The precise identification of tumor boundaries during radical prostatectomy remains a major clinical challenge. As positive surgical margins occur in 15-40% of prostate cancer cases, the risk of postoperative recurrence and functional impairment is significantly increased.

In a study published in Nature Biomedical Engineering on Nov. 18, researchers from Shenzhen Institute of Advanced Technology (SIAT) of the Chinese Academy of Sciences and Fudan University developed a label-free intraoperative navigation system based on surface-enhanced Raman scattering (SERS), which enables real-time, in situ grading of prostate tumor malignancy by simultaneously detecting tissue acidity and prostate-specific antigen (PSA) activity.

This system integrates a microfluidic sampling pen, a nanoimprinted SERS array, and an artificial intelligence-powered spectral analysis module. The pen automatically extracts biomarkers from tissue surfaces within six seconds per point without causing tissue damage. The extracted droplets are then transferred to a highly uniform SERS array functionalized with Raman reporters sensitive to pH and PSA. A custom two-dimensional deep learning model processes the spectral data and outputs quantitative biomarker readings in under two minutes.

In a clinical trial involving 144 prostate cancer patients, the system achieved an area under the receiver operating characteristic curve (AUC) of 0.890 for identifying high-grade tumors (Gleason Grade Group ≥ 3), significantly outperforming traditional intraoperative methods such as frozen section analysis which is time-consuming and operator-dependent.

Notably, the system is automated, making it a promising platform for real-time "molecular pathology" during surgery. Besides, by providing surgeons with a "malignancy map" of the resection area, it helps optimize the balance between maximal tumor removal and minimal damage to functional tissues.

This SERS navigation system represents an advance in intraoperative tumor malignancy assessment with broad implications for precision cancer surgery. The study provides a new platform for precise tumor boundary identification during surgery.