Advancing endometrial cancer diagnosis through optical biopsy: novel minimally invasive approach

View at publisher
Authors: Sanskar Thakur, Yixiao Lin, Jinhua Xu, Haolin Nie, Sarah Badwan, Lukai Wang, Lindsay Kuroki, Ian Hagemann, Quing Zhu
Cited by: 0
References: 8

Abstract

Endometrial cancer is the most common malignancy of the female reproductive system, with over 69,000 cases and 13,000 deaths estimated in 2025. The current standard of care relies on invasive diagnostic procedures such as hysteroscopy or biopsy, which also carry a false negative rate exceeding 10%. Optical Coherence Tomography (OCT), with its micron-level resolution and millimeter-scale imaging depth, offers a promising non-invasive alternative for biopsy. In this pilot study, we demonstrate the feasibility of using OCT for diagnosing endometrial cancer in human uteri samples. A custom 3.1mm OCT catheter operating at 1310nm and 20kHz was developed to acquire 3D volumetric images of the endometrial cavity. A total of 69 patients—diagnosed as normal, high-risk endometrial intraepithelial neoplasia (EIN), grade 1 cancer, or cancer with EIN—were enrolled. From December 2024 to June 2025, 57 uteri samples were imaged. Each 3D OCT volume featured a sequence of 2D B-scans with spatial and depth resolution of approximately 10μm, with a longitudinal spacing of 250μm. Optical scattering and structural radiomics features were extracted from the volumes. Dimensionality reduction was performed using cosine similarity-based network graph, followed by classification using a linear regression to distinguish normal from cancerous tissue. The model achieved an area under the curve (AUC) of 0.95. These results suggest that 3D OCT imaging, combined with feature-based analysis, holds significant potential as a non-invasive tool for screening and diagnosing endometrial cancer.

Fields of Study & Subjects

Optical Coherence Tomography Applications AI in cancer detection Photoacoustic and Ultrasonic Imaging

Publication details

Journal / source
Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine
Pages
25–25
Publication date
5 March 2026
Language
EN