Toward optical spectroscopy‐guided lung biopsy: Demonstration of tissue‐type classification

Abstract

The diagnostic yield of standard tissuesampling modalities of suspected lung cancers, whether by bronchoscopy or interventional radiology, can be nonoptimal, varying with the size and location of lesions. What is needed is an insitu sensor, integrated in the biopsy tool, to objectively distinguish among tissue types in real time, not to replace biopsy with an optical diagnostic, but to verify that the sampling tool is properly located within the target lesion. We investigated the feasibility of elastic scattering spectroscopy (ESS), coupled with machine learning, to distinguish lung lesions from the various nearby tissue types, in a study with freshly-excised lung tissues from surgical resections.Optical spectra were recorded with an ESS fiberoptic probe in different areas of the resected pulmonary tissues, including benign-margin tissue sites as well as the periphery and core of the lesion. An artificial-intelligence model was used to analyze, retrospectively, 2032 measurements from excised tissues of 35 patients. With high accuracy, ESS was able to distinguish alveolar tissue from bronchi, alveolar tissue from lesions, and bronchi from lesions. This ex vivo study indicates promise for ESS fiberoptic probes to be integrated with surgical intervention tools, to improve reliability of pulmonary lesion targeting.

Distinguishing tumor from healthy tissue in human liver ex vivo using machine learning and multivariate analysis of diffuse reflectance spectra

“…Almost all previous studies that use DRS to distinguish between human tumors and tumor-free liver tissues have used model-based analysis based on diffusion theory [7,38,39,41,42]. However, Dremin et al [43] trained ANNs using estimates of blood oxygen saturations obtained from MC-simulated DRS spectra.…”

“…Mathematical models, such as semi‐empirical models, the diffusion approximation to the radiative transfer equation and Monte Carlo (MC) methods have been used to derive physiological parameters that characterize the tissue. However, only a few studies have applied DRS in human liver tissue characterization [7, 8, 38–43, 54, 55], and all but two of these previous studies used the standard analytical model derived from diffusion theory proposed by Farrell et al [56]. Evers et al [39] used DRS with an extended wavelength range from 500 to 1600 nm and examined liver specimens immediately after resection from patients with colorectal liver metastases to discriminate tumors from surrounding tissues with a sensitivity and specificity both of 94% and an accuracy of 100% for each patient.…”

“…Evers et al [39] used DRS with an extended wavelength range from 500 to 1600 nm and examined liver specimens immediately after resection from patients with colorectal liver metastases to discriminate tumors from surrounding tissues with a sensitivity and specificity both of 94% and an accuracy of 100% for each patient. After rigorous modeling, as described by Nachabe et al [38], the authors also showed that healthy liver tissue contains more bile and blood compared to tumorous liver tissues and concluded that bile is a key discriminatory chromophore. Using the same technique in the first in vivo study, Tanis et al [41] distinguished liver tumors from surrounding tumor‐free tissues with a sensitivity of 95% and a specificity of 92%.…”

“…Many research groups are attempting to establish diagnostic criteria, or a form of "optical biopsy", based on the correlation between the histology and optical signatures of tissues, for a variety of human tissues, such as colon [9][10][11], skin [12][13][14], breast [15][16][17][18][19][20], cervical [21][22][23], oral [23][24][25][26][27], prostate [28][29][30][31], nasopharynx [32,33] and lung tissues [34][35][36][37]. Studies have also showed the potential of probe-based DRS to discriminate between human liver tumors and tumor-free surrounding tissue [7,8,[38][39][40][41][42][43].…”

Optical spectroscopy and imaging in surgical management of cancer patients

“…The objective of so-called "optical biopsy" is to classify the tissue without taking actual tissue samples. This has been a particular focus in endoscopic detection of tumors or premalignant lesions in, for example, the lung [39] and GI tract [13,40,41]. Alternatively, optical measurements can be integrated with the physical biopsy device to increase the likelihood that the tissue sample will be representative of the highest-grade tumor and to minimize the risk of damaging significant blood vessels.…”