Fiber Bragg Grating Sensors Combined With Artificial Intelligence for Noninvasive Breast Tumor Identification

Noninvasive breast palpation is a common preliminary examination to detect hard inclusions (e.g., tumors) within breast tissue. In this context, sensorized tactile probes are emerging as a promising solution to quantitatively identify alterations in tissue stiffness and support medical practice. However, these systems are still affected by low spatial resolution and high encumbrance, which can hinder their effectiveness and usability. To overcome the existing challenges, the present study proposed a novel smart tactile probe based on fiber Bragg grating (FBG) for noninvasive breast palpation, which exhibits enhanced spatial resolution and ergonomic design to ensure patient comfort during the examination. The smart probe was designed, fabricated, and validated through palpation tests conducted under controlled and uncontrolled conditions on phantoms simulating breast tissue with an embedded tumor. Moreover, in this study we investigated the feasibility of artificial intelligence (AI) algorithms to automatically detect the presence of tumors from the data collected by the proposed system. Here, for the first time, the FBG technology was combined with AI to develop an AI-driven approach for noninvasive breast tumor identification. The results demonstrated high accuracy in tumor detection, with a minimal rate of false negatives (i.e., recall above 95%). As a result, integrating FBG technology with AI shows great potential for accurately detecting the presence or absence of hard inclusions. These encouraging results will foster future optimization of both the tactile probe design and AI algorithms, aiming to further improve the reliability of breast palpation examinations and early tumor diagnosis.