Open Access
Add to BibSonomy
Abstract
In this work, a common-path optical coherence tomography (OCT) system is demonstrated for characterizing the tissue in terms of some optical properties. A negative axicon structure chemically etched inside the fiber tip is employed as optical probe in the OCT. This probe generates a quality Bessel beam owning a large depth-of-field, ∼700 µm and small central spot size, ∼3 µm. The OCT system is probing the sample without using any microscopic lens. For experimental validation, the OCT imaging of chicken tissue has been obtained along with estimation of its refractive index and optical attenuation coefficient. Afterwards, the cancerous tissue is differentiated from the normal tissue based on the OCT imaging, refractive index, and optical attenuation coefficient. The respective tissue samples are collected from the human liver and pancreas. This probe could be a useful tool for endoscopic or minimal-invasive inspection of malignancy inside the tissue either at early-stage or during surgery.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Recent advancements in photonics technology have stimulated the growth in development of the novel optical approaches for clinical application especially in the field of oncology. The malignancy identification inside tissue is always a time-taking, labelled and complicated procedure (e.g., histopathology), owing to the partial resection of tissue specifically during surgery which can stimulate recurrence of the disease afterwards. Hence, it is imperative to develop a time-saving, and label-free malignancy detection method that can be applicable during surgery as well as early-stage diagnostics. Optical coherence tomography (OCT) [1] has been recently explored for imaging different kinds of cancers e.g., gastrointestinal [2], oral [3], brain [4], skin [5], cervical [6], breast [7], etc. It has the capability to visualize abnormalities [8] within complex biological tissue. OCT is a non-invasive and label-free approach which provides cross-sectional imaging with micron-order resolution and a few mm penetration depth. It also consumes less time and can investigate a larger area in comparison to the standard histopathology approach.
OCT systems usually suffer from a trade-off between the depth-of-field (DOF) and lateral resolution. This trade-off can be minimized by extending the DOF while simultaneously limiting the spot size of the beam. This can be achieved by employing a non-diffracting beam, namely the Bessel beam [9] which provides an extended DOF with a small central spot size. Different methods for Bessel beam generation are reported in the literature for OCT systems. A micro-optic axicon lens in free space [10] has offered ∼8 µm lateral resolution, and ∼400 µm DOF but it suffers from ambient noise and misalignment issue. A simple setup employing an in-fiber positive axicon tip [9] has provided ∼4.4 µm lateral resolution and ∼600 µm DOF. However, the axicon tip fabrication is easy but its prone to physical damage during probing the sample. An all-fiber optical probe made up of gradient-index (GRIN) lens, and no core fiber (NCF) [11] provides a ∼7 µm lateral resolution and ∼980 µm DOF, but the probe fabrication is relatively complex. Vairagi et. al. [12] has achieved a ∼700 μm DOF and ∼3.3 μm lateral resolution by using fiber-based negative axicon. This probe has simple fabrication approach, high reproducibility, longevity, and cost-effectiveness. The negative axicon probe also offers better trade-off between lateral resolution and DOF in comparison to the other Bessel beam probe. Hence, the negative axicon based OCT system is utilized in this work for tissue characterization.
However, an image only offers qualitative information of the sample about its morphology and anatomy. The quantitative knowledge in comparison to the qualitative information provides more accuracy. Therefore, the optical properties of tissue like refractive index (RI) and optical attenuation coefficient (OAC) have also been included in this study for characterizing tissue quantitatively. The progression of cancer shows a correlation with the RI of the tissue thus, it has capability to distinguish between the malignant and normal tissues [13,14]. Several studies in the literature have utilized RI for malignancy identification can be found in: 1) [15], the mean RI of pancreatic normal tissue as 1.372 and cancerous tissue as 1.396; 2) [16], mean RI of colorectal healthy tissue as 1.3473 and cancerous tissue as 1.3542; 3) [17], RI of the normal cells (n = 1.35–1.37) and cancerous cells (n = 1.39-1.40).
