Magnetic Microrobot With Drilling‐Sensing Dual Functionality for Targeted Biopsy of Deep‐Seated Tracheal Microlesions

Biopsy remains gold standard for definitive diagnosis of numerous diseases, yet conventional approaches confront inherent limitations encompassing procedural invasiveness, limited spatial resolution, poor deep-tissue accessibility, and complicated post-operative histopathology examinations. Herein, a multifunctional magnetically actuated microrobotic platform for targeted biopsy of deep-seated micro-lesion in lung for potential use of early lung cancer diagnosis is reported. This conical helix-structured microrobot features controllable gold nanospike decoration at its tip, integrating dual functionality in drilling-enabled tissue sampling and surface-enhanced Raman scattering (SERS) biosensing. Utilizing real-time tracheoscopy and fluoroscopy guidance, this platform achieves magnetic navigation through saline-filled bronchi to target lesions, followed by on-demand sampling and retrieval, whose feasibility is systematically verified through ex vivo porcine lung models and in vivo rabbit trials. The collected biomarker and tissue samples can be rapidly recognized and quantified based on SERS spectroscopy. Moreover, through establishing a convolutional neural network-based deep learning model, it is succeed to distinguish between the healthy and various cancerous lung tissues from clinical patients, achieving a recognition accuracy of over 94.3% within just 3 min. Therefore, this work represents a microrobot-based biopsy techniques for noninvasive and precise sampling in lung with diagnosis capability of early-stage lung cancer, offering substantial improvements over conventional biopsy methods.