Raman spectroscopy method and system

Abstract

The present disclosure provides a system and method for obtaining structural information relating to tissue. More particularly, the present disclosure provides an optical system which utilizes a long focal length lens arrangement to focus light incident on a needle probe. Focusing the light across a distance, before the light reaches the needle probe ensures that the maximum energy is retained within the light without introducing background noise signal from additional lenses or telescopes. In turn, this enables the miniaturization of the needle probe, therefore, making the system applicable for use in vivo examinations. The disclosure also describes a method and system for quantitative Raman spectral analysis of tissue which provides absolute biochemical compositions of tissue.

Description

[0001] RAMAN SPECTROSCOPY METHOD AND SYSTEM

[0002] Technical Field

[0003] The present disclosure generally relates to methods and systems for use in Raman spectroscopy techniques. More specifically, part 1 of the present disclosure relates to methods and systems for non-polarized and polarized Raman spectroscopy, the methods and systems utilizing long distance working lens and a thin hollow needle. Part 2 of the present disclosure relates to a method and system for Raman spectroscopy that is able to undertake quantitative Raman histology to obtain absolute biochemical concentration values.

[0004] Part 1 of the present disclosure seeks to improve the problem wherein the methods and systems for polarized Raman spectroscopy are unsuitable for in vivo examination due to their relative size. In other words, the needle probe of polarized Raman spectroscopy must be thin enough to enable compatibility with patient examination. In contrast, part 2 is agnostic to the specific needle probe used, and any Raman probe can be used, with the processing of the gathered Raman information then being used in the embodiments of part 2 of the present disclosure to obtain quantitative Raman histology information.

[0005] Background and Related Art

[0006] Background to Part 1

[0007] Osteoarthritis (OA) is a chronic, debilitating painful condition, characterized by the structural degeneration of articular cartilage, the load-bearing connective tissue that lines the ends of long bones. It is the most widespread cause of disability in adults; over 20% of the adult population of the US, or over 50 million individuals, are afflicted with the disease and its incidence is predicted to rise sharply over coming decades. Currently, no clinical therapy exists to halt the progression of the disease.

[0008] Articular cartilage degeneration during osteoarthritis (OA) occurs in stages. Early stage degeneration is marked by difficult-to-detect tissue changes: an initial loss of glycosaminoglycans (GAG) from the topmost cartilage layers and disorganization (alignment loss) of collagen at the articular surface. This is followed by far more substantial physical erosion of the collagen matrix, until bone-on-bone contact is reached. Interestingly, the early stage of the disease, before substantial tissue erosion has occurred, represents a critical clinical window, when intervention strategies (e.g., drug therapies, physical therapy, lifestyle changes) may be most effective at reversing the onset of cartilage degeneration. However, the ability to diagnose early-stage OA remains a considerable clinical challenge; conventional imaging platforms (e.g., radiography, CT, MRI) are predominantly suited for diagnosing late stages of the disease due to lack of resolution and molecular specificity. For this reason, there is a great need to introduce novel biomolecular sensitive optical techniques for the in vivo early-stage diagnosis of OA.

[0009] The structure and composition of hyaline cartilage is optimized for its mechanical performance. It is comprised of a type-II collagen (COL) fibril network that affords structure and tensile strength, complemented by a negatively charged sulfated glycosaminoglycan (GAG) matrix that provides compressive properties and retains interstitial water. More than 90% of applied joint load is supported by pressurization of entrapped water (interstitial fluid load support), yielding the tissue's characteristic low frictional properties. Additionally, cartilage is inhomogeneous and structurally anisotropic, where the collagen configuration varies with depth (segregated into zones) and is optimized for the mechanical loading of each zone: superficial zone (SZ) - comprised of collagen arranged parallel to the articular surface and resists surface shear and improves gliding; transitional (middle) zone (MZ) - composed of mix-aligned collagen and GAGs responsible for generating interstitial fluid load support; deep radial zone (DZ) - collagen fibers perpendicular to the subchondral plate, anchors cartilage to subchondral bone. During OA, degradation of cartilage occurs in stages. Early on, GAG is depleted from the SZ with concomitant loss of superficial collagen fiber organization and alignment. Loss of GAG and COL alignment reduces fluid load support, transferring loads to the collagen matrix, and leading to cartilage erosion through the MZ and DZ, culminating in significant cartilage volume loss until bone-on-bone contact is reached.

[0010] As a major clinical challenge, changes in joint structure and function that account for OA morbidity and disability appear relatively late in the disease process. Currently, OA is diagnosed from clinical symptoms (pain, swelling, impaired function) and image-based assessments (radiographs and magnetic resonance imaging (MRI)) that are biased towards late-stage OA pathoanatomy (cartilage volume loss, bone marrow edema, subchondral bone thickening, cysts, and osteophytes) 10. The irreversible breakdown of cartilage occurs before clinical symptoms and radiographic signs are evident. Therefore, OA diagnosed at a late stage, after changes in tissue structure have transpired, restricts treatment options. The inability to identify mild or early cartilage damage, when therapeutic strategies will be most effective, remains a significant clinical obstacle.

[0011] Raman spectroscopy is an inelastic light scattering technique that offers an in vivo optical biopsy of tissues at the molecular level. Using fiber-optic probes, Raman spectra can be obtained from tissues in vivo (e.g., by the use of endoscopes). The Raman spectrum scales linearly with concentrations and therefore contains a wealth of quantitative information, allowing for the potential extraction of contents of key extracellular matrix (ECM) constituents in the tissue. At National University of Singapore (NUS), Dr Bergholt pioneered real-time "Raman endoscopy" for non-invasive in vivo "optical biopsy" in the gastrointestinal (GI) tract (Gastroenterology 2014: IF 20.877). This technique was applied in more than 800 patients over 5 years, lead to 3 patent applications that has recently been commercialised in form of a medical device for endoscopy (IMDX™, Endofotonics Pte Ltd). In turn, this work allowed expansion of this system as a platform technology to a plurality of organs including: oral cavity, nasopharynx, larynx, esophagus, gastric and colon in humans (J Raman spectroscopy, 2012; J Biomed Opt, 2012; J Biophotonics, 2016a). In contrast, there exists no clinical Raman technique for the diagnosis of OA.

[0012] In more detail, Raman spectroscopy is predicated on the inelastic scattering of photons. When monochromic laser light induces a change in molecular polarizability during vibrations, a small proportion of the incident photons (~1 in 108) are scattered with a change in wavelength. The Raman scattered light indicates the vibrational modes of constituent molecules; the absorbed energy corresponds to specific Raman active vibrational modes that define a molecule's "fingerprint". Therefore, the Raman spectra of cartilage carry information about individual molecular vibrational bonds that correspond to specific biochemical building blocks (amides, sulfates, carboxylic acids, hydroxyls) of the constituents of hyaline cartilage (GAG, COL, H2O). Prior work demonstrates that Raman spectra of cartilage exhibit statistical changes in response to mechanical damage and OA. However, the implementation of Raman spectroscopy as a diagnostic tool for cartilage health has been impeded by: 1) lack of clinically compatible intra-articular, fiber-optic Raman needle probes for in vivo diagnostics, and 2) inability to extract specific and quantitative biochemical and structural metrics diagnostic of early stage OA. As a consequence, to date, a platform capable of achieving in vivo Raman diagnostics of early-OA has not been demonstrated.

[0013] The key advantage of polarized Raman spectroscopy compared to conventional Raman spectroscopy is that it offers additional specific information about molecular organization / symmetries and tissue structure.

[0014] Our previous application US 2023 / 341330 Al (the entirety of which is incorporated herein by reference) presented a Raman platform that provides multiplexed polarized Raman signals. This system used a tightly focused lens to uniquely preserve polarisation information from the shallow tissue surface that otherwise would be scrambled. By multiplexing the two polarizations, it allowed for measurement of polarized spectra in clinically relevant time scales. There is however a key challenge that needs to be addressed to translate polarized Raman spectroscopy into in vivo examinations: The needle probe needs to be thin to be compatible with patient examination.

