DOI：10.12201/bmr.202605.00040

Real-Time Pathogen Visualization: Fluorescence ROSE versus Kyoto Classification for H. pylori Infection Activity and Eradication Assessment

•  

  Abstract: Objective To compare the diagnostic value of fluorescence rapid on‑site evaluation (F‑ROSE) and Kyoto classification‑based endoscopic direct diagnosis in determining Helicobacter pylori (Hp) infection, distinguishing infection activity, and assessing post‑eradication status, to clarify the differences between the two methods in low‑bacterial‑load infection, active infection, and residual infection after eradication, and to establish an integrated endoscopic precision diagnostic pathway for Hp covering qualitative, quantitative, activity and eradication assessment. Methods A total of 210 consecutive patients undergoing gastroscopy were enrolled and underwent both fluorescence ROSE and Kyoto classification endoscopic evaluation. ROSE used AO‑EB dual fluorescent staining, and viable, non‑viable and total bacteria were counted under ×400 high‑power field (HPF). Kyoto classification classified patients as uninfected, currently infected, or past/eradicated infection based on regular arrangement of collecting venules (RAC), diffuse redness, atrophy, intestinal metaplasia, nodularity and other signs. Latent class analysis (LCA) combined with a Bayesian model was used to infer true infection and activity status. ROC curves, consistency tests and subgroup analyses were performed to compare the diagnostic performance of the two methods, focusing on low‑bacterial‑load infection, post‑eradication residual infection, and endoscopic gray‑zone cases. Results By model inference, among the 210 patients, 107 had true current infection, 68 had past/eradicated infection, and 35 were uninfected. The positive rate of fluorescence ROSE (50.5%) was significantly higher than that of Kyoto classification (44.3%, P<0.05). With true infection as reference: fluorescence ROSE showed sensitivity 92.5%, specificity 95.3%, Kappa=0.891, AUC=0.948; Kyoto classification showed sensitivity 79.4%, specificity 90.2%, Kappa=0.726, AUC=0.863. Among 57 patients with true low‑bacterial‑load infection (4‑9 viable bacteria/HPF) inferred by the model, the detection rate of ROSE was 100.0%, while that of Kyoto classification was only 35.7%. In post‑eradication cases, the sensitivity of ROSE in identifying viable residual/recurrent infection was 88.2%, significantly higher than 41.2% of Kyoto classification. Among 41 endoscopic gray‑zone (suspicious infection) cases by Kyoto classification, 73.2% were confirmed as true viable infection by ROSE. Conclusions Kyoto classification relies on mucosal morphology to indirectly infer Hp status, which is prone to miss low‑bacterial‑load and post‑eradication residual infection and cannot distinguish viable from dead bacteria. Fluorescence ROSE enables real‑time visual identification of viable bacteria, accurate quantification, direct judgment of infection activity and eradication effect, breaking through the limitations of endoscopic morphology. It is an optimal technique for current infection, low‑bacterial‑load infection and post‑eradication evaluation, and can complement Kyoto classification to form a dual‑dimensional precision diagnosis system of morphology %2B pathogen.

  Key words: fluorescence ROSE; Helicobacter pylori; Kyoto classification of gastritis; endoscopic direct diagnosis; infection activity; post‑eradication evaluation; low‑bacterial‑load infection

  提交时间：2026-05-29

  版权声明：作者本人独立拥有该论文的版权，预印本系统仅拥有论文的永久保存权利。任何人未经允许不得重复使用。
• html

• 图表

• 张林, 侯艳红, 李炳慧, 吴凯, 张静, 杨汨. 基于荧光 ROSE 的幽门螺杆菌实时可视化技术在HP除菌治疗全程动态评估中的应用研究. 2026. doi: 10.12201/bmr.202606.00007

  李炳慧, 张林, 侯艳红, 吴凯, 张静, 杨汨. 基于荧光 ROSE 的幽门螺旋杆菌实时可视化诊断新方法：一项基于无金标准潜在类别分析的大样本对照研究. 2026. doi: 10.12201/bmr.202606.00055

  李炳慧, 张林, 侯艳红, 吴凯, 韩家峰, 刘君钰, 张静, 杨汨. 一小时幽门螺杆菌表型药敏检测：基于荧光代谢活性的免培养新技术初步验证. 2026. doi: 10.12201/bmr.202607.00011

  莫巧玲, 刘菲, 张孟维, 杨日华, 邱全喜, 廖凯明. 香连丸治疗脾胃湿热型慢性萎缩性胃炎合并幽门螺杆菌感染的临床观察. 2026. doi: 10.12201/bmr.202604.00037

• 序号	提交日期	编号	操作
  1	2026-05-10	10.12201/bmr.202605.00040V1	下载

• 公开评论  匿名评论  仅发给作者 提交评论