Diagnostic Utility of Fluorescence and Diffuse Reflectance Imaging in Gastrointestinal Luminal Pathology

BACKGROUND: Cancer is a fatal disease with significant global mortality. Because cancer is so diverse and lacks biomolecular indicators, it is difficult to diagnose at an early stage using current histopathological diagnostic techniques. However, recent cutting-edge fluorescence and diffuse reflectance (DR) imaging techniques, which use sub-cellular spectral features, have proven beneficial in cancer diagnostics. This study aimed to investigate the diagnostic potential of these methods for discriminating malignant from normal gastrointestinal tissue in an ex vivo setting. MATERIALS AND METHODS: An experimental trial was conducted on 42 patients attending the Department of General Surgery, Kasturba Medical College, Manipal. Fluorescence and DR at 399 sites were measured using a handheld multimodal device. The spectral readings were taken immediately after the surgical resection of the specimen. Light-emitting diodes (LED) emitting at 375 nm, 545 nm, 575 nm, and 610 nm wavelengths were used for tissue illumination, and the backscattered light was analyzed using proprietary software. The DR ratio (R610/R545) intensity values were determined for all the different sites examined, and a scatter plot diagram was made to correlate with tissue pathology. RESULTS: Spectral readings of histologically confirmed malignant tissues were compared with those of histologically verified normal tissues and were used as controls. In total, 399 sites yielded 766 spectral findings. In an ex vivo context, our study demonstrated a sensitivity of 78.8% and specificity of 98.3% for differentiating between malignant and normal tissues. Additionally, the study showed a 90% specificity and 92% sensitivity for differentiating between benign and malignant tissues. CONCLUSION: The diagnostic utility of fluorescence and DR imaging demonstrates its potential to discriminate between normal tissues and pathological lesions of the gastrointestinal tract in real-time.