The OCT based tissue RI measurement methods [18,19] including (1) focus tracking, and (2) optical path length (OPL) have been widely adopted by the researchers due to label-free approach. However, these approaches either require tissue sample of limited thickness or not suitable for common-path configuration. The Gaussian beam is used by these methods for probing the sample. Therefore, the RI measurement using Fresnel equation and Bessel beam in common-path configuration as reported in Refs. [20,21] is explored for measuring tissue RI. Due to self-healing nature of the Bessel beam, their intensity profile in scattering media provides more ballistic photons in comparison to the Gaussian beam [22], making the Bessel beam ideal choice for RI measurement.
In addition, the OAC has also been studied for cancerous tissue. OCT based OAC estimation is generally obtained by using exponential fitting on the intensity decay observed in the OCT depth profiles [23,24]. Such data fitting methods are preferred only for homogenous samples. Hence, these approaches have limited practical implementation and offer poor resolution. To overcome this issue, the OAC measurement using a depth-resolved estimation (DRE) method [25] has been proposed which estimates the OAC corresponding to each pixel with higher resolution. However, the focal plane must be positioned above the sample [26] for accurate OAC measurement. This problem can be resolved by either considering the confocal function or using the beam having the large DOF such as, Bessel beam [27]. The DRE method is utilized for OAC estimation in our OCT system without considering the confocal function which makes the OAC calculation less complex.
In this paper, a common-path OCT system with an axicon probe is demonstrated for tissue characterization refer to the optical properties. Moreover, common-path OCT [28] has additive competency to limit the polarization mismatch, environmental vibration, dispersion effect, etc. A unique negative axicon structure is fabricated at the optical fiber tip which is spliced with the sample arm, generating a quality non-diffracting Bessel beam. This system does not require any microscopic objective lens as the beam is already focused. Initially, the chicken tissue is tested for validating our proposed OCT system to obtain OCT image, RI, and OAC. Subsequently, the OCT system is employed to differentiate the normal and cancerous tissue samples from human liver and pancreas on the basis of OCT imaging, RI and OAC. The rest of the works have been presented in the following sections as Experimental setup, Methodology, Material, Results and discussion, and Conclusion.
2. Experimental setup
2.1 Negative axicon probe
For fabrication of the negative axicon probe, the distal end of a germanium doped photosensitive single-mode optical fiber (Nufern GF4A) superficially touches the hydrofluoric (HF) acid solution which removes the material from the fiber core and the surrounding cladding, and travel upward into the core as long as the capillary action dominates. Thereafter, the solution recedes gradually and finally leaves behind the full-grown axicon in ∼40 minutes. It is a self-ending process. The dimensions of the 40 min axicon probe are typically measured as ∼290 µm long and ∼63 µm wide at the opening which is reduced gradually to ∼3 µm at the apex as shown in Fig. 1(a). The brief description about axicon fabrication can be found in our previous work [29]. There may be slight variation in external features of the probe depending upon the batch-to-batch optical fiber but it hardly impacts the quality of Bessel beam from the axicon. However, we have demonstrated axicon length optimization by controlling the etching time under strict environmental condition, such as temperature [30]. It is observed that the probe etched at 26°C for 7 min provides an optimized axicon length. Still, it does not show significant change in the generated Bessel beam except its location from the axicon tip. However, the 7 min etching optimization cannot be guaranteed due to the sensitivity of the process to the slight variation in ambient temperature. Even a small fluctuation in the temperature can significantly alter the etching rate of the optical fiber in HF, affecting the axicon length. The full grown axicon is independent from the variation of such thermal condition, therefore it provides better reproducibility. A typical picture of the transverse profile of Bessel beam captured through camera for Helium-neon laser source (633 nm) and Thorlabs beam profiler for super-luminescent diode source (840 nm) are shown in Fig. 1(b) and (c), respectively. A full grown axicon structure delivers the quality Bessel beam having a large DOF, ∼700 µm at 808 nm [12] with ∼3 µm central spot size.
Fig. 1. (a) Microscopic picture of a negative axicon probe. Transverse beam profile of the Bessel beam recorded at the (b) camera through a screen for Helium-neon laser source (632 nm) and (c) Thorlabs beam profiler for super-luminescent diode source (840 nm).
