Fat/water ratios measured with diffuse reflectance spectroscopy to detect breast tumor boundaries

Abstract

Recognition of the tumor during breast-conserving surgery (BCS) can be very difficult and currently a robust method of margin assessment for the surgical setting is not available. As a result, tumor-positive margins, which require additional treatment, are not found until histopathologic evaluation. With diffuse reflectance spectroscopy (DRS), tissue can be characterized during surgery based on optical parameters that are related to the tissue morphology and composition. Here we investigate which optical parameters are able to detect tumor in an area with a mixture of benign and tumor tissue and hence which parameters are most suitable for intra-operative margin assessment. DRS spectra (400-1600 nm) were obtained from 16 ex vivo lumpectomy specimens from benign, tumor border, and tumor tissue. One mastectomy specimen was used with a custom-made grid for validation purposes. The optical parameter related to the absorption of fat and water (F/W-ratio) in the extended near-infrared wavelength region (~1000-1600 nm) provided the best discrimination between benign and tumor sites resulting in a sensitivity and specificity of 100 % (excluding the border sites). Per patient, the scaled F/W-ratio gradually decreased from grossly benign tissue towards the tumor in 87.5 % of the specimens. In one test case, based on a predefined F/W-ratio for boundary tissue of 0.58, DRS produced a surgical resection plane that nearly overlapped with a 2-mm rim of benign tissue, 2 mm being the most widely accepted definition of a negative margin. The F/W-ratio provided excellent discrimination between sites clearly inside or outside the tumor and was able to detect the border of the tumor in one test case. This work shows the potential for DRS to guide the surgeon during BCS.

In vivo tumor identification of colorectal liver metastases with diffuse reflectance and fluorescence spectroscopy

“…Technical advances in the treatment of liver malignancies, such as laparoscopic and robotic surgery as well as more frequent administration of neo-adjuvant systemic treatment, call for new technologies aiding the surgeon to correctly locate and resect the entire tumor volume in the liver. In previous studies, we have shown that DRS and FS can be used to identify various types of tissue in ex vivo experiments in lung and breast [8,22,23]. Also in liver tissue, we have demonstrated that DRS can accurately discriminate normal liver tissue from CRLM in an ex vivo model with a sensitivity and specificity of 94% [12].…”

Quantitative in vivo detection of adipose tissue browning using diffuse reflectance spectroscopy in near‐infrared II window

“… It has been reported that extended NIR wavelength region measurement renders a better discrimination between benign and tumor sites as well as accurate estimation of chromophores present in the tissue. Recently, detection of classical WAT and BAT in vivo using optical techniques, such as diffuse optical spectroscopy and imaging (DOSI) or time resolved spectroscopy, was reported. In one of these studies, Ganesan et al used DOSI to detect the changes in subcutaneous adipose tissue during weight loss in humans.…”

Imaging depth variations in hyperspectral imaging: Development of a method to detect tumor up to the required tumor‐free margin width

“…Based on previous research, one of the main differences in optical properties between tumor and healthy breast tissue is the higher water and lower fat content in tumors compared to healthy tissue . Therefore, we focused on the NIR wavelength range, where water and fat are the main absorbers of light.…”

“…We estimated the optical penetration depth and diffuse reflectance in breast tissue using known optical properties of breast tissue . Since water and fat are the main absorbers of light in the NIR region, equations for the absorption coefficient and the reduced scattering were approximated by: In which [water] and [fat] represent the water and fat concentration in the tissue, and ε water and ε fat their extinction spectra.…”

“…To develop this method, we first identified wavelength regions at which the penetration depth of photons does not depend on whether the tissue is healthy or malignant. Previous research showed that the amount of water and fat is one of the main optical differences between tumor and healthy breast tissue . Therefore, we focused on the near‐infrared wavelength range, where water and fat are the main absorbers.…”