[0015] Background to Part 2

[0016] Quantitative histology is essential for understanding the structural and biochemical composition of connective tissues, which play critical roles in maintaining the mechanical integrity of various organs and tissue parts. Connective tissues, including cartilage, bone, tendons, and ligaments, exhibit unique compositions of extracellular matrix (ECM) proteins such as collagen, elastin, and glycosaminoglycans (GAG). These components govern their biomechanical properties and are dynamically altered in response to physiological stress, aging, and pathological conditions like fibrosis, osteoarthritis, and connective tissue disorders. Accurate characterization of these changes at the molecular level is key to understanding disease mechanisms, monitoring therapeutic efficacy, and developing regenerative strategies. For instance, the composition of healthy articular cartilage is optimized for its mechanical performance, comprised of water and negatively charged sulfated GAG constrained by a fibrous type-II collagen network (COL). Early-stage cartilage degeneration is characterized by changes in the composition of the tissue, notably loss of sulfated GAG and / or tissue swelling contributing to mechanical softening, followed by more severe erosion of the collagen network.

[0017] Traditional histological staining techniques of tissue sections (e.g., hematoxylin, eosin, alcian blue, safranin o, picrosirius red) is widely utilized for, such as picrosirius red or alcian blue staining, provide limited obtaining insights into tissue architecture ECM distribution and cellular components. However, these techniques are plagued by a host of limitations including, 1) requirement of laborious preprocessing, 2) artifacts generated by processing protocols (e.g., fixation, embedding, staining), and 3) there inability to provide absolute quantifications of tissue composition— these techniques, are at best semi-quantitative, and limited in their ability to quantify biochemical composition. Additionally alternative, other techniques such as biochemical assays require tissue homogenization, losing spatial information about the tissue structure distribution. These limitations underscore the need for a non-destructive, spatially resolved method to analyze quantify the composition of connective tissues composition quantitatively.

[0018] Raman spectroscopy is an inelastic light scattering technique that yields a non-destructive, quantitative, point-wise optical fingerprint of a tissue's molecular building blocks (amides, sulfates, carboxylic acids, and hydroxyls), thus allowing recognition of the predominant molecular constituents of articular cartilage: glycosaminoglycans (GAG), collagen (COL), and water (H2O). However, Raman spectroscopy is generally not quantitative for tissue samples due to tissue light absorption and scattering. For this reason, it is common to normalize Raman spectra (to the integrated area under the Raman curve) thereby losing any quantitative information.

[0019] Summary of the Invention

[0020] Summary of Part 1 In part 1, the present disclosure provides a method and system for use in Raman spectroscopy, the method and system utilizing a long working distance lens and a thin needle probe, wherein the long distance working lens is arranged to focus incident light into the back aperture of the thin needle probe.

[0021] Further, the present disclosure provides a system and method for obtaining structural information relating to tissue. More particularly, the present disclosure provides an optical system which utilizes a long focal length lens arrangement to focus light incident on a needle probe. Focusing the light across a distance, before the light reaches the needle probe ensures that the maximum energy is retained within the light without introducing background signal from additional lenses or telescopes. In turn, this enables the miniaturization of the needle probe, therefore, making the system applicable for use in vivo examinations.

[0022] More particularly, an embodiment of the present disclosure provides a system for obtaining structural information relating to tissue, the system may comprise: a needle probe projecting from a probe housing body, a probe housing body comprising a long focal length lens arrangement, the long focal length lens arrangement positioned within the probe housing body, distal from the junction of the needle probe and the probe housing body; the needle probe being configured to direct polarized light to the tissue and collect reflected Raman scattered light; a beam splitter configured to split the Raman scattering into two polarized components; and a spectrometer configured to simultaneously and separately image the two polarized components to produce two Raman spectra.

[0023] An embodiment of the present disclosure provides a method for obtaining structural information relating to tissue, the method may comprise: directing polarized light to the tissue using a needle probe; focusing the polarized light onto the tissue using a long focal length lens arrangement such that the polarized light is reflected from the tissue producing Raman scattering, wherein the long focal length lens arrangement is positioned within a probe housing body, distal from the junction of the needle probe and the probe housing body; collecting the Raman scattering using the needle probe; splitting the Raman scattering into two polarised components using a beam splitter; simultaneously and separately imaging the two polarised components using a spectrometer to produce two Raman spectra; and processing and analysing the two produced Raman spectra using a computer program to obtain structural information relating to the tissue.

[0024] An aspect of the present disclosure provides a system for obtaining structural information relating to tissue, the system comprising one or more of: a needle probe projecting from a probe housing body, a probe housing body comprising a long focal length lens arrangement, the long focal length lens arrangement positioned within the probe housing body, distal from the junction of the needle probe and the probe housing body; the needle probe being configured to direct light to the tissue and collect reflected Raman scattered light; and a spectrometer configured to image the reflected Raman scattered light to produce at least one Raman spectrum.

[0025] The long focal length lens arrangement may have a focal length substantially equal to the distance between the long focal length lens arrangement and the junction of the needle probe and the probe housing body.

[0026] The needle probe may comprise a short focal length lens arrangement, the short focal length lens arrangement being positioned at a proximal end of the needle probe.

[0027] The length of the needle probe may be substantially equal to the focal length of the long focal length lens arrangement.

[0028] The probe housing body may further comprise a band pass filter, the band pass filter may be positioned between the long focal length lens arrangement and the junction of the needle probe and the probe housing body.

[0029] The needle probe may be configured to direct polarized light to the tissue.

[0030] The system may further comprise: a beam splitter which may be configured to split the Raman scattering into two polarized components; and the spectrometer may be configured to simultaneously and separately image the two polarized components to produce two Raman spectra.

[0031] The light may be provided by a multi-mode laser.

[0032] The needle probe may have a diameter of less than 5mm, or may have a diameter of less than 3mm, or may have a diameter of less than 1mm.

[0033] A further aspect of the present disclosure provides a method for obtaining structural information relating to tissue, the method comprising one or more of: directing light to the tissue using a needle probe; focusing the light onto the tissue using a long focal length lens arrangement such that the light is reflected from the tissue producing Raman scattered light, wherein the long focal length lens arrangement is positioned within a probe housing body, distal from the junction of the needle probe and the probe housing body; collecting the Raman scattered light using the needle probe; imaging the Raman scattered light using a spectrometer to produce at least one Raman spectrum; and processing and analysing the at least one Raman spectra using a computer program to obtain structural information relating to the tissue. The focusing of the light onto the tissue may at least in part be dependent upon the long focal length lens arrangement having a focal length that may be substantially equal to the distance between the long focal length lens arrangement and the junction of the needle probe and the probe housing body.

[0034] The method may further comprise focusing the light from the long focal length lens arrangement onto the tissue via a band pass filter, the band pass filter may be positioned within the probe housing body between the long focal length lens arrangement and the needle probe.

[0035] The directing of the light to the tissue may be achieved by the needle probe and a short focal length lens arrangement, the short focal length lens arrangement may be positioned at a proximal end of the needle probe.

[0036] The method may further comprise directing light from a multi-mode laser to the tissue using a needle probe.

[0037] The method may further comprise directing polarized light to the tissue using needle probe.

[0038] The method may further comprise: splitting the Raman scattered light into two polarised components using a beam splitter; simultaneously and separately imaging the two polarised components using a spectrometer to produce two Raman spectra; and processing and analysing the two produced Raman spectra using a computer program to obtain structural information relating to the tissue.

[0039] Summary of Part 2

[0040] Part 2 of the present disclosure provides a method and system for determining the biochemical composition of tissue using Raman spectroscopy. Firstly, a number of phantom specimens having known biochemical compositions at obtained, and then Raman spectra of the respective phantom specimens are taken using a Raman spectroscope. The Raman spectra are then processed by a suitably programmed computer to obtain relative biochemical constituent concentrations, and then a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions is found using curve fitting techniques. Once the non-linear transformation curve has been found for known phantom tissue, it can then be applied to characterise unknown tissue samples, the biochemical composition of which is to be found. To do this a Raman spectrum of the unknown tissue is taken, which is then processed to obtain relative tissue constituent composition data indicating the relative biochemical tissue constituent proportions in the tissue sample. The previously found non-linear transformation curve is then applied to the relative tissue constituent composition data to determine quantitative absolute measurements of the biochemical tissue constituents.

[0041] In view of the above, from one aspect the present disclosure provides a method of determining the biochemical composition of tissue, comprising: obtaining at least one Raman spectrum of a tissue sample, the biochemical composition of which is to be found; processing the Raman spectrum to obtain relative tissue constituent composition data indicating the relative biochemical tissue constituent proportions in the tissue sample; and then applying a non-linear fitted curve model to the relative tissue constituent composition data to determine quantitative absolute measurements of the biochemical tissue constituents.

[0042] In one embodiment the relative tissue constituent composition data is obtained using a multivariate least-squares linear regression model, whereas the non-linear fitted curve model is obtained using a a 2D polynomial curve-fit of a plurality of selected Raman data.