2.2 Common-path optical coherence tomography (OCT) system
Figure 2 illustrates a common-path spectral-domain OCT system with a negative axicon optical probe. Briefly, a super-luminescent diode (Superlum) with a central wavelength of 840 nm and full-width half maxima (FWHM) of 45 nm is used as a light source to illuminate the tissue sample. The source is coupled to Port 1 of an optical circulator from Dk photonics with operating wavelength, 850 nm ± 100 nm. The tissue sample is placed inside the sample holder fixed to the 3-axis translational stage (NanoMax300 Thorlabs) as shown in Fig. 2. The top surface of the sample holder is made up of the glass coverslip, ∼150 μm. At the distal end of the sample arm, the negative axicon probe is spliced and fixed inside the optical fiber holder which is mounted on the customized translation stage with micron-order resolution. It helps in placing the probe vertically near to the sample. The distance between the sample and probe is adjusted at a position to obtain the maximum interference. While, the optical probe is fixed, the sample is scanned by moving the translational stage. The interference spectra are acquired from the spectrometer and synchronised with 3-axis stage using the customised program written on the LabVIEW platform. The further processing of the data has been done in the MATLAB@2016 software to obtain the cross-sectional OCT image, and optical properties of the tissue.
Fig. 2. Schematic of the experimental setup demonstrating common-path optical coherence tomography system. TS: 3-axis Translational Stage; CL: Collimating Lens; DF: Diffraction grating; FL: Focusing Lens; LSC: Line Scan Camera.
3. Methodology
In this section, the methodology for the image construction, refractive index and optical attenuation coefficient measurement are described.
3.1 Image construction
Fast Fourier transform (FFT) based image reconstruction method has been applied on the acquired interference spectra to obtain the sample cross-sectional image [31]. The detected interference spectrum (discrete form) can be described as:
where, ${R_s}$: intensity field reflected from the sample; ${R_r}$: intensity field from reference i.e., air-glass interface; $n$: refractive index of sample; $2z$: path difference between the sample and reference beams.The presence of the reference or dc information can cause the fixed-pattern noise, resulting into bold horizontal lines in the constructed image. Therefore, the mean subtraction method is applied for removal of fixed-pattern noise [32]. To reconstruct the A-scan in terms of depth, the spectrum is re-sampled from the wavelength domain into the linearly sampled wavenumber domain (k-space) before applying the FFT [31], as follows:
Here, integer i = 0,1,2,3……(N-1); $k$: wavenumber; ${\lambda _{max}}$: maximum wavelength; ${\lambda _{min}}$: minimum wavelength.
Afterwards, the FFT is applied to the interpolated data to obtain the axial depth profile of the sample, considering only the magnitude of complex FFT. Subsequently, each FFT point represents a single A-scan and multiple A-scans in the lateral direction are combined to form a 2D cross-section image known as a B-scan.
3.2 Refractive index (RI) measurement
RI is an essential optical property which determines how much light reflect or refract while propagating through a media. Here, we are using the Fresnel equation for RI calculation as stated in Ref. [20]. According to Fresnel, the relationship between two media for 0° incidence angle is descried below:
The FFT is applied to the acquired interference spectrum which provides two distinct peaks from air-glass and glass-tissue for an A-scan as represented in Fig. 3. The first peak indicates the reflectance from the air-glass interface only because the glass is homogenous medium. Hence the reflectance within the glass coverslip is negligible, leading to a small peak width. The second peak is generated from the glass-tissue, having also the reflectance from within tissue sample due to its heterogeneity. As the refractive index within the sample does not vary much, resulting in the absence of separate peaks from internal interfaces. Therefore, the entire peak is considered and the measured refractive index indicates the arage value. The following steps are involved to evaluate the tissue RI:
- - The reflected powers corresponding to air-glass and glass-tissue are computed by squaring and summing the intensity points under the respective peak.
- - The reflected power from air-glass interface is utilized in Eq. (3) to obtain the incident powers on the glass-tissue because the reflectance at air-glass is ∼4% for ${n_{air}}$=1 and ${n_{glass}}$. = 1.52.
- - Afterwards, the reflected power corresponding to glass-tissue is corrected by incorporating the loss. Then, RI of the tissue is calculated by employing the incident and corrected reflected power on glass-tissue with the known RI of the glass in Eq. (3).
Fig. 3. A typical A-scan, resulting into two reflected peaks corresponding to the air-glass and glass-tissue interfaces as marked by the circle.
For every single A-scan, a corresponding RI value is calculated and the mean RI for the sample is obtained after removing outliers using the boxplot as further discussed in section 5. Here, the calculated RI is the average RI of the sample along the imaging depth. The change in RI for incident angle 1° and 2° instead of 0° have been estimated for intralipid solution (highly scattering) as a sample. It is found that the change in RI for 1° and 2° are in the scale of 3-4 decimal and 2-3 decimal, respectively as compared to the 0° incident angle. The change in 3-4 decimal of the measured RI is considerable for the scattering media. Also, it is not challenging task to maintain the incidence angle less than or equal to 1°. Hereafter, the Fresnel equation-based RI method is implemented on the chicken and human tissue samples in section 5.
3.3 Optical attenuation coefficient (OAC) estimation
We have applied the DRE method [25] to determine the OAC of the sample tissue. This method assumes that the maximum light gets attenuated inside the tissue imaging range. The model in the DRE method is given below:
The OAC estimation is explained in the following steps as illustrated in Fig. 4 through a flow diagram.
- 1. The tissue surface is identified with respect to the background for each A-scan.
- 2. The tissue is separated from the background information using a mask.
- 3. For each A-scan, the normalized sensitivity losss is calculated along the depth profile. Then, the acquired signal, I in Eq. (4) is corrected by adjusting the loss.
- 4. The OAC, $\mu $ is calculated using Eq. (5) for each A-scan.
- 5. For a B-scan, the mean value of the OAC is obtained row by row up to the imaging depth.
Fig. 4. Flow diagram of the steps for estimating the optical attenuation coefficient using DRE method.
Later, the obtained OAC values are plotted using a boxplot and the mean values are considered after removing the outliers. The influence of focal position is negligible for the larger Rayleigh range or DOF [27], therefore, it is not considered here because the Bessel beam has a large DOF as mentioned in Section 2.
For validation of the OAC estimation on the standard sample, we have utilized various concentration of intralipid from 1-5%, prepared by diluting the intralipid 20% (Fresenius-Kabi) with distilled water. Estimated OAC for various intralipid concentrations is given in Table 1 which is also compared with the extrapolated OAC values from Ref. [33].
Table 1. Comparison of estimated OAC with reported OAC for various intralipid concentration.
4. Material
4.1 Sample preparation
The fresh chicken tissue is brought from the market before each experiment and the experiments are conducted within 30-60 minutes at room temperature. Human tissue samples are collected from the Post Graduate Institute of Medical Education & Research (PGMIER), Chandigarh, within 2-3 hour after the tissue is removed from the body, and kept at room temperature in formalin. The sample is also preserved in a formalin solution after the experiments for histology to be conducted later for establishing ground truth.
4.2 Histopathology
To verify the whether the tissue is malignant or not, the histology examinations of the samples are conducted. The tissue slices are sectioned at the same locations from where a B-scan is acquired. Afterwards, the tissue samples are fixed with formalin, and these fixed samples are sent to the histopathologist who carried out a standard histology process. In this process, the tissue slices are stained with hematoxylin-eosin (H&E) dye, then corresponding glass-slides with tissue imprints are examined under the microscope to obtain the histological images.
5. Results and discussion
Before starting the real-time experiments, the system performance parameters are computed. Theoretically, the lateral resolution of our system is ∼3 µm which is equivalent to the central spot size of the Bessel beam. Experimentally, the lateral resolution of our system is found to be ∼3.3 μm after scanning the transmission grating having 7500 lines per inch, which is in accordance with theoretical lateral resolution (see Supplement 1). The theoretical axial resolution of our system in the air is ∼6.9 μm whereas, the experimentally axial resolution of our system is found to be ∼8μm (see Supplement 1). At the spectrometer-end, the measured reference beam power is 5 µW, and the sample beam power is drop by only ∼2 µW over the 1 mm distance.
The speed of data acquisition is 10KHz A-scans. The point spread function (PSF) is determined with respect to different distances for the broadband mirror as a sample and the respective plot is illustrated in Fig. 5. It is observed that the peak intensity initially increases after reaching the maximum value, then it starts decaying gradually because the Bessel beam starts forming after some distance from the tip end. It is found that the sensitivity loss of our system is ∼5 dB for 1 mm as shown in Fig. 5, which is calculated by after the maximum intensity value.
For validation of our OCT system, the fresh chicken muscle tissue is tested to obtain the cross-sectional OCT image and RI along with the OAC. It is observed in Fig. 6 that the three layers of the cross-section chicken tissue are distinctly visible, as reported in the literature [34].
Fig. 6. The cross-sectional image of chicken muscle tissue from our OCT system showing the epidermis and dermis layer. Scale bar: 100μm.
The RI of the chicken muscle tissue is estimated by solving the Fresnel relation stated in Eq. (3) as described in the subsection 3.2. In Table 2, the mean RI and mean OAC from different chicken muscle tissue samples are calculated after removing the outliers using the boxplot. Five different samples of the chicken having similar breed have been taken to show the deviation in the obtained RI and OAC values. The RI of the chicken tissue varies from the 1.39 to 1.44 as in Ref. [35,36]. The mean RI of the chicken tissue from our system is estimated as ∼1.41 for all samples with standard deviation, 0.012. Similarly, the OAC is also calculated by processing the spectrum using DRE method. The mean OAC for all 5 samples is obtained as ∼2.40 with standard deviation, 0.36. These results indicate that our system is efficiently working for estimating the RI and OAC of the chicken tissue.
Table 2. Refractive index (RI) and optical attenuation coefficient (OAC) for muscle tissue from different chicken sample.
Furthermore, the different human tissues with both normal and malignant types are examined by our OCT system. The cross-sectional OCT images of the human liver tissues along with the corresponding histopathological images are represented in Fig. 7. Both the normal and malignant tissue samples are collected from PGIMER, Chandigarh after surgery from different patients detected with metastatic malignancy, as per their biopsy reports. Both normal and malignant tissues of single patient are further manually divided into the sub-samples of ∼5mm2 in size with a thickness of ∼5 mm. For normal liver sample, both the images from OCT system and histopathology exhibit a valley like structure at the surface of the sample as marked by a rectangular box, Fig. 7(a) and (b). The marking is done manually to represent that the histopathology and OCT images are obtained from the same location.
Fig. 7. The cross-sectional histopathology images of (a) normal, and (c) malignant human liver tissue. The cross-sectional images from the proposed OCT system of (b) normal, and (d) malignant human liver tissue. Scale bar: 100 µm.
The presence of malignancy in the human liver sample is confirmed by histopathologist from PGIMER. The image from our OCT system of the malignant human liver tissue along with its histology image is illustrated in Fig. 7(c) and (d). It is observed that the malignant regions are present in both the images which are marked by rectangular and square boxes. The regions marked by boxes are identified with the help of histopathologist. The histology image is zoomed in as represented in Fig. 7(a) and (c) to compare the maximum area with OCT images. It shows that the histology images are taken from the same position where the B-scan has been acquired. Hence, our system can effectively visualize the abnormalities existing in the tissue structure without using a microscopic objective lens.
However, one could not identify whether the tissue is malignant or not by visualizing the OCT image only. Therefore, it would be helpful to study the quantitative information like RI and OAC which are considered in this work. The mean RI for each sample is calculated after the removing the outliers using boxplot as presented in Fig. 8(b). The mean RI of the malignant and normal liver tissues obtained from our OCT system are 1.3837 ± 0.0071 and 1.3694 ± 0.0053, respectively as presented in Table 3 with the other statistical parameters including maximum, minimum, and interquartile range (IQR). The RI of the normal human liver tissue is in close match with the value reported in the literature [37]. The RI of malignant tissue in our case is found to be on the higher side than the normal tissue which is in accordance with the Ref. [13].
Fig. 8. Boxplot of the (a) optical attenuation coefficient (OAC) and (b) refractive index (RI) for human liver and pancreas tissues. LM: Liver Malignant; LN: Liver Normal; PM: Pancreas Malignant; PN: Pancreas Normal.
