AB057. SOH25_AB_382. Predicting cancer in rectal neoplasia using artificial intelligence and fluorescence angiography: a comparison of methods
General Surgery I

AB057. SOH25_AB_382. Predicting cancer in rectal neoplasia using artificial intelligence and fluorescence angiography: a comparison of methods

Patrick Boland1, Pol Mac Aonghusa1, Ashokkumar Singaravelu1, Philip McEntee1, Ronan Ambrose Cahill1,2

1UCD Centre for Precision Surgery, UCD School of Medicine, UCD Catherine McAuley Education & Research Centre, Dublin, Ireland; 2Department of Colorectal Surgery, Mater Misericordiae University Hospital, Dublin, Ireland

Background: Management of significant rectal polyps and early cancers is often complicated by suboptimal preoperative characterization. The CLASSICA Project aims to validate the use of artificial intelligence (AI) and indocyanine green fluorescence angiography (ICGFA) to characterise rectal neoplasia in situ. Here we compare two methods developed during this project.

Methods: Patients with rectal neoplasia referred for transanal management underwent ICGFA assessment. Video recordings and patient demographics and outcomes were recorded prospectively. Recordings were analysed using boutique tracking and fluorescence extraction software, producing time-series curves. Patients who underwent prior excisions and those with low rectal lesions not visible with a TAMIS port in situ were excluded. Two AI classification methods, a user selected region (USR) of interest method and a method based on intratumoural variability (ITV) of fluorescent signals compared to standard preoperative investigation methods.

Results: In total, 100 cases were enrolled (54 malignant). The USR method achieved unseen testing accuracy sensitivity and specificity of 80%, 88%, and 63% respectively (on a cohort of 89 post-exclusions). The ITV method achieved unseen testing accuracy sensitivity and specificity 80%, 83.33% and 75% respectively (cohort of 88 post exclusions, it was not possible to process one case with this method). Endoscopic biopsy was 65% accurate, 23% sensitive and 100% specific in the same group.

Conclusions: The results of this multicentre European study show that ICGFA and AI methods accurately characterize rectal neoplasia in situ. These methods outperform traditional approaches such as endoscopic biopsy and magnetic resonance imaging (MRI). The results of this multicentre European study show that ICGFA and AI methods accurately characterise rectal neoplasia in situ. These methods outperform traditional approaches such as endoscopic biopsy and MRI.

Keywords: Artificial intelligence (AI); fluorescence; preoperative staging; rectal cancer; rectal neoplasia

Acknowledgments

None.

Footnote

Funding: None.

Conflicts of Interest: The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

doi: 10.21037/map-25-ab057
Cite this abstract as: Boland P, Mac Aonghusa P, Singaravelu A, McEntee P, Cahill RA. AB057. SOH25_AB_382. Predicting cancer in rectal neoplasia using artificial intelligence and fluorescence angiography: a comparison of methods. Mesentery Peritoneum 2025;9:AB057.

Article Options

Download Citation

Share