[0043] In some embodiments the Raman data were selected by minimizing model error against known constituent concentrations in a plurality of phantoms used to obtain test data to calibrate the non-linear fitted curve model.

[0044] The multivariate least-squares linear regression model is used to obtain relative tissue constituent composition data. This model is advantageous because it can handle multiple independent variables simultaneously, allowing for the analysis of complex relationships between these variables and the dependent variable. This method is particularly useful when the relationship between the variables is linear, as it minimizes the sum of the squares of the differences between the observed and predicted values, providing a best-fit line for the data.

[0045] On the other hand, the non-linear fitted curve model, obtained using a 2D polynomial curve- fit, is used to calibrate the model against known constituent concentrations in a plurality of phantoms. This approach is beneficial when the relationship between the variables is nonlinear, as it can capture more complex patterns in the data. By minimizing model error against known concentrations, this method ensures that the model is accurately calibrated, leading to more precise and reliable predictions.

[0046] In summary, the multivariate least-squares linear regression model is advantageous for handling multiple linear relationships, while the 2D polynomial curve-fit is better suited for capturing non-linear relationships and ensuring accurate model calibration.

[0047] In some embodiments the calibration of the non-linear fitted curved model comprises: obtaining a plurality of Raman spectra relating respectively to a plurality of phantom specimens having known biochemical compositions; processing the Raman spectra to obtain relative biochemical constituent concentrations; and then finding a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions; wherein the found non-linear transformation curve is then used as the nonlinear fitted curve model.

[0048] In some embodiments the tissue is animal connective tissue, and the biochemical tissue constituents comprise at least GAG, collagen, and water. In such a case the phantom specimens include specimens having different known concentrations of GAG, collagen, and water.

[0049] Another aspect of the disclosure provides a method of calibrating a non-linear transformation model for the quantification of tissue subject to Raman spectroscopy, comprising: obtaining a plurality of phantom specimens having known biochemical compositions; obtaining a plurality of Raman spectra relating respectively to the plurality of phantom specimens; processing the Raman spectra to obtain relative biochemical constituent concentrations; and then finding a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions.

[0050] A yet further aspect of the disclosure provides a system, comprising : a Raman spectrometer; and a processor arranged to receive and process Raman spectrum data obtained by the Raman spectrometer; the system further comprising processor readable instructions that when executed by the processor cause the processor to: receive at least one Raman spectrum of a tissue sample, the biochemical composition of which is to be found; process the Raman spectrum to obtain relative tissue constituent composition data indicating the relative biochemical tissue constituent proportions in the tissue sample; and then apply a non-linear fitted curve model to the relative tissue constituent composition data to determine quantitative absolute measurements of the biochemical tissue constituents.

[0051] Moreover, another aspect of the present disclosure provides a system, comprising: a Raman spectrometer; and a processor arranged to receive and process Raman spectrum data obtained by the Raman spectrometer; the system further comprising processor readable instructions that when executed by the processor cause the processor to: obtain a plurality of Raman spectra relating respectively to a plurality of phantom specimens having known biochemical compositions; process the Raman spectra to obtain relative biochemical constituent concentrations; and then find a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions.

[0052] Further features and advantages will be apparent from the appended claims. Brief Description of the Drawings

[0053] Further features and advantages of the present disclosure will become apparent from the following description of an embodiment thereof, presented by way of example only, and with reference to the accompanying drawings, wherein like reference numerals refer to like parts, and wherein:

[0054] Figures for Part 1 :

[0055] Figure 1A shows a schematic of an optical design comprising a long working distance lens to couple light into the needle for polarized Raman spectroscopy, in accordance with an embodiment of the present disclosure;

[0056] Figure IB shows a schematic of an optical design comprising a long working distance lens to couple light into the needle for conventional (non-polarized) Raman spectroscopy;

[0057] Figure 1C shows a schematic of an optical design comprising a long working distance lens to couple light into the needle and a telescope for beam shrinking, for use in conventional (nonpolarized) Raman spectroscopy;

[0058] Figure ID shows a schematic of an optical design comprising a long working distance lens to couple light into the needle and a telescope for beam shrinking, for use in polarized Raman spectroscopy;

[0059] Figure 2A shows images of undertaking needle polarized Raman analysis of healthy and dysplastic lesion;

[0060] Figure 2B shows a graph of average P-polarized Raman spectra of dysplastic lesion and normal edge showing subtle but important spectral differences across the spectra;

[0061] Figures 2C and 2D show graphs of depolarisation analysis and probabilistic partial least squares discriminant analysis (PLS-DA) showing significant spectral (structural) differences are associated with dysplasia;

[0062] Figure 3A shows images of healthy and lesion afflicted femoral condyles disarticulated from skeletally mature bovines;

[0063] Figure 3B shows an image of a sectioned osteochondral block for site specific analysis; Figure 3C shows an example image of the Raman spectroscopy needle probe, in accordance with an embodiment of the present disclosure;

[0064] Figures 3D and 3E show representative 2D stacked area plots of cumulative contribution of GAG, Col and H20;

[0065] Figure 4A shows a graphical representation showing the levels of GAG content, collagen content, and water content for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks;

[0066] Figure 4B shows a graphical representation showing the Raman GAG score, Raman COL score, Raman H2O score, and Raman OH area for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks;

[0067] Figure 4C shows a graphical representation showing Raman GAG score against assay- measured GAG content for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of cartilage blocks;

[0068] Figure 4D shows a graphical representation showing the change in elastic modulus for different tissue specimens, namely, healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks;

[0069] Figure 4E shows graphical representation showing Raman GAG score against elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks;

[0070] Figure 4F shows graphical representation showing linear combination of Raman GAG score + Raman OH area against elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks;

[0071] Figure 5A shows representative anatomical site variability of Raman GAG score over femoral condyles human cadaver donor specimens.

[0072] Figure 5B shows a graphical representation showing GAG content vs Raman GAG score for all analyzed human tissue blocks;

[0073] Figure 5C shows a graphical representation showing Raman GAG score vs elastic modulus for all analyzed human tissue blocks; Figure 5D shows a graphical representation showing linear combination of Raman biomarkers vs elastic modulus for all analyzed human tissue blocks;

[0074] Figure 6A shows a graphical representation showing Outerbridge grade vs elastic modulus for all analyzed human tissue blocks;

[0075] Figure 6B shows a graphical representation showing GAG content vs Raman GAG score for subset of Outerbridge grade 0 human tissue blocks;

[0076] Figure 6C shows a graphical representation showing Raman GAG score vs elastic modulus for subset of Outerbridge grade 0 human tissue blocks;

[0077] Figure 6D shows a graphical representation showing linear combination of Raman biomarkers vs elastic modulus for subset of Outerbridge grade 0 human tissue blocks;

[0078] Figure 6E shows a graphical representation showing OARSI scores for representative healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks.

[0079] Figure 6F shows a graphical representation showing OARSI scores for representative human cartilage blocks.

[0080] Figure 6G shows a graphical representation showing OARSI score vs elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks.

[0081] Figure 6H shows a graphical representation showing OARSI score vs elastic modulus for human cartilage blocks.

[0082] Figure 61 shows a graphical representation showing linear combination of Raman biomarkers vs elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks.

[0083] Figure 6J shows a graphical representation showing linear combination of Raman biomarkers vs elastic modulus for human cartilage blocks.

[0084] Figure 7A shows graphical representation of T2* MRI relaxation time vs elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks. Figure 7B shows graphical representation of T2 MRI relaxation time vs elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks.

[0085] Figure 7C shows graphical representation of Raman GAG score vs elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks.

[0086] Figure 7D shows graphical representation of linear combination of Raman biomarkers vs elastic modulus for healthy condyle (HC), lesion sites (LS), and lesion peripheral sites (LPS) of bovine cartilage blocks.

[0087] Figure 7E shows graphical representation of T2* MRI relaxation time vs elastic modulus for all analyzed human tissue blocks.

[0088] Figure 7F shows graphical representation of T2 MRI relaxation time vs elastic modulus for all analyzed human tissue blocks.

[0089] Figure 7G shows graphical representation of Raman GAG score vs elastic modulus for all analyzed human tissue blocks.

[0090] Figure 7H shows graphical representation of linear combination of Raman biomarkers vs elastic modulus for all analyzed human tissue blocks.

[0091] Figure 71 shows graphical representation of T2* MRI relaxation time vs elastic modulus for early OA (Outerbridge grade < 2.0) human tissue blocks.

[0092] Figure 7J shows graphical representation of T2 MRI relaxation time vs elastic modulus for early OA (Outerbridge grade < 2.0) human tissue blocks.

[0093] Figure 7K shows graphical representation of Raman GAG score vs elastic modulus for early OA (Outerbridge grade < 2.0) human tissue blocks.

[0094] Figure 7L shows graphical representation of linear combination of Raman biomarkers vs elastic modulus for early OA (Outerbridge grade < 2.0) human tissue blocks.

[0095] Figure 8A shows an image of the Raman spectroscopy needle probe sensing patient tissue regeneration, in accordance with the present disclosure; Figures 8B, 8C and 8D show graphical representations of the resultant tissue generation in relation to the Young's modulus, the sGAG content and the Raman GAG score, respectively, for each human donor + / - TGF-beta administration;

[0096] Figure 8E shows a graphical representation of the Raman GAG score against the sGAG content for all tested constructs;

[0097] Figure 8F shows a graphical representation of the Raman GAG score vs the Young's modulus for all tested constructs;

[0098] Figure 9A shows a block diagram depiction of a Raman spectroscopy needle probe analysing cartilage using tissue phantoms;

[0099] Figure 9B shows an image of a compositional space of tissue phantom sets for varying GAG, collagen (COL), and water (H2O);

[0100] Figure 9C shows 2D stacked area plots depicting contribution of GAG, COL, H2O to representative tissue phantoms;

[0101] Figure 9D shows a bivariate polynomial fitting of derived Raman biomarkers to phantom composition.

[0102] Figure 9E shows a live tissue cartilage explant model for generation of cartilage with varied compositional states;

[0103] Figure 9F shows a graphical representation of a compositional space of live tissue explants after 14 days chemical stimulation in culture;

[0104] Figure 9G shows a regression of Raman phantom corrected predictions of GAG and H2O vs assay-measured composition;

[0105] Figure 10A shows a depiction of Ex vivo Raman biochemical mapping of GAG and H2O content (% wet weight) over surfaces of equine trochlea;

[0106] Figure 10B shows a graph of regression Raman biochemical analysis vs. DMMB measured GAG content;

[0107] Figure IOC shows images of a Raman spectroscopy probe in contact with trochlea surface; Figure 10D shows a graph of in vivo Raman biochemical mapping of GAG content over the surface of equine trochlea;

[0108] Figure 11A shows representative spectra and decomposition to ECM components with(A) low bone contribution and (B) high bone contribution. (C) Intraoperative Raman probe measuring tissue composition adjacent to the 2 circular chondral defects. (D) Map of in vivo Raman sample sites in neocartilage and surrounding uninjured native cartilage. Photo is arthroscopy from Horse 1 proximal at 6 months.

[0109] Figure 11B shows a representative Bonescore anatomic site map before defect (healthy) and 3, 6, and 12 months after defect and microfracture for horse 1 proximal.

[0110] Figure 11C shows the temporal evolution in Bonescore and GAGscore of neocartilage(blue) and native cartilage (orange) for each horse and defect over 12 months, with 12-month photo shown for each site. Dashed lines group data and images from the same defect.

[0111] Figures for Part 2:

[0112] Figure 12 shows, in order, (A) GAG content, (B) H2O content, and (C) EY of treated explants over time. The dashed line at day 15 is when IL-lo was added to groups previously cultured without it. (D) shows the composition of all explants in terms of GAG and estimated collagen content. Diagonal orange lines are isolines of equal H2O concentration.

[0113] Figure 13 shows representative stacked area plots of Raman fingerprint spectra for (A) + Dex / -ILlo and (B) -Dex / +ILlo groups. (C) shows representative high-wavenumber spectra from the same, and (D) shows a 3D contour plot of a Raman corrector function predicting GAG.

[0114] Figure 14 shows the correlation of Raman metrics to biochemically measured composition (see A-C) before and (see E-D) after phantom correction. (F) shows the correlation of mechanical properties to Raman estimates of GAG and H2O concentration.

[0115] Figure 15 is a block diagram of a Raman spectroscopic probe that can be used to obtain Raman spectra for use as input information into embodiments of this part of the present disclosure.

[0116] Figure 16 is diagram of the inputs and steps used in an embodiment of Part 2 of the present disclosure.

[0117] Figure 17 is a data flow diagram illustrating the operation of the embodiment of the second part of the present disclosure. Figure 18 is a series of photographs showing how the phantoms are produced in the embodiments of the second part of the disclosure.

[0118] Figure 19 is a false colour map visualizing the distribution of GAGs within the tissue from Raman spectral data.

[0119] Detailed Description of the Preferred Embodiments

[0120] Description of embodiments for part 1

[0121] In part 1, the present disclosure provides a method and system for use in Raman spectroscopy, the method and system utilizing a long working distance lens and a thin needle probe, wherein the long distance working lens is arranged to focus incident light into the back aperture of the thin needle probe.

[0122] In more detail, the present disclosure relates to a system and method for obtaining structural information relating to tissue. More particularly, the present disclosure provides an optical system which utilizes a long focal length lens arrangement to focus polarized light incident on a needle probe. Focusing the polarized light across a distance, before the polarized light reaches the needle probe ensures that the maximum energy is retained within the polarized light without introducing background signal from additional lenses or telescopes. In turn, this enables the miniaturization of the needle probe, therefore, making the system applicable for use in vivo examinations.

[0123] An embodiment of the present disclosure provides a system for obtaining structural information relating to tissue, the system comprising: a needle probe projecting from a probe housing body, a probe housing body comprising a long focal length lens arrangement, the long focal length lens arrangement positioned within the probe housing body, distal from the junction of the needle probe and the probe housing body; the needle probe being configured to direct light to the tissue and collect reflected Raman scattered light; and a spectrometer configured to image the reflected Raman scattered light to produce at least one Raman spectrum.

[0124] A further embodiment of the present disclosure provides a method for obtaining structural information relating to tissue, the method comprising: directing light to the tissue using a needle probe; focusing the light onto the tissue using a long focal length lens arrangement such that the light is reflected from the tissue producing Raman scattered light, wherein the long focal length lens arrangement is positioned within a probe housing body, distal from the junction of the needle probe and the probe housing body; collecting the Raman scattered light using the needle probe; imaging the Raman scattered light using a spectrometer to produce at least one Raman spectrum; and processing and analysing the at least one Raman spectra using a computer program to obtain structural information relating to the tissue.

[0125] More particularly, a further embodiment of the present disclosure provides a system for obtaining structural information relating to tissue, the system comprising: a needle probe projecting from a probe housing body, a probe housing body comprising a long focal length lens arrangement, the long focal length lens arrangement positioned within the probe housing body, distal from the junction of the needle probe and the probe housing body; the needle probe being configured to direct polarized light to the tissue and collect reflected Raman scattered light; a beam splitter configured to split the Raman scattering into two polarised components; and a spectrometer configured to simultaneously and separately image the two polarised components to produce two Raman spectra.

[0126] A further embodiment of the present disclosure provides a method for obtaining structural information relating to tissue, the method comprising: directing polarized light to the tissue using a needle probe; focusing the polarized light onto the tissue using a long focal length lens arrangement such that the polarized light is reflected from the tissue producing Raman scattering, wherein the long focal length lens arrangement is positioned within a probe housing body, distal from the junction of the needle probe and the probe housing body; collecting the Raman scattering using the needle probe; splitting the Raman scattering into two polarised components using a beam splitter; simultaneously and separately imaging the two polarised components using a spectrometer to produce two Raman spectra; and processing and analysing the two produced Raman spectra using a computer program to obtain structural information relating to the tissue.

[0127] The present disclosure will now be discussed in more detail in relation to the associated Figures.

[0128] In Figures 1A-1D, the following key applies - M: mirror, PB: polarization beam splitter, LP: long pass filter, DM: dichroic mirror, BP: band pass filter, P: Polarizer.

[0129] The present disclosure provides and system and method which places a single long working distance lens and focuses the light into the needle back aperture (Figures 1A-B). This efficiently allows 40% of the light to be transmitted. This is not intuitive for the following reasons: (1) the laser light will be incident on the bandpass filter with a small angle. (2) the laser light will not be fully collimated. Conventional wisdom would place a beam shrinker (telescope) to guide the laser light into the needle (Figures 1C-D). The telescope approach is highly inefficient as multimode laser beams leads to imperfect beam shrinking. This comes with a massive disadvantage since the beam shrinker creates background signal by being in the Raman optical path. Further, guiding laser light into a needle with a beam shrinker cannot be done efficiently using a multimode laser beam. The present disclosure enables miniaturization of the probe (without background signal) and the use of thin needles compatible with in vivo examinations. The present disclosure enables the diameter of the needle probe to be reduced to less than 0.5mm. However, in some embodiments of the present disclosure the diameter of the needle probe may be less than 5mm, or less than 3mm, or less than 1mm.

[0130] It is worth noting that the present disclosure is suitable for use with both polarized and nonpolarized light, these may be considered as separate embodiments of the present disclosure as shown in Figure 1A and IB, respectively.

[0131] Supporting results (Oral cancer.)

[0132] The following description details the application of the present disclosure. Particularly, the following description details the application of the thin Raman spectroscopy need probe for cancer diagnosis in the oral cavity.

[0133] Polarized Raman needle probe predicts cancer in the oral cavity. Head & neck cancers including oral cancers are severe life-limiting diseases, particularly when diagnosed in the later stages. Earliest possible detection in precancer (dysplasia) or early cancer stages is the single most important measure for reducing oral cancer patients' morbidity and mortality rates. Conventional diagnosis relies on visual inspection guiding the biopsy of suspicious oral lesions, which has resulted in 5 year survivals of up to 90% for early tumours or dysplasia. This survival rate however declines to at best 50% for advanced cancers, highlighting the necessity of early diagnosis as a primary factor in determining patient outcomes. Visual guided tissue sampling suffers from inter-observer dependence, sampling errors and difficulty in identifying and targeting flat dysplastic lesions in particular. The differentiation between inflamed / scared tissue and dysplasia, establishing cancer depth, local / distant spread assessment (staging) and grading (aggression), and margin assessment of malignant lesions all represent monumental clinical challenges. Further, biopsy also remains a significant deterrent to patient attendance. There is a significant unmet clinical need for a rapid, non-invasive, outpatient-deployable diagnostic method to improve surveillance, early diagnosis, intra-operative margin assessment and follow up I management of oral cancer patients.

[0134] Optical imaging and spectroscopy have recently offered great promise to address unmet clinical needs since they are non-invasive and can capture molecular / structural information without prior tissue preparation. These technologies offer new, clinically relevant biomarkers such that essential point-of-care decisions can be made with the use of safe (non-ionizing) levels of optical radiation at a much lower cost than with MRI, CT and PET. Raman spectroscopy is a label-free, rapid and minimally invasive optical technique using laser light that provides a point-wise optical fingerprint of the myriad of inter- and intra-cellular building blocks (i.e., proteins, lipids and DNA) of tissue at the biomolecular level. Raman spectroscopy offers label-free diagnosis of cancers in vivo. In the last two decades, there has been accumulating evidence on the accurate diagnostic capability of Raman spectroscopy through comprehensive in vitro, ex vivo and in vivo studies. The key advantage of polarized Raman spectroscopy compared to conventional Raman spectroscopy is that it offers additional specific information about molecular organisation / symmetries and tissue structure.

[0135] Polarized Raman spectroscopy was conducted on oral dysplasia and the distal edge of the lesion (normal tissues) using the miniaturized needle probe, in accordance with the present disclosure (Figure 2A-B). The analysis involved Partial Least Squares Discriminant Analysis (PLS-DA) of the depolarization ratio (DPR) in oral cancer and benign tissue obtained from the same patient ex vivo (Figure 2C). The probabilistic results (Figure 2D) indicate that diagnostic information crucial for distinguishing between cancer and benign tissues can be derived solely from molecular symmetries / ordering and tissue structure.

[0136] Supporting results (Cartilage)

[0137] The following description showcases the different applications of the present disclosure. Particularly, the following description showcases the different applications of the thin Raman spectroscopy needle probe when used for analysis of cartilage health.

[0138] Raman needle probe predicts bovine cartilage properties. The capability of the miniaturized needle Raman probe was assessed to predict degeneration of cartilage properties. This analysis was initially performed on degenerated skeletally mature bovine femoral condyle specimens with naturally occurring focal chondral lesions.

[0139] The femoral articular surface of 12 disarticulated, skeletally mature, bovine knees were examined macroscopically and divided into a healthy femoral condyle [HC] group (n=6) and a lesion-afflicted condyle [LAC] group (n=6) possessing a ~01O-15mm focal chondral lesion within the load bearing region of one of the femoral condyles, but no apparent macroscopic evidence of degeneration in the peripheral cartilage (Figure 3A). For tissue analysis, one condyle from each knee was sectioned along a rectangular grid using a fine-tooth handsaw to produce ~6x6x6mm osteochondral tissue blocks (n=10-15 / condyle). For LAC specimens, analyses were performed on specimens excised from the chondral lesion site (LS) and surrounding cartilage (i.e., lesion peripheral sites (LPS); Figure 3A). The material properties and composition of 128 osteochondral specimens excised from 12 condyles were examined by Raman probe spectroscopy, mechanical indentation, biochemical assay, and Outerbridge visual grading. Raman derived ECM biomarkers were regressed to cartilage biochemical content and tissue modulus. Histopathologic OARSI grading was compared to tissue modulus for a random subset of 30 specimens. An additional five intact bovine condyles were evaluated by MRI, then sectioned into 54 osteochondral block specimens and subjected to Raman probe spectroscopy and mechanical testing. Predictions of cartilage composition and tissue modulus achieved by Raman probe spectroscopy were compared to predictions from quantitative MRI.

[0140] Raman spectra were acquired with the Raman needle probe (Figure 3C), in accordance with the present disclosure, and the fingerprint range (800-1800cm-1) was preprocessed and subjected to a multivariate least-squares linear regression model to calculate the relative contribution of the predominant cartilage constituents (GAG, COL, and H2O) to the composite cartilage Raman spectra (Eqn 1).

[0141] Cartilage spectra=GAGscore* (GAGREF) + COLscore*(COLREF) + H2Oscore*( H2OREF) (1) where the derived regression coefficients (GAGscore, COLscore, H2Oscore) represent the relative contribution of the predominant constituents of cartilage ECM (GAG, COL, H2O) to the composite cartilage Raman spectra, calculated using the spectra of purified reference chemicals for each ECM constituent: GAGREF=chondroitin sulfate (shark cartilage; Sigma[C4384]); COI_REF=type-II collagen (chicken sternal cartilage; Sigma[C9301]); H2OREF= PBS. For each constituent, the Raman score ranges from a value of zero to one based on the constituent's weighted contribution to the tissue spectra. The high wavenumber range (2700-3800cm-1) was preprocessed using a linear baseline subtraction, area-under- curve normalization, followed by measurement of the area under the organic-content- associated CH2 region (CHarea) and water-content-associated OH region (OHarea) .

[0142] The elastic modulus of the chondral layer for each tissue block was derived by indentation testing and the central 03mm chondral core of each block was sub-punched and analyzed for GAG content. OARSI scores (gradeXstage) were assessed by a board-certified pathologist on Saf O sections for a random subset of tissues from all analyzed regions (N=30). MRI examination was performed using a 9.4T Bruker BioSpec MRI scanner to obtain T2* relaxation times for an additional 2 healthy and 2 lesion-afflicted condyles, followed by the aforementioned sectioning, Raman analysis, and mechanical testing protocol.

[0143] Compared to HC, both the LS and the adjacent LPS cartilage were degraded compositionally (Figure 4): GAG (p<0.001) and collagen contents (p<0.001) were reduced, while H2O content increased (p<0.002), indicative of increased porosity (Figure 4A). Multivariate regression models fit to the Raman spectra, where the regression coefficients (GAGscore, COLscore, H2Oscore) represented the relative contribution of each ECM constituent to the composite cartilage Raman spectra, accounted for 91±1.5% of the variation of the fingerprint region. Raman spectroscopy identified similar changes in cartilage GAG composition: GAGscore (Figure 4B) was decreased for both LS (p<0.001) and LPS (p<0.001) cartilage compared to HC, and the GAGscore described 78% of the variation of GAG content for all specimens (R2=0.78; p<0.001; Figure 40). These alterations in ECM composition resulted in significant softening of both LS (p<0.001) and LPS (p<0.001) tissue compared to HC tissue (Figure 4D). Changes in the Raman derived ECM biomarkers accounted for changes in the material properties of the degraded cartilage (Figures 4E,F): GAGscore accounted for 71% of the variation in elastic modulus (R2=0.71; p<0.001), as did a linear combination of GAGscore and OHarea describing 72% of the variation in modulus (R2= 0.72; p<0.001).

[0144] Raman needle probe predicts mechanical properties of human cartilage ex vivo.

[0145] A follow-up study was performed on ex vivo specimens acquired from human cadavers, representing a large range of degenerative states. The medial and lateral femoral condyles of 10 disarticulated human cadaver knees procured from the National Disease Research Interchange, ranging in age from 27 to 75 years, were examined macroscopically and cartilage morphology classified according to Outerbridge visual grading scheme prior to sectioning into ~6x6x6mm osteochondral specimens. Raman spectroscopy derived ECM biomarkers were regressed to GAG content (n=125) and tissue modulus (n=235 blocks).

[0146] There was wide variation in tissue composition and stiffness for hyaline cartilage comprising the human femoral condyles (Figure 5) that depended on donor age and anatomic location (i.e., medial vs. lateral condyle). Raman needle probe spectroscopy revealed higher GAGscore for stiffer tissues (Figure 5A). GAGscore accounted for 60% of the variation in GAG content (Figure 5B; R2=0.60; p<0.001) and 60% of the variation in elastic modulus (Figure 5C; R2=0.60; p<0.001). A linear combination of the Raman derived ECM biomarkers GAGscore and OHarea accounted for 71% of the variation in elastic modulus (Figure 5D; R2=0.71, p<0.001).

[0147] Raman vs macro- and micro-scopic assessment of cartilage pathoanatomy.

[0148] An additional analysis was performed to test the hypothesis that Raman-probe derived ECM biomarkers predict cartilage material properties better than arthroscopic visual grading schemes (Outerbridge) and histopathologic grading (OARSI), For aforementioned bovine specimens, histopathologic OARSI grading was compared to tissue modulus for a random subset of 30 specimens. For aforementioned human specimens, OARSI scores were compared to cartilage mechanical properties for a random subset of 71 specimens. Outerbridge classification of chondromalacia was based on direct visualization of the articular cartilage augmented by magnified macrophotography images. Using a superimposed grid corresponding to subsequent sectioned osteochondral tissue blocks, a grade of 0 through IV was assigned to the chondral area of interest: Grade 0 = normal cartilage; Grade I = no break in surface integrity but softened tissue when pushed with a probe; Grade II = partialthickness defect with fissures <0.5 inches in diameter that do not reach subchondral bone; Grade III = cartilage fissuring with an area >0.5 inches that reach subchondral bone; Grade IV - cartilage erosion that exposes subchondral bone. For OARSI score assessments, 03mm cylindrical cartilage cores were formalin fixed and paraffin embedded, and sections were stained with safranin-O / fast-green. OARSI scores were determined as the product of grade (0 to 6) and stage (0 to 4) by a board-certified pathologist.

[0149] Outerbridge grade accounted for only 5% of the variation in elastic modulus of human specimens (Figure 6A; R2=0.045; p<0.001). Emphasizing the relative insensitivity of macroscopic assessments of cartilage health, for the subset of Outerbridge 0 human specimens, Raman ECM biomarkers accounted for 70% of the variation in GAG content (Figure 6B; R2= 0.70, p<0.001) and 74% of the variation in modulus (Figure 6C; R2= 0.74, p<0.001); a linear combination of GAGscore and OHarea accounted for similar variation in elastic modulus (R2=0.75; p<0.001; Figure 6D). For bovine specimens, OARSI scores varied according to the injury group: 3.0±3.4 for HC, 5.7±2.0 for LPS, and 7.4±3.6 for LS (Figure 6E). For human specimens, OARSI scores ranged from 0 to 18 (Figure 6F). OARSI score accounted for 54% of the variation in elastic modulus of bovine tissue (Figure 6G; R2=0.54; p<0.001) but only accounted for 12% of the variation in elastic modulus of human tissue (Figure 6H; R2=0.12; p<0.001). In comparison, a linear combination of Raman derived ECM biomarkers accounted for 86% of the variation in modulus for bovine tissue (Figure 61; R2=0.86; p<0.001) and 70% of the variation in modulus for human tissue (Figure 6J;

[0150] R2=0.70; p<0.001).

[0151] Raman vs MRI assessments of cartilage properties

[0152] An additional analysis was performed to test the hypothesis that Raman-probe derived ECM biomarkers predict cartilage material properties better than quantitative compositional MRI (T2*and T2 mapping). An additional five intact bovine condyles were evaluated by MRI, then sectioned into 54 osteochondral block specimens and subjected to Raman probe spectroscopy and mechanical testing. Predictions of cartilage composition and tissue modulus achieved by Raman probe spectroscopy were compared to predictions from quantitative MRI. The medial / lateral femoral condyles from one human cadaver donor and tibial plateaus from five human cadaver donors were subjected to MRI then sectioned into 88 osteochondral block specimens and subjected to Raman spectroscopy and mechanical testing. Raman versus MRI comparisons were further performed for a subset of tissue blocks with early OA pathoanatomy as defined by originating from a chondral surface with a mean Outerbridge grade <2.0. T2* and T2 relaxation times were obtained on the intact bovine and human specimens using a 9.4T Bruker BioSpec MRI scanner, mapped for the full cartilage thickness and co-registered to the position of each excised osteochondral specimens. MRI relaxation times predicted cartilage tissue properties moderately. For bovine specimens, T2* & T2 relaxation times (Figure 7A-B) described 27% and 30% of the variation in elastic modulus (p<0.001), respectively. In comparison, the Raman derived ECM biomarkers better predicted cartilage tissue properties: GAGscore accounted for 70% of the variation in elastic modulus (R2=0.70; p<0.001) for bovine specimens (Figure 7C); a linear combination of GAGscore and OHarea accounted for similar variation in elastic modulus (R2=0.72; p<0.001; Figure 7D). For human specimens, T2* & T2 relaxation times (Figure 7E-F) described 15% and 25% of the variation in elastic modulus respectively (p<0.001). For human specimens, Raman GAGscore accounted for 67% of the variation in elastic modulus (R2=0.67; p<0.001) (Figure 7G); a linear combination of GAGscore and OHarea accounted for similar variation in elastic modulus (R2=0.71; p<0.001; Figure 7H). For early-OA-associated human specimens, T2* & T2 described 15% and 45% of the variation in modulus (p<0.001; Figure 7I-J). For these specimens, GAGscore accounted for 82% of the variation in elastic modulus (R2=0.82; p<0.001; Figure 7K); a linear combination of GAGscore and OHarea accounted for similar variation in elastic modulus (R2=0.84; p<0.001; Figure 7L).

[0153] Raman needle probe for monitoring of patient specific cartilage regeneration profiles. The ability of the Raman needle probe to monitor the response of cartilage to emerging chondroregenerative treatments was examined. To this end, the ability of the Raman probe to monitor the development of regenerating neocartilage using an established cartilage tissue engineering protocol with varying clinical human cell populations was examined. Human chondrocytes were isolated from the femoral condyles of 3 human donors (HD) (NDRI; HD1: 58 / f, HD2: 36 / m, HD3: 53 / m) and expanded. For each HD, chondrocytes were seeded at 45X106cells / mL in 2% w / v type VII agarose to generate 03x2mm cylindrical neocartilage constructs. Constructs were cultured in chondrogenic medium supplemented with or without continuous supplementation of lOng / mL TGF-beta3 and subjected to Raman (Figure 8A), mechanical, and biochemical analysis at days 28 and 56 (n=15 constructs per timepoint per donor). Multivariate linear regression of the Raman fingerprint range was subjected to the aforementioned multivariate linear regression model (Eqn 1) with the incorporation of an additional agarose scaffold term (Agarose score), yielding the regression coefficient Raman scores for each ECM / scaffold constituent. Constructs were analyzed for compressive Young's modulus (EY), sGAG content (DMMB assay), and water content. TGF- beta exposure enhanced sGAG content (16-fold to 52-fold) and EY (3.6-fold to 8.3-fold) for all donors. At day 56, donors exhibited variable levels of sGAG content and EY (Figure 8B8<C): HD1 exhibited sGAG and EY levels that were respectively 2.9-fold and 2.8-fold higher than HD2, (p<0.001), and respectively both 1.7-fold higher than HD3, (p<0.005; Figure 8B&C). Raman sGAG and COL scores generally increased with culture duration, while H2O and Agarose scores decreased (not shown), indicating replace of the agarose scaffold by neocartilage ECM. At day 56, Raman sGAG scores reflected differences in sGAG and EY between donors: HD1 exhibited a Raman sGAG score that was 2.1-fold and 1.6-fold higher than HD2 and HD3, respectively (p<0.001; Figure 8D). For all donors, Raman sGAG scores predicted 87% and 88% of the variation in construct sGAG content and EY, respectively (Figure 8E8iF). This demonstrates the ability of the Raman needle probe to optically biopsy evolving neocartilage and achieve patient-specific quantitative assessments of its changing composition. The Raman sGAG score statistically distinguished the variable tissue composition profiles of each donor.

[0154] Quantitative Raman measurement of cartilage composition via tissue phantom calibration. The Raman platform, in accordance with the present disclosure, outputs biomarkers that are defined in spectral intensity units, and as such, provide only semi- quantitative assessments of tissue composition. The capability of using compositional tissue phantoms to convert relative Raman biomarkers from cartilage tissue specimens to absolute biochemical concentration values was explored. Tissue phantoms are composed of hydrogels with prescribed ratios of key cartilage constituents (GAG, COL, H2O) (Figure 9A&B) and used to generate a high quality training set that converts non-linear, relative Raman biomarker measures (Figure 9C) into an absolute biochemical concentration of each constituent (Figure 9D). This conversion algorithm was validated with an in vitro model system, whereby live cartilage explants were monitored in response to a range of biochemical compositional changes akin to those present in OA, consisting of 1) GAG depletion induced by media supplementation of the catabolic cytokine interleukin-la (IL-la), and 2) cartilage swelling induced by culture in the absence of corticosteroid dexamethasone (DEX) (Figure 9E). This culture protocol generated cartilage tissue specimens that span a wide range of compositional states (GAG: 0-10%, H2O: 70-90%), serving as a robust model to test the phantom correction algorithm (Figure 9F). Results demonstrated that phantom-based Raman biochemical predictions achieved outstanding quantitative agreement with direct assaybased biochemical measures (Figure 9G). This platform can serve as a standardized analytical tool to nondestuctively quantitatively measure the composition of cartilage in response to degeneration and treatment across a range of preclinical and clinical R+D model systems, spanning from ex vivo explant studies, to in vivo animal studies, the ROT clinical trials.

[0155] Raman arthroscopy for in vivo quantitative monitoring of cartilage composition in equine stifle joint. A study was performed where real-time, in vivo Raman spectroscopic assessments were performed in the operating room to achieve quantitative measures of the composition of cartilage in an equine stifle joint. Raman biomarkers were converted to absolute quantitative measures with the previously developed phantom data sets. Initially, for validation of the approach, analysis was performed on an excised equine trochlea ex vivo in order to compare Raman predictions to assay-based compositional measures (Figure 10A). Here, 3mm full thickness cartilage plugs were harvested at 169 discrete anatomical sites over the surface of an excised 5-year-old equine femoral trochlear groove. The composition of each plug was determined by Raman spectral analysis and DMMB GAG assay. Raman derived biomarkers accounted for 78% of the variance in GAG content (Figure IOC) and depicted a biochemical spatial gradient along the length of the trochlear groove, with GAG content increasing and H2O content decreasing from distal to proximal (Figure 10A-B). Subsequently, the Raman needle probe, in accordance with the present disclosure, was implemented in vivo to measure cartilage composition at discrete anatomical sites along the equine trochlear groove. After IACUC approval Raman needle probe spectral analysis was performed on the vertical wall of the lateral trochlear groove of two 5-year-old TB horses through a 5-6 cm mini-arthrotomy between and parallel to the middle and lateral patellar ligaments before and after creating two 15mm focal defects removing uncalcified and calcified cartilage down to the subchondral bone (Figure IOC). The needle tip was autoclaved and the optics draped in sterile plastic film. Measures were performed at defined anatomical sites using a PDMS template with 4mm diameter holes at sites along a defined polar coordinate system overlaying the trochlear facet. At each site, the probe lens gently contacted the cartilage surface, with spectra acquired over a 5 sec integration time. In the surgical setting, the needle probe achieved high quality spectral acquisition of cartilage, as evidenced by a high SNR and the strong fit of the multivariate regression model to the cartilage spectra (R2=0.93+ / -0.03). In-vivo Raman GAG measures exhibited a similar range and spatial distribution as the ex vivo analysis (Figure 10D). This study establishes the capability of the Raman arthroscopic platform, in accordance with the present disclosure, to achieve accurate quantitative in vivo measures of the biochemical content of cartilage in the surgical setting. The analysis is fully automated, which provides the clinician real time measurement of tissue composition, allowing diagnostic assessments at defined sites. The accuracy of ECM composition calculated from the in-vivo Raman spectral analysis is supported by ex vivo measures of tissue composition at discrete anatomical sites over the equine trochlea, where Raman derived GAG measures accounted for 78% of the variation in GAG content. The successful implementation of this platform in-vivo supports the use of Raman arthroscopy to objectively monitor standardized biomarkers that reflect cartilage function in health and disease and to (non-destructively) longitudinally monitor the efficacy of OA treatments that preserve or regenerate the cartilage ECM in pre-clinical animal models and human clinical trials. Further Raman arthroscopy assessments of the regenerate tissue generated by micro-fracture at the chondral defects at 38<. 6-month follow-up on these horses, providing the first-ever longitudinal in vivo Raman assessment of tissue response to a chondroregenerative therapy.

[0156] In-Vivo Raman Spectroscopy Can Assess Microfracture Neocartilage Formation and Maturation Longitudinally by Serial Optical "Biopsy" Measurements in the Equine Stifle Joint. A further equine study was performed to assess the ability of the sensing platform to monitor cartilage regeneration. Hyaline cartilage is a viscoelastic, biphasic, composite material optimized for its mechanical performance. The tissue is comprised of a type-II collagen (COL) fibril network that provides structure and tensile strength, constraining a negatively charged sulfated glycosaminoglycan (GAG) matrix that retains interstitial water, imparting fluid load support in compression. These extracellular matrix (ECM) components act synergistically, bestowing the rheological and tribological material properties essential to cartilage function. Osteoarthritis (OA) is an incapacitating condition in which hyaline cartilage progressively breaks down because of mechanical overloading, provoking a cascade of structural failure (fissuring / fibrillation / superficial zone delamination), tissue swelling, GAG depletion and COL network derangement. Development of chondro- protective / -regenerative therapies is impeded by a lack of non-destructive clinical assessments of cartilage composition germane to the tissue's mechanical function. Quantitative MRI (Tip, T2*) and arthroscopic-based cartilage grading (Outerbridge) are only moderately correlated with metrics characterizing cartilage composition and material properties1. Raman spectroscopy is an inelastic light scattering technique that reflects the vibrational modes of the biochemical building blocks (amides, sulfates, hydroxyls) of key cartilage ECM-specific biomarkers (GAG, COL, H2O) that contribute to the material properties intrinsic to tissue function. We developed a novel Raman spectroscopy needle probe and real-time spectral analysis platform capable of performing both ex-vivo and in-vivo measurement of ECM-specific compositional biomarkers for cartilage with a high degree of accuracy (R2=0.8-0.94). Additionally, the acquired Raman spectra can detect signal from the subchondral bone, reflecting chondral thickness. The aim of this study was to demonstrate that sequential, in-vivo, real-time Raman spectroscopic analysis obtained using a custom needle probe could assess the formation and maturation of neocartilage formed in an osteochondral defect created on the trochlear facet of the equine stifle joint treated by microfracture as well as monitor changes in tissue composition in the native cartilage surrounding the defect.

[0157] Repair Model'. With IACUC approval, two (proximal and distal) 15mm diameter chondral defects were created by removing all the cartilage to the calcified cartilage on the lateral trochlear facet of two 5-year-old thoroughbred horses, (Figure 11A (C)) which was then repaired by microfracture.

[0158] Raman Probe'. A custom Raman probe (In Photonics) comprised of a threaded needle tip (02.75mm) with a distal 02mm sapphire ball lens was fiber-coupled to a battery powered 785nm laser (lOOmW output; IPS) and spectrometer (Eagle; Ibsen), allowing portability.

[0159] Raman Spectra-. High SNR spectra in the fingerprint (800-1800cm-1) and high wavenumber (800-3800cm-1) ranges were acquired. Spectra were preprocessed by background subtraction and area-under-curve normalization. The cartilage fingerprint spectra were fit to a multivariate linear regression model:

[0160] Carti la g espectra = GAGscore* (GAGREF) + COLscore* (CO l_REF) + H2Oscore* ( H2OREF) + B0n6score* ( B0 neREF) where: GAGREF, COLREF, H20REF, and BoneREF are reference spectra of purified reference chemicals of each ECM constituent; "scores" are the "fit" regression coefficients reflecting the relative contribution of each ECM constituent and subchondral bone to the acquired composite spectra (Figure 11A(A,B)). Raman spectral analysis was performed in real time, with ECM biomarker scores and quality metrics displayed to the operator via a custom GUI.

[0161] In-Vivo Measures’. The needle probe was autoclaved, and optics draped in sterile plastic. Raman spectra were collected by inserting the needle probe through a parapatellar arthrotomy between the middle and lateral patellar ligaments (Figure 11A(C)). Holding the probe normal to the surface with minimal contact force, spectra were acquired for 10 seconds at multiple sites along the lateral trochlear facet within each defect and the surrounding native cartilage, referenced to a polar coordinate system defined by the margins of the defect and the proximodistal trochlear axis (Figure 11A(D)). Sequential in-vivo assessments at each site were performed before creating the defect, and @ 3, 6, and 12 months after microfracture (Figure 11B).

[0162] High quality spectral acquisition of the cartilage and neocartilage ECM biomarkers and subchondral bone signal was achieved in vivo in real time. The most obvious Raman spectra changes were elevated bone signals in the neocartilage with little change in the surrounding native cartilage over time, reflecting that the repair tissue was thinner (Figure 11C). At 3 months the Bonescore was significantly elevated for the neocartilage forming within 3 of 4 repair sites (horse 1 proximal defect; horse 2 proximal & distal defects) compared to the corresponding native cartilage pre-defect (p<0.001) and the native cartilage surrounding the defect (p<0.001). For both horses the Bonescore at the proximal defect decreased over time such that @ 12 months it was not different from the surrounding native cartilage, indicating progressive neocartilage formation. However, for horse 2 the Bonescore at the distal defect remained higher than the surrounding native cartilage (p<0.001) in stark contrast to horse 1 where the neocartilage Bonescore for the distal defect was not different from that of the surrounding native tissue at all timepoints, suggesting accelerated repair (Figure 11C). At both the microfracture site and surrounding native cartilage the GAGscore increased at 3 months for horse 1 at the proximal defect and for horse 2 at both defects, signifying an overall anabolic response stimulated by microfracture. While the GAGscore subsequently decreased @ 6 and 12 months, compared to baseline it remained elevated in both the native and neocartilage. This study establishes the capability of our Raman spectroscopic probe platform to perform in vivo real time assessments of the composition of emerging neocartilage repair tissue allowing for serial monitoring of cartilage regeneration. The Bonescore reflects the thickness of the emerging repair tissue, which initially was thinner than the original pre-defect native cartilage, but as the neocartilage evolved over time the Bonescore decreased and the GAGscore increased signifying the evolution of neocartilage over time and space. Sequential Bonescore and GAGscore measurements were not uniform across horses or defect sites highlighting the ability of the Raman needle probe spectroscopy platform to examine differential repair responses for individual patients.

[0163] Raman spectra acquired through our needle probe platform provided an "optical biopsy" of native and neocartilage tissue composition, non-destructively and in real time. The in-vivo longitudinal assessment of neocartilage formation after microfracture supports the clinical use of our needle probe Raman spectroscopy platform to examine the composition of cartilage in health and disease and to monitor the efficacy of chondroprotective / regenerative therapies in-vivo via an arthrotomy or arthroscopically.

[0164] Description of embodiments for part 2

[0165] Quantitative Raman histology

[0166] In part 2 we demonstrate a novel phantom-based approach that permits for the quantitative Raman histology of connective tissues. We demonstrate the capability of using compositional tissue phantoms to convert acquired relative Raman biomarkers from cartilage tissue specimens to absolute biochemical concentration values. This conversion algorithm is created using a phantom approach on known concentrations and validated with an in vitro model system, whereby live cartilage explants are monitored in response to a range of biochemical compositional changes, consisting of— 1) GAG depletion induced by media supplementation of the catabolic cytokine interleukin-lo (IL-lo)[3], and 2) cartilage swelling induced by culture in the absence of the corticosteroid dexamethasone (DEX) [4]. We demonstrate the accurate phantom-calibrated quantitative Raman measures are superior to standard Raman based analysis.

[0167] In summary, the phantom-based approach in quantitative Raman histology described herein works by using compositional tissue phantoms to convert relative Raman biomarkers from tissue specimens into absolute biochemical concentration values. The process involves creating tissue phantoms with known concentrations of glycosaminoglycans (GAG), collagen (COL), and water (H2O) to serve as reference samples. The Raman spectra of these reference phantoms are then measured, with the obtained Raman spectra providing an indication of the molecular composition of the phantoms. The obtained Raman spectra are then processed by being normalized and subjected to a multivariate least-squares linear regression model to calculate the relative contribution (or scores) of the predominant cartilage constituents (GAG, COL, and H2O). The relative scores are then converted to absolute metrics using a non-linear curve fitting approach. This involves creating a mapping function that converts the relative Raman scores to highly accurate measures of the concentration of biochemical constituents in the tissue.

[0168] The accuracy of the phantom-calibrated quantitative Raman measures is validated using an in vitro model system. Live cartilage explants are monitored in response to biochemical compositional changes, and the phantom-based approach is used to predict the concentrations of GAG, COL, and H2O in these explants. This method allows for accurate, non-destructive, and quantitative analysis of the biochemical composition of connective tissues, which is superior to standard Raman-based analysis.

[0169] With respect to the non-linear curve fitting, this is an important step to convert relative Raman scores into absolute biochemical concentrations. The phantom samples with known concentrations of glycosaminoglycans (GAG), collagen (COL), and water (H2O) have thein Raman spectra measured, and the relative Raman scores obtained from the tissue samples are then converted to absolute metrics using the non-linear curve fitting approach. This involves using a 2D polynomial curve-fit of two Raman metrics. In some embodiments of the disclosure, the Matlab curve fitting toolbox (sfit) is used for this purpose. During the curve fitting some model optimization is performed by selecting relative Raman scores that minimize the model error against the known constituent concentrations in the phantoms. This ensures that the transformation is accurate and reliable.

[0170] After the non-linear transformation (i.e. the curve fitting), the accuracy of the predicted concentrations is validated against assay-measured GAG and H2O contents in cultured explants. The results show a high degree of accuracy, with a correlation coefficient (R2) of 0.89 for water and 0.82 for GAG1. The non-linear transformation therefore allows for accurate quantitative analysis of the biochemical composition of connective tissues, making the phantom-based approach superior to standard Raman-based analysis.

[0171] Once the curve fitting has been performed using the phantoms, it is then possible to use the system to analyse real-world tissue samples, the biochemical composition of which it is desired to find. This is performed by obtaining Raman spectra from across the tissue sample using a Raman microscope. This involves scanning the tissue sections to acquire Raman images, which provide a detailed optical fingerprint of the molecular composition of the tissue. The acquired Raman spectra are then normalized and subjected to a multivariate leastsquares linear regression model to calculate the relative contribution (or scores) of the predominant cartilage constituents (GAG, COL, and H2O), and then the relative Raman scores are converted to absolute metrics using the non-linear transformation curves obtained from the phantom calibration. This step involves applying the non-linear curve fitting model to map the relative scores to absolute biochemical concentration. By repeating this process for each Raman spectrum obtained from across the sample, a 2D tissue map of absolute tissue concentrations can be obtained.

[0172] Further details will become apparent from the following detailed description.

[0173] Creation of an explant model system to validate quantitative Raman of connective tissues:

[0174] In order to test the phantom-based approach to quantitative Raman spectroscopy, we first utilized an experimental model system to generate cartilage specimens with a large variation in composition— i.e. different concentration of extracellular matrix (ECM) constituents (GAG, collagen, water) in the tissue. This was achieved by cultivating live bovine cartilage explanted tissues over a period of 20 days, and subjecting them to different chemical environments that induce compositional changes (see Figure 11A and further details below). For explants cultivated in a basal media formulation, only modest changes in GAG, H2O, and mechanical properties (Young's modulus [EY]) were observed over time. For -Dex / -ILlo, explants increased H2O and lost EY over time. For +Dex / +ILlo, explants lost GAG & EY but increased H2O over time. For -Dex / +ILlo, explants exhibited a further loss of GAG & EY and increased further in H2O over time. These different treatment groups were able to produce explants with a wide range of compositions (see Figure ID that shows the range of explants having different collagen, GAG, and water compositions by percentage of each constituent) that is useful to validate our quantitative approach.

[0175] In order to culture the explants, cