Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 May 10;17(10):1620.
doi: 10.3390/cancers17101620.

The Association Between Heatmap Position and the Diagnostic Accuracy of Artificial Intelligence for Colorectal Polyp Diagnosis

Ayla Thijssen  1   2 , Nikoo Dehghani  3 , Ruud W M Schrauwen  4 , Eric T P Keulen  5 , Eveline J A Rondagh  5 , Mark H P van Avesaat  5 , Khalida Soufidi  5 , Ankie Reumkens  5 , Paul H A Bours  5 , Quirine E W van der Zander  1   2 , Peter H N de With  3 , Bjorn Winkens  6   7 , Fons van der Sommen  3 , Erik J Schoon  2   8
Affiliations

The Association Between Heatmap Position and the Diagnostic Accuracy of Artificial Intelligence for Colorectal Polyp Diagnosis

Ayla Thijssen et al. Cancers (Basel). .

Abstract

Background/objectives: Artificial intelligence (AI) algorithms for diagnosing colorectal polyps are emerging but not yet widely used. Trust in AI is lacking and could be improved by visually explainable AI, such as heatmaps. This study aims to investigate the association between heatmap position and AI accuracy for the endoscopic characterization of colorectal polyps.

Methods: Four AI algorithms diagnosed 2133 prospectively collected images of 376 colorectal polyps from two hospitals, using histopathology as the gold standard. Heatmap position was compared to the human-annotated polyp position. Generalized estimating equations were used to assess the association between heatmap position and a correct AI diagnosis.

Results: Higher percentages of heatmap covering the colorectal polyp were associated with correct diagnoses in all four algorithms (OR 1.013 [95% CI 1.006-1.019], OR 1.025 [95% CI 1.011-1.039], OR 1.038 [95% CI 1.024-1.053], and OR 1.039 [95% CI 1.020-1.058]-all p < 0.001). A higher percentage of polyp not covered by heatmap was associated with a correct diagnosis of Algorithm 1 (OR 1.006 [95% CI 1.003-1.010], p < 0.001), while in Algorithm 2, a lower percentage was associated with a correct diagnosis (OR 0.992 [95% CI 0.985-1.000], p 0.044). Algorithms 3 and 4 showed negative, but not statistically significant, associations.

Conclusions: Higher percentages of heatmap covering the polyp were associated with correct diagnoses of four AI algorithms. This indicates that it is clinically relevant to strive for AI predictions with heatmaps covering as much colorectal polyp tissue as possible. Knowing how to interpret heatmaps could increase trust in AI and, with that, benefit the implementation of AI in clinical practice.

Keywords: colonoscopy; colorectal polyps; computer-aided diagnosis; visually explainable artificial intelligence.

PubMed Disclaimer

Conflict of interest statement

F.v.d.S. received research support from Olympus outside the submitted work. E.S. received research support and speaker fees from Fujifilm outside the submitted work. Q.v.d.Z. was supported by Fujifilm Inc. to attend scientific meetings outside the submitted work. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Figures

Figure A1
Figure A1
Examples of heatmap coverage from (A,C,E) Algorithm 1 and (B,D,F) Algorithm 2 in correctly classified cases.
Figure 1
Figure 1
Example of human-annotated colorectal polyp position, showing (A) a labeled polyp and (B) the corresponding mask.
Figure 2
Figure 2
Example of (A) a training image and (B) the selected region of interest (ROI) from the central area of this image.
Figure 3
Figure 3
Example of a colorectal polyp image with a heatmap, showing (A) the heatmap on the polyp, (B) the part of the heatmap covering polyp in green, the part of the heatmap covering tissue surrounding the polyp in red, and the part of the polyp not covered by the heatmap in orange (3B).
Figure 4
Figure 4
Examples of different heatmap coverage: (A) a heatmap covering only colorectal polyp tissue but missing part of the polyp, (B) a heatmap covering little colorectal polyp and much tissue surrounding the colorectal polyp, (C) a heatmap covering only colorectal polyp and little tissue surrounding the colorectal polyp, and (D) a heatmap covering the colorectal polyp tissue and covering much tissue surrounding the colorectal polyp.
See this image and copyright information in PMC

References

    1. Hassan C., Balsamo G., Lorenzetti R., Zullo A., Antonelli G. Artificial Intelligence Allows Leaving-In-Situ Colorectal Polyps. Clin. Gastroenterol. Hepatol. 2022;20:2505–2513.e2504. doi: 10.1016/j.cgh.2022.04.045. - DOI - PubMed
    1. Kader R., Cid-Mejias A., Brandao P., Islam S., Hebbar S., Puyal J.G., Ahmad O.F., Hussein M., Toth D., Mountney P., et al. Polyp characterization using deep learning and a publicly accessible polyp video database. Dig. Endosc. 2023;35:645–655. doi: 10.1111/den.14500. - DOI - PMC - PubMed
    1. Houwen B., Hassan C., Coupe V.M.H., Greuter M.J.E., Hazewinkel Y., Vleugels J.L.A., Antonelli G., Bustamante-Balen M., Coron E., Cortas G.A., et al. Definition of competence standards for optical diagnosis of diminutive colorectal polyps: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy. 2022;54:88–99. doi: 10.1055/a-1689-5130. - DOI - PubMed
    1. Rex D.K., Kahi C., O’Brien M., Levin T.R., Pohl H., Rastogi A., Burgart L., Imperiale T., Ladabaum U., Cohen J., et al. The American Society for Gastrointestinal Endoscopy PIVI (Preservation and Incorporation of Valuable Endoscopic Innovations) on real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest. Endosc. 2011;73:419–422. doi: 10.1016/j.gie.2011.01.023. - DOI - PubMed
    1. Schulz P.J., Lwin M.O., Kee K.M., Goh W.W.B., Lam T.Y.T., Sung J.J.Y. Modeling the influence of attitudes, trust, and beliefs on endoscopists’ acceptance of artificial intelligence applications in medical practice. Front. Public. Health. 2023;11:1301563. doi: 10.3389/fpubh.2023.1301563. - DOI - PMC - PubMed

Grants and funding

LinkOut - more resources