Volumetric tumor delineation and assessment of its early response to radiotherapy with optical coherence tomography

Abstract

Texture analyses of optical coherence tomography (OCT) images have shown initial promise for differentiation of normal and tumor tissues. This work develops a fully automatic volumetric tumor delineation technique employing quantitative OCT image speckle analysis based on Gamma distribution fits. We test its performance in-vivo using immunodeficient mice with dorsal skin window chambers and subcutaneously grown tumor models. Tumor boundaries detection is confirmed using epi-fluorescence microscopy, combined photoacoustic-ultrasound imaging, and histology. Pilot animal study of tumor response to radiotherapy demonstrates high accuracy, objective nature, novelty of the proposed method in the volumetric separation of tumor and normal tissues, and the sensitivity of the fitting parameters to radiation-induced tissue changes. Overall, the developed methodology enables hitherto impossible longitudinal studies for detecting subtle tissue alterations stemming from therapeutic insult.

Longitudinal In-Vivo Quantification of Tumour Microvascular Heterogeneity by Optical Coherence Angiography in Pre-Clinical Radiation Therapy

“…The temporal fluctuations in 𝜆, essentially not observed in 𝐷𝐿𝐹 150 , likely arise from the high sensitivity of this curve-fitting based metric to the short distance range where we expect most speckle artefacts missed during the post-processing (see Section 2.4). With improved automation of both tumour [42] and especially vessel segmentations [43], increasing accuracy and inter-timepoint consistency in the OCA post-processing, observed temporal fluctuations in the 𝜆 metric should be greatly reduced.…”

OCT-specific signal features for semi-automatic semantic scans annotation and segmentations

“…This tool can be further specifically tuned for semantic isolation/annotation of specific biological structures (such as epidermis, vessels, tumor borders, edema etc) taking into account specificity of various OCT devices, tissues and tasks. This tool is ready for potential integration of approaches presented in papers devoted to OAC-based layer-by-layer assessment of the intestinal wall 15 , lymph vessels visualization 16,17,18 , and speckle statistics based volumetric tumor delineation 16 , and speckle-contrast analysis of breast cancer tissue 20 as well as other structures 21 .…”

Antimicrobial PDT effectively destroys E. coli and E. faecalis orthopaedic biofilms compared to low efficacy of a tobramycin and vancomycin mixture: an in vitro study using optical coherence tomography

“…For this, phantom-based measurements of attenuation coefficients specific to our particular OCT system were performed following the procedure described in detail in ref. [13], with the signal depth-decay compensation performed from the top to the bottom 14 . Figure 1f shows cross-sectional OCT images (B-scans) of one microfluidic device before and after signal depth-decay compensation.…”