We read with considerable interest the article by Akazawa et al. evaluating an automated microvascular blood flow rate (MBFR) system for distinguishing early gastric cancer (EGC) from patchy redness 1. This innovative approach introduces an objective physiological parameter into magnifying endoscopy and deserves recognition. Nevertheless, we wish to highlight a key issue affecting its real-world clinical applicability. All videos used for algorithm development and validation were obtained under optimal imaging conditions—maximal magnification, water immersion, and stable visualization of subepithelial microvessels—performed by experienced operators. In routine endoscopic practice, such ideal conditions are not consistently achievable. Many lesions, particularly in post-Helicobacter pylori eradication stomachs, present with subtle morphological changes and unstable microvascular views, making it difficult to capture reproducible flow signals. Prior research has shown that the diagnostic performance of magnifying endoscopy decreases significantly when image quality or operator experience is suboptimal 2, 3. Because the authors excluded lesions without clearly visible microvessels, the study population may represent a selected subset with better visualization than typically encountered in practice. This limitation is noteworthy because the excluded lesions often represent the most diagnostically challenging cases, which are precisely where an objective tool like MBFR would be most valuable. Including unselected, consecutively encountered lesions—regardless of vascular visibility—and reporting the frequency and characteristics of cases in which MBFR measurement fails would provide a clearer understanding of the system's practical utility. Furthermore, recent studies indicate that EGC after H. pylori eradication frequently demonstrate faint or indistinct microvascular patterns, complicating dynamic flow assessment 4. Evaluating the algorithm in such presentations may help clarify its true clinical role. In summary, the work by Akazawa et al. represents an important step toward objective endoscopic evaluation. Addressing real-world imaging variability and incorporating diagnostically difficult lesions in future validation studies will substantially enhance the generalizability and clinical relevance of this promising system. We congratulate the authors on their valuable contribution. The authors have nothing to report. The authors declare no conflicts of interest. Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
Liu et al. (Thu,) studied this question.