peng P Xu Hai-Miao HM Tang Hong-Ping HP Wu Wen-Qing WQ Seery Samuel S Han Xiao X Ye Hu H Jiang Yu Y Qiao You-Lin YL

Acta obstetricia et gynecologica Scandinavica 20230615 8

<h4>Introduction</h4>Cytology-based triaging is commonly used to manage the care of women with positive human papillomavirus (HPV) results, but it suffers from subjectivity and a lack of sensitivity and reproducibility. The diagnostic performance of an artificial intelligence-enabled liquid-based cytology (AI-LBC) triage approach remains unclear. Here, we compared the clinical performance of AI-LBC, human cytologists and HPV16/18 genotyping at triaging HPV-positive women.<h4>Material and methods ...[more]

PMID: 37318036

Similar Datasets

Project description:BackgroundSelf-sampling for human papillomavirus (HPV) testing is a feasible option to improve the cervical screening coverage. However, an ideal triage method for HPV-positive self-samples does not yet exist. The aim of this study was to explore the utility of HPV genotyping and p16INK4a immunostaining (p16) in triaging HPV-positive self-samples, focusing on HPV-positive, cytology-negative (HPCN) women.MethodsA total of 73,699 women were screened in a cervical screening project in China via SeqHPV assay on self-samples. HPV-positive women were called-back and collected cervical sample for p16 immunostaining and liquid-based cytology, those who met any result of HPV16/18+ or visual inspection with acetic acid (VIA) + or p16+ were referred for colposcopy, and HPCN women with adequate data on p16 and pathology were analyzed. A triage strategy was considered acceptable if the negative predictive value (NPV) for cervical intraepithelial neoplasia 3 or worse (CIN3+) was 98% or more, combined with an improvement of sensitivity and specificity for CIN2+/CIN3+ in reference to the comparator, being HPV16/18 + .ResultsA total of 2731 HPCN women aged 30-64 years were enrolled, 136 (5.0%) CIN2+ and 53 (1.9%) CIN3+ were detected. Five triage strategies met the criteria: p16+; HPV16/33+; 'HPV16+ or HPV33/58/31/35+&p16+'; 'HPV16/33+ or HPV58/31/35+&p16+'; HPV16/18/31/33/45/52/58 + & p16+. These strategies required less or similar colposcopy referrals, and less colposcopies to detected one case of CIN2+/CIN3+, achieving favorable false positive (negative) rates to the comparator. Among them, p16 staining detected 83.1% (79.2%) of underlying CIN2 + (CIN3+) in HPCN women. Moreover, three triage strategies were favorable in sensitivity and/or specificity to the 'HPV16/33+' strategy: p16+; 'HPV16+ or HPV33/58/31/35 + &p16+'; HPV16/18/31/33/45/52/58 + &p16 + .ConclusionsGenotyping for HPV16/33 could be utilized to optimize the management of HPCN women. Moreover, p16 immunostaining, either alone or combined with extended genotypes, is more effective than HPV genotypes alone in the triage of HPCN women.

Project description:BackgroundLaboratory detection of Human papillomavirus (HPV), Chlamydia trachomatis and Neisseria gonorrhoeae in liquid-based cervicovaginal cytology specimens is now based on identification of the DNA sequences unique to these infectious agents. However, current commercial test kits rely on nucleotide probe hybridization to determine DNA sequences, which may lead to diagnostic errors due to cross-reactivity. The aim of this study was to find a practical approach to perform automated Sanger DNA sequencing in clinical laboratories for validation of the DNA tests for these three infectious agents.MethodsA crude proteinase K digest of 5% of the cells collected in a liquid-based cervicovaginal cytology specimen was used for the detection of DNA molecules specific for HPV, C trachomatis and N gonorrhoeae, and for preparation of materials suitable for direct automated DNA sequencing. Several sets of commercially available polymerase chain reaction (PCR) primers were used to prepare nested PCR amplicons for direct DNA sequencing.ResultsSome variants of HPV-16 and HPV-31 were found to share an at least 34-base long sequence homology downstream of the GP5+ binding site, and all HPV-6 and HPV-11 variants shared an upstream 34-base sequence including part of the GP5+ primer. Accurate HPV genotyping frequently required more than 34-bases for sequence alignments to distinguish some of the HPV genotype variants with closely related sequences in this L1 gene hypervariable region. Using the automated Sanger DNA sequencing method for parallel comparative studies on split samples and to retest the residues of samples previously tested positive for C trachomatis and/or for N gonorrhoeae, we also found false-negative and false-positive results as reported by two commercial nucleic acid test kits.ConclusionIdentification of a signature DNA sequence by the automated Sanger method is useful for validation of HPV genotyping and for molecular testing of C trachomatis and N gonorrhoeae in liquid-based cervicovaginal Papanicolaou (Pap) cytology specimens for clinical laboratories with experience in molecular biology to increase the specificity of these DNA-based tests. However, even a highly specific test for high-risk HPV genotyping may have unacceptably low positive predictive values for precancer lesion in populations with a low cervical cancer prevalence rate.

Project description:BackgroundHealthcare is expected to increasingly integrate technologies enabled by artificial intelligence (AI) into patient care. Understanding perceptions of these tools is essential to successful development and adoption. This exploratory study gauged participants' level of openness, concern, and perceived benefit associated with AI-driven healthcare technologies. We also explored socio-demographic, health-related, and psychosocial correlates of these perceptions.MethodsWe developed a measure depicting six AI-driven technologies that either diagnose, predict, or suggest treatment. We administered the measure via an online survey to adults (N = 936) in the United States using MTurk, a crowdsourcing platform. Participants indicated their level of openness to using the AI technology in the healthcare scenario. Items reflecting potential concerns and benefits associated with each technology accompanied the scenarios. Participants rated the extent that the statements of concerns and benefits influenced their perception of favorability toward the technology. Participants completed measures of socio-demographics, health variables, and psychosocial variables such as trust in the healthcare system and trust in technology. Exploratory and confirmatory factor analyses of the concern and benefit items identified two factors representing overall level of concern and perceived benefit. Descriptive analyses examined levels of openness, concern, and perceived benefit. Correlational analyses explored associations of socio-demographic, health, and psychosocial variables with openness, concern, and benefit scores while multivariable regression models examined these relationships concurrently.ResultsParticipants were moderately open to AI-driven healthcare technologies (M = 3.1/5.0 ± 0.9), but there was variation depending on the type of application, and the statements of concerns and benefits swayed views. Trust in the healthcare system and trust in technology were the strongest, most consistent correlates of openness, concern, and perceived benefit. Most other socio-demographic, health-related, and psychosocial variables were less strongly, or not, associated, but multivariable models indicated some personality characteristics (e.g., conscientiousness and agreeableness) and socio-demographics (e.g., full-time employment, age, sex, and race) were modestly related to perceptions.ConclusionsParticipants' openness appears tenuous, suggesting early promotion strategies and experiences with novel AI technologies may strongly influence views, especially if implementation of AI technologies increases or undermines trust. The exploratory nature of these findings warrants additional research.

Dataset Information

Assessing artificial intelligence enabled liquid-based cytology for triaging HPV-positive women: a population-based cross-sectional study.

Introduction

Material and methods

Results

Conclusions

Publications

Assessing artificial intelligence enabled liquid-based cytology for triaging HPV-positive women: a population-based cross-sectional study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets