Artificial Intelligence-Based Device for Detection of Oral Cancer

An Artificial Intelligence-Based Device for Detection of Oral Cancer is disclosed. The Device integrates in a handheld, battery-powered instrument: a multi-modal oral mucosal imaging probe providing co-registered white-light reflectance (20 MP CMOS, 5,500K LED), autofluorescence (400–450 nm excitation, long pass filtered), and narrow-band (415 nm and 540 nm) images within a 3-second acquisition cycle; a 785 nm confocal Raman spectroscopy module providing depth-selective epithelial tissue biochemical characterization in the 600–1,800 cm⁻¹ fingerprint region; a disposable electrochemical salivary biomarker cartridge measuring IL-6, IL-8, TNF-α, VEGF, EGF, sCD44, LDH, and HPV-16/18 LAMP nucleic acid detection, an onboard NPU-based AI multi-modal fusion engine implementing a multi-branch CNN for imaging analysis, Raman spectral classification, and salivary biomarker risk scoring — fused into an Oral Mucosal Risk Score (OMRS) and five-category lesion classification with explainable AI diagnostic report within 90 seconds; a real-time image quality assurance algorithm; a step-by-step examination protocol guide with multilingual voice output; fingerprint biometric patient identification; longitudinal lesion progression tracking; store-and-forward and real-time video teleconsultation via 4G/LTE; GPS-tagged screening record logging; and HL7 FHIR cloud registry integration. The Device enables non-invasive, AI-driven, early oral cancer and OPMD detection at point-of-care by primary healthcare practitioners without specialized oncology expertise, addressing the critical screening gap in high-incidence populations in South Asia, Southeast Asia, and other low- and middle-income settings.