Peter Eskil Andersen

Ninety percent of drugs fail during clinical trials, mainly due to lack of clinical efficacy. Recent developments in in vitro models such as 3D tumor heterospheroids have led to improvements in failure rates, but the relative lack of standardized evaluation methods for 3D cultures limits their utility in high-throughput screening. Optical coherence tomography (OCT) shows significant promise for high-throughput screening of 3D models; however, the optimal classification model and key image features for assessing drug efficacy in OCT images of spheroids has yet to be explored in detail. In this study, we investigate whether OCT combined with machine learning methods can be used to identify biomarkers of drug efficacy in 3D tumor spheroid models. We further compare the performance of two different models to determine the optimal configuration for accurate classification. Volumetric OCT images were acquired of co-cultured HT29 spheroids treated with 3 different concentrations of cisplatin. A two-dimensional multi-view ResNet model and a three-dimensional ResNet model were used to classify the images and to identify key image features associated with each group. Differences between spheroids treated with different concentrations of cisplatin are clearly visible in the OCT images. Our model was able to classify the images based on cisplatin concentration with 71.9% accuracy using the 2D multi-view model and 91.2% accuracy using the 3D model. Key features in the 3D model significantly improved the model accuracy. These results underscore the possibility that OCT could be used for high-throughput screening of drugs using 3D in vitro models and highlight key identifying features for further investigation.

Alterations in tissue, such as the onset of disease, have been observed to affect the absorption and scattering properties—which, in turn, change the attenuation of light or sound waves propagating in the tissue.¹ In computed tomography (CT) and ultrasound, tissue attenuation is used to differentiate among tissue types, providing detailed insights into tissue structure and function.^2,3 In CT, the tissue attenuation is measured by the loss of intensity of the X-ray beam passing through the sample, mainly due to water absorption.^2,4 However, contrast in CT is limited by the small difference in absorption coefficient among soft tissues, making it better suited for higher density tissues such as bone.² Ultrasound is sensitive to the decrease in the amplitude of the acoustic wave propagating through tissue, which is affected by not only absorption but also scattering, reflection, refraction, and diffraction.³ The change in wave amplitude enables better distinction among tissues with similar density. However, the resolution of ultrasound does not currently allow for the identification of microscopic changes in tissue, which is important in satellite lesion detection and in vivo margin assessment.³

The penetration depth of optical coherence tomography (OCT) reaches well beyond conventional microscopy; however, signal reduction with depth leads to rapid degradation of the signal below the noise level. The pursuit of imaging at depth has been largely approached by extinguishing multiple scattering. However, in OCT, multiple scattering substantially contributes to image formation at depth. Here, we investigate the role of multiple scattering in OCT image contrast and postulate that, in OCT, multiple scattering can enhance image contrast at depth. We introduce an original geometry that completely decouples the incident and collection fields by introducing a spatial offset between them, leading to preferential collection of multiply scattered light. A wave optics–based theoretical framework supports our experimentally demonstrated improvement in contrast. The effective signal attenuation can be reduced by more than 24 decibels. Notably, a ninefold enhancement in image contrast at depth is observed in scattering biological samples. This geometry enables a powerful capacity to dynamically tune for contrast at depth.

The current gold standard for clinical diagnosis of melanoma is excisional biopsy and histopathologic analysis. Approximately 15–30 benign lesions are biopsied to diagnose each melanoma. In addition, biopsies are invasive and result in pain, anxiety, scarring, and disfigurement of patients, which can add additional burden to the health care system. Among several imaging techniques developed to enhance melanoma diagnosis, optical coherence tomography (OCT), with its high-resolution and intermediate penetration depth, can potentially provide required diagnostic information noninvasively. Here, we present an image analysis algorithm, “optical properties extraction (OPE),” which improves the specificity and sensitivity of OCT by identifying unique optical radiomic signatures pertinent to melanoma detection. We evaluated the performance of the algorithm using several tissue-mimicking phantoms and then tested the OPE algorithm on 69 human subjects. Our data show that benign nevi and melanoma can be differentiated with 97% sensitivity and 98% specificity. These findings suggest that the adoption of OPE algorithm in the clinic can lead to improvements in melanoma diagnosis and patient experience.Significance:This study describes a noninvasive, safe, simple-to-implement, and accurate method for the detection and differentiation of malignant melanoma versus benign nevi.

Supplementary file includes two sections entitled "Preprocessing" and "Mie simulations". 14 figures and 8 tables. Fig S1 shows a schematic diagram of the Michelson Diagnostics SS-OCT, Fig S2 shows histologic photographs and OCT images of melanoma and nevus cases, Fig S3 is the pre-processing procedure optimization flowchart, Fig S4, S5, S6 and S7 are related to optical properties of melanoma and benign nevus, Fig S8 indicates the results of classifier, Fig S9 and 10 show ROC and AUC curves for the selected classifiers, Fig S11 shows absorption coefficients with linearized X-axis, Fig S12, S13 and 14 are related to preprocessing optimization strategies. Table S1 shows the material used in phantom construction, Table S2 lists the specifications of melanoma and benign lesions, Table S3 shows definitions of diagnostic statistics, Table S4 and S5 are related to classification results, Table S6, S7 and S8 show the relation between physical characteristics of particles and their optical properties.

The current gold standard for clinical diagnosis of melanoma is excisional biopsy and histopathologic analysis. Approximately 15–30 benign lesions are biopsied to diagnose each melanoma. In addition, biopsies are invasive and result in pain, anxiety, scarring, and disfigurement of patients, which can add additional burden to the health care system. Among several imaging techniques developed to enhance melanoma diagnosis, optical coherence tomography (OCT), with its high-resolution and intermediate penetration depth, can potentially provide required diagnostic information noninvasively. Here, we present an image analysis algorithm, “optical properties extraction (OPE),” which improves the specificity and sensitivity of OCT by identifying unique optical radiomic signatures pertinent to melanoma detection. We evaluated the performance of the algorithm using several tissue-mimicking phantoms and then tested the OPE algorithm on 69 human subjects. Our data show that benign nevi and melanoma can be differentiated with 97% sensitivity and 98% specificity. These findings suggest that the adoption of OPE algorithm in the clinic can lead to improvements in melanoma diagnosis and patient experience.Significance:This study describes a noninvasive, safe, simple-to-implement, and accurate method for the detection and differentiation of malignant melanoma versus benign nevi.

Supplementary file includes two sections entitled "Preprocessing" and "Mie simulations". 14 figures and 8 tables. Fig S1 shows a schematic diagram of the Michelson Diagnostics SS-OCT, Fig S2 shows histologic photographs and OCT images of melanoma and nevus cases, Fig S3 is the pre-processing procedure optimization flowchart, Fig S4, S5, S6 and S7 are related to optical properties of melanoma and benign nevus, Fig S8 indicates the results of classifier, Fig S9 and 10 show ROC and AUC curves for the selected classifiers, Fig S11 shows absorption coefficients with linearized X-axis, Fig S12, S13 and 14 are related to preprocessing optimization strategies. Table S1 shows the material used in phantom construction, Table S2 lists the specifications of melanoma and benign lesions, Table S3 shows definitions of diagnostic statistics, Table S4 and S5 are related to classification results, Table S6, S7 and S8 show the relation between physical characteristics of particles and their optical properties.

The guest editorial introduces the Journal of Biomedical Optics special collection, Perspectives in Biophotonics, https://www.spiedigitallibrary.org/biophotonics-tutorials.