Table 3. Statistical parameters for the refractive index and optical attenuation coefficient of human liver and pancreas tissue samples.a
Similarly, the mean OAC is estimated using the DRE method as described in subsection 3.3 for every sample after removing the outliers. The corresponding variation of the OAC values for both malignant and normal tissue types are also presented in the form of boxplot in Fig. 8(a). The OAC of the human liver for normal and malignant tissue are estimated as 3.22 ± 0.45 and 1.97 ± 0.50, respectively. Using DRE method, the estimated OAC of the malignant tissue is found to be less than the normal tissue. The similar trend is reported in the literature [4].
In order to verify this study further, the normal and malignant human pancreatic tissue having pancreatic adenocarcinoma from different patients are considered for analysis. The samples collected from PGIMER of single patient are further dissected into sub-samples of ∼5 mm2 with ∼5 mm thickness for experimentation. In Fig. 9, the cross-sectional image from the OCT system for (a) normal pancreatic tissue and (b) malignant pancreatic tissue (pancreatic adenocarcinoma) are presented. The sample in Fig. 9(b) is completely malignant as verified by the histopathology report obtained from PGIMER. The marked areas in yellow boxes are highly malignant which are identified with the help of histopathologist.
Fig. 9. The cross-sectional image of (a) normal human pancreas tissue along with the (b) malignant human pancreas tissue (pancreatic adenocarcinoma) from the OCT system. Scale bar: 100 µm.
The boxplot of the obtained RI and OAC values for both malignant and normal pancreatic tissue samples are shown in Fig. 8. The statistical values of the RI and OAC for the human pancreas tissues are presented in Table 3. The mean RI of the human normal and malignant pancreatic tissue are obtained as 1.3617 ± 0.0063 and 1.3861 ± 0.0049, respectively which are approximately close to the reported RI values in Ref. [15]. The mean OAC of the human normal and malignant pancreas tissue are also estimated as 3.65 ± 0.52 and 2.05 ± 0.48, respectively. From Table 3, it is observed that the malignant tissue has higher RI than the respective normal tissue. The normal tissue of human liver and pancreas offers higher OAC in comparison to the malignant tissue. It implies that our system can differentiate the cancerous/malignant tissue from the normal tissue of different human organs both qualitatively and quantitively.
6. Conclusion
In summary, we have demonstrated the axicon probe based common-path optical coherence tomography system on chicken and ex-vivo human tissues. The probe has a unique negative axicon structure which produces a quality Bessel beam with a large depth-of-field ∼700 µm and central spot size, ∼3 µm. Our OCT system provides lateral resolution, ∼3.3 µm and axial resolution, ∼8 µm experimentally. The system is not using any microscopic lens for probing the tissue sample. In this work, the quantitative parameter like refractive index and optical attenuation coefficient are also taken into consideration for differentiating the cancerous tissue from normal tissue. The malignancy inside the human liver and pancreatic tissues are visualized by OCT imaging. The malignant tissue is differentiated quantitively using refractive index and optical attenuation coefficient based on the statistical parameters from normal tissue. The common-path and negative axicon probe configuration has potential to be an endoscopic or minimal-invasive probe in future for identifying cancerous tissue either at early-stage or during surgery.
Funding
Council of Scientific and Industrial Research, India (6/1/FIRST/2020-RPPBDD-TMD-SeMI).
Acknowledgments
The author, Pooja Gupta, would like to thank the Council of Scientific & Industrial Research, New Delhi, India, for Research Associate fellowship. We would like to thank Dr. Harjeet Singh from PGIMER, Chandigarh for helping us in providing the human samples.
Disclosures
The authors declare no conflict of interest.
Data availability
The data presented in this study are not publicly available and will be provided on request by the corresponding author.
Supplemental document
See Supplement 1 for supporting content.
References
1. J. M. Schmitt, “Optical Coherence Tomography (OCT): a review,” IEEE J. Sel. Top. Quantum Electron. 5(4), 1205–1215 (1999). [CrossRef]
2. H. Luo, S. Li, Y. Zeng, et al., “Human colorectal cancer tissue assessment using optical coherence tomography catheter and deep learning,” J. Biophotonics 15(6), e202100349 (2022). [CrossRef]
3. A. E. Heidari, S. P. Sunny, B. L. James, et al., “Optical Coherence Tomography as an Oral Cancer Screening Adjunct in a Low Resource Settings,” IEEE J. Sel. Top. Quantum Electron. 25(1), 1–8 (2018). [CrossRef]
4. W. Yuan, C. Kut, W. Liang, et al., “Robust and fast characterization of OCT-based optical attenuation using a novel frequency-domain algorithm for brain cancer detection,” Sci. Rep. 7, 44909 (2017). [CrossRef]
5. L. van Manen, J. Dijkstra, C. Boccara, et al., “The clinical usefulness of optical coherence tomography during cancer interventions,” J. Cancer Res. Clin. Oncol. 144(10), 1967–1990 (2018). [CrossRef]
6. Y. Gan, W. Yao, K. M. Myers, et al., “Analyzing three-dimensional ultrastructure of human cervical tissue using optical coherence tomography,” Biomed. Opt. Express 6(4), 1090 (2015). [CrossRef]
7. A. Butola, A. Ahmad, V. Dubey, et al., “Volumetric analysis of breast cancer tissues using machine learning and swept-source optical coherence tomography,” Appl. Opt. 58(5), A135 (2019). [CrossRef]
8. S. A. Boppart, D. L. Marks, A. L. Oldenburg, et al., “Optical coherence tomography: a review of clinical development from bench to bedside,” J. Biomed. Opt. 12(5), 051403 (2007). [CrossRef]
9. K. M. Tan, M. Mazilu, T. H. Chow, et al., “In-fiber common-path optical coherence tomography using a conical-tip fiber,” Opt. Express 17(4), 2231–2235 (2009). [CrossRef]
10. K.-S. Lee and J. P. Rolland, “Bessel beam spectral-domain high-resolution optical coherence tomography with micro-optic axicon providing extended focusing range,” Opt. Lett. 33(15), 1696 (2008). [CrossRef]
11. D. Lorenser, X. Yang, and D. D. Sampson, “Ultrathin fiber probes with extended depth of focus for optical coherence tomography,” Opt. Lett. 37(10), 1616 (2012). [CrossRef]
12. K. Vairagi, A. Pandey, P. Gupta, et al., “Common-path optical coherence tomography using the bessel beam from negative axicon optical fiber tip,” IEEE J. Sel. Top. Quantum Electron. 25(1), 1–6 (2019). [CrossRef]
13. Z. Wang, “Tissue refractive index as marker of disease,” J. Biomed. Opt. 16(11), 1 (2011). [CrossRef]
14. R. Khan, B. Gul, S. Khan, et al., “Refractive index of biological tissues: Review, measurement techniques, and applications,” Photodiagn. Photodyn. Ther. 33, 102192 (2021). [CrossRef]
15. S. Y. Lee, W. R. Lloyd, M. Chandra, et al., “Characterizing human pancreatic cancer precursor using quantitative tissue optical spectroscopy,” Biomed. Opt. Express 4(12), 2828 (2013). [CrossRef]
16. S. Carvalho, N. Gueiral, E. Nogueira, et al., “Wavelength dependence of the refractive index of human colorectal tissues: comparison between healthy mucosa and cancer,” J. Biomed. Photonics Eng. 2(4), 040307 (2016). [CrossRef]
17. X. J. Liang, A. Q. Liu, C. S. Lim, et al., “Determining refractive index of single living cell using an integrated microchip,” Sens. Actuators, A 133(2), 349–354 (2007). [CrossRef]
18. G. J. Tearney, M. E. Brezinski, B. E. Bouma, et al., “Determination of the refractive index of highly scattering human tissue by optical coherence tomography,” Opt. Lett. 20(21), 2258 (1995). [CrossRef]
19. A. Knüttel and M. Boehlau-Godau, “Spatially confined and temporally resolved refractive index and scattering evaluation in human skin performed with optical coherence tomography,” J. Biomed. Opt. 5(1), 83 (2000). [CrossRef]
20. P. Gupta, A. Pandey, K. Vairagi, et al., “Solving Fresnel equation for refractive index using reflected optical power obtained from Bessel beam interferometry,” Rev. Sci. Instrum. 90(1), 015110 (2019). [CrossRef]
21. A. Pandey, P. Gupta, K. Vairagi, et al., “Packaged negative axicon optical fiber probe and Bessel beam interferometry for refractive index measurement of hazardous liquid samples,” J. Lightwave Technol. 37(24), 6121–6126 (2019). [CrossRef]
22. C. Gohn-Kreuz and A. Rohrbach, “Light needles in scattering media using self-reconstructing beams and the STED principle,” Optica 4(9), 1134 (2017). [CrossRef]
23. D. J. Faber, F. J. van der Meer, M. C. G. Aalders, et al., “Quantitative measurement of attenuation coefficients of weakly scattering media using optical coherence tomography,” Opt. Express 12(19), 4353 (2004). [CrossRef]
24. E. C. C. Cauberg, D. M. de Bruin, D. J. Faber, et al., “Quantitative measurement of attenuation coefficients of bladder biopsies using optical coherence tomography for grading urothelial carcinoma of the bladder,” J. Biomed. Opt. 15(6), 066013 (2010). [CrossRef]
25. K. A. Vermeer, J. Mo, J. J. A. Weda, et al., “Depth-resolved model-based reconstruction of attenuation coefficients in optical coherence tomography,” Biomed. Opt. Express 5(1), 322 (2014). [CrossRef]
26. G. T. Smith, N. Dwork, D. O’Connor, et al., “Automated, Depth-Resolved Estimation of the Attenuation Coefficient from Optical Coherence Tomography Data,” IEEE Trans. Med. Imaging 34(12), 2592–2602 (2015). [CrossRef]
27. B. Ghafaryasl, K. A. Vermeer, J. Kalkman, et al., “Attenuation coefficient estimation in Fourier-domain OCT of multi-layered phantoms,” Biomed. Opt. Express 12(5), 2744 (2021). [CrossRef]
28. A. B. Vakhtin, D. J. Kane, W. R. Wood, et al., “Common-path interferometer for frequency-domain optical coherence tomography,” Appl. Opt. 42(34), 6953 (2003). [CrossRef]
29. K. Vairagi, R. A. Minz, S. Kaur, et al., “Deep Seated Negative Axicon in Selective Optical Fiber Tip and Collimated Bessel Beam,” IEEE Photonics Technol. Lett. 29(10), 786–789 (2017). [CrossRef]
30. K., Vairagi, P. Gupta, S. A. Mondal, et al., “Reflective axicon based energy-efficient extended depth of focus quasi-Bessel beam probe for common-path optical coherence tomography,” Applied Optics 62(3), 511–517 (2023). [CrossRef]
31. M. Ali and R. Parlapalli, “Signal processing overview of optical coherence tomography systems for medical imaging,” Texas Instruments Whitepapers, 1–22 (2010).
32. S. Moon, S.-W. Lee, and Z. Chen, “Reference spectrum extraction and fixed-pattern noise removal in optical coherence tomography,” Opt. Express 18(24), 24395 (2010). [CrossRef]
33. B. Aernouts, R. Van Beers, R. Watté, et al., “Dependent scattering in Intralipid® phantoms in the 600-1850nm range,” Opt. Express 22(5), 6086 (2014). [CrossRef]
34. H. B. Lillywhite, “Water relations of tetrapod integument,” J. Exp. Biol. 209(2), 202–226 (2006). [CrossRef]
35. M. Ohmi, Y. Ohnishi, K. Yoden, et al., “In vitro simultaneous measurement of refractive index and thickness of biological tissue by the low coherence interferometry,” IEEE Trans. Biomed. Eng. 47(9), 1266–1270 (2000). [CrossRef]
36. P. Sun and H. Sun, “Determination of the anisotropy complex refractive indices of chicken tissues in vitro at 650 nm,” JEOS:RP 5, 10030 (2010). [CrossRef]
37. P. Giannios, K. G. Toutouzas, M. Matiatou, et al., “Visible to near-infrared refractive properties of freshly-excised human-liver tissues: Marking hepatic malignancies,” Sci. Rep. 6(1), 27910 (2016). [CrossRef]
