Diagnostic accuracy for eosinophilic esophagitis (EoE) was higher when white light images were used with linked color imaging rather than alone, suggesting that linked color imaging may be used to reduce diagnostic disparities among endoscopists.
Researchers suggest that linked color imaging, a recently developed image-enhanced endoscopy technique, could lead to reduced diagnostic disparities among endoscopists.
Linked color imaging provides clear images using preprocessing technology with radiation of short-wavelength laser light and postprocessing technology with the enhancement of color differences. This process allows for red areas to appear redder and white areas to appear whiter. Compared with white light images, linked color imaging maintains a brighter field of endoscopic view.
Researchers conducted a preliminary retrospective study to investigate whether endoscopic observation using additional linked color imaging improved the accuracy of endoscopic diagnosis of EoE compared with white light images alone.
They collected 30 white light images and 30 white light images with linked color imaging from patients with and without EoE. These images were randomly and blindly reviewed by 10 endoscopists, including 4 experts and 6 nonexperts.
They rated edema, ring, exudate furrows, and strictures on the adjusted EoE endoscopic reference score. The diagnosis of EoE was also assessed for accuracy.
Researchers found that diagnostic accuracy for EoE was higher for white light imaging and linked color imaging than for white light imaging alone (0.85 vs 0.70, respectively).
The findings demonstrated that linked color imaging was useful for the diagnosis of EoE, particularly for endoscopists with no experience with EoE and for cases with milder endoscopic abnormalities.
Researchers expect that linked color imaging will contribute to the screening and diagnosis of EoE in clinical practice where endoscopy is not always performed by expert endoscopists.
The results of inter- and intra-observer agreements indicated that the additional effect of linked color imaging on white light images could be mainly attributed to the improvement of diagnostic consistency in furrows, strictures, and rings, especially in diffuse types of EoE.
The inter-observer agreement for white light images with linked color imaging statistically surpassed white light images alone for furrows (kappa = 0.73 vs 0.67, respectively; P = .0013), stricture (kappa = 0.51 vs 0.39, respectively; P = .0072), and diagnosis (kappa = 0.67 vs 0.57, respectively; P < .0001) of EoE.
Additionally, the intra-observer agreement for white light images with linked color imaging surpassed white light images alone for a ring (kappa = 0.62 vs 0.43; P = .0052), and a similar trend was found in exudates, furrows, and diagnosis regardless of expert or nonexpert endoscopists.
However, diagnostic ability was much lower when abnormal findings were localized to a small area at the lower end of the esophagus compared with the diffuse type. No diagnostic improvement was found with the addition of linked color imaging.
For the localized type, the inter-observer agreement in edema and intra-observer agreement in stricture were found to be significantly higher in white light imaging than in white light imaging with linked color imaging. The authors note that a lack of images with the localized type might affect these results.
They also suggest that the advantages of linked color imaging might not be observed in images inherently capturing closer areas of the localized type than images of the diffuse type. Additionally, the degree of endoscopic abnormality in the localized type was originally less severe than that in the diffuse type, which may influence the results.
The additional effect of linked color imaging on the diagnostic consistency of EoE was confirmed to be statistically significant even when the localized type was excluded from the analysis.
Overall, the findings of this study indicate that linked color imaging may yield detailed surface information up to a more distant area of the esophagus and result in improved diagnostic ability and consistency for EoE compared with other imaging methods.
The study had some limitations. Given that linked color imaging is created by preprocessing and postprocessing technology, it was impossible for researchers to use the images of white light imaging and linked color imaging taken under the same conditions.
However, researchers prepared images with as many similar compositions as possible. Additionally, the study was designed to compare the difference between white light imaging alone vs white light imaging with linked color imaging, rather than comparing white light imaging alone against linked color imaging alone.
Abe Y, Sasaki Y, Yagi M, et al. Linked color imaging improves the diagnostic accuracy of eosinophilic esophagitis. DEN Open. 2023;3(1):e146. doi:10.1002/deo2.146