Ultrafast, label-free near-intraoperative pathology using machine-learning-assisted broadband photoacoustic spectral analysis, a case study with orofacial tumors

Ultrafast, label-free histopathological assessment of intraoperative margins remains a critical challenge. Here, we present a near-intraoperative diagnostic approach using broadband Photoacoustic Spectral Analysis (bPASA). With broadband Ultraviolet Photoacoustic Microscopy (bUV-PAM), we acquired spectral signatures from unprocessed tissue specimens and integrated these data into a Logistic Regression (LR) algorithm to develop diagnostic models for ultrafast, label-free assessment. The models were established and validated using over 1000 mouse and human orofacial tumor samples. For murine Oral Squamous Cell Carcinoma (OSCC) and Adenoid Cystic Carcinoma (ACC), the models achieved average Area Under the Curve (AUC) values of approximately 0.86, 0.84, and 0.95 for sections, fresh tissues, and in vivo samples, respectively. In human clinical specimens, we developed sequential diagnostic models covering seven pathological states, achieving an average AUC of approximately 0.87. These findings demonstrate that the method enables ultrafast, label-free assessment of surgical margins within 2 minutes, a substantial time saving over conventional frozen-section analysis, and positions it as a promising tool for real-time intraoperative pathology. Despite these encouraging results, several limitations should be noted, including the ex vivo nature of most clinical validation, the limited sample sizes in some external validation subgroups, the exclusion of mixed pathological samples, the lack of prospective intraoperative testing, and incomplete evidence for Formalin-Fixed Paraffin-Embedded (FFPE)‑to‑fresh transferability across all diagnostic categories.

Authors

Institutions

Publication Details

Journal
Photoacoustics
Published
2026-11-18
DOI
https://doi.org/10.1016/j.pacs.2026.100906
Primary Topic
Photoacoustic and Ultrasonic Imaging
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Ultrafast, label-free near-intraoperative pathology using machine-learning-assisted broadband photoacoustic spectral analysis, a case study with orofacial tumors

尚建伟 Shang Jianwei, Changyu Shen, Yingbin Yan, Lei Hou et al.
1 links
Photoacoustics
Photoacoustic and Ultrasonic Imaging
article

Ultrafast, label-free near-intraoperative pathology using machine-learning-assisted broadband photoacoustic spectral analysis, a case study with orofacial tumors

尚建伟 Shang Jianwei, Changyu Shen, Yingbin Yan, Lei Hou, Fan Yang, Xiaocong Yuan, Zhiling Zhang, Yong Yang, Li Zhang, Xiaojing Wu
article en

Abstract

Ultrafast, label-free histopathological assessment of intraoperative margins remains a critical challenge. Here, we present a near-intraoperative diagnostic approach using broadband Photoacoustic Spectral Analysis (bPASA). With broadband Ultraviolet Photoacoustic Microscopy (bUV-PAM), we acquired spectral signatures from unprocessed tissue specimens and integrated these data into a Logistic Regression (LR) algorithm to develop diagnostic models for ultrafast, label-free assessment. The models were established and validated using over 1000 mouse and human orofacial tumor samples. For murine Oral Squamous Cell Carcinoma (OSCC) and Adenoid Cystic Carcinoma (ACC), the models achieved average Area Under the Curve (AUC) values of approximately 0.86, 0.84, and 0.95 for sections, fresh tissues, and in vivo samples, respectively. In human clinical specimens, we developed sequential diagnostic models covering seven pathological states, achieving an average AUC of approximately 0.87. These findings demonstrate that the method enables ultrafast, label-free assessment of surgical margins within 2 minutes, a substantial time saving over conventional frozen-section analysis, and positions it as a promising tool for real-time intraoperative pathology. Despite these encouraging results, several limitations should be noted, including the ex vivo nature of most clinical validation, the limited sample sizes in some external validation subgroups, the exclusion of mixed pathological samples, the lack of prospective intraoperative testing, and incomplete evidence for Formalin-Fixed Paraffin-Embedded (FFPE)‑to‑fresh transferability across all diagnostic categories.

PhotoacousticsVol. 52
Jiangnan University (CN), Shenzhen University (CN), Nankai University (CN), Zhejiang Lab (CN), Tianjin Stomatological Hospital (CN), China Jiliang University (CN)
National Natural Science Foundation of China, Nankai University, Shenzhen University, Natural Science Foundation of Zhejiang Province
Good health and well-being
Photoacoustic and Ultrasonic Imaging
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon