Abstract
<title>Abstract</title> <p>Background and Aims Fluorescent rapid on-site evaluation (ROSE) is an emerging technique that enables real-time cytological assessment during endoscopic procedures. This study aims to demonstrate the morphological imaging features of fluorescent ROSE for real-time visualization of Helicobacter pylori (H. pylori) during gastroscopy, and to explore its feasibility in differentiating viable from non-viable bacteria, as well as its correlation with endoscopic manifestations and clinical characteristics. Methods A total of 210 patients with suspected H. pylori infection who underwent gastroscopy between January 2026 and June 2026 were enrolled. Gastric mucosal tissues were collected during gastroscopy and processed using fluorescent ROSE with acridine orange-ethidium bromide (AO-EB) double staining. Fluorescent morphology, signal intensity, and distribution patterns of H. pylori were recorded. Fluorescence criteria for distinguishing viable and non-viable bacteria were defined based on staining characteristics. Endoscopic findings, clinical symptoms, and prior eradication history were collected for correlation analysis. All patients underwent 13C-urea breath test (13C-UBT) prior to enrollment; 13C-UBT results were used as reference for H. pylori infection status but diagnostic accuracy metrics were not the primary objective of this descriptive study. Results: Of the 210 patients, fluorescent ROSE identified 174 (82.9%) as H. pylori-positive, of which 122 (58.1%) showed viable bacteria (yellowish-green fluorescence) and 52 (24.8%) showed non-viable bacteria (orange-red fluorescence); 36 (17.1%) were negative. Under fluorescence microscopy, viable bacteria appeared as rod-shaped, arc-shaped, or S-shaped structures with bright yellow-green signals and intact morphology; non-viable bacteria exhibited orange-red fluorescence with fragmented or coccoid forms. Nodular antral gastritis demonstrated the highest viable bacteria positivity rate (92.3%), significantly higher than other endoscopic types (P<0.001). The incidence of dyspeptic symptoms in the viable bacteria group (82.8%) was significantly higher than in the non-viable group (36.5%, P<0.001). Patients with prior H. pylori eradication therapy had a significantly lower viable bacteria rate compared to untreated patients (31.3% vs. 74.6%, P<0.001). Conclusion Fluorescent ROSE provides real-time, intuitive visualization of H. pylori with clear morphological characteristics and demonstrates potential for distinguishing viable from non-viable bacteria during gastroscopy. Nodular antral gastritis shows the strongest association with viable H. pylori colonization. This technique may assist endoscopists in preliminary real-time assessment of infection status, though further diagnostic accuracy studies are warranted. This descriptive study serves as a technical demonstration of the method's feasibility and clinical correlation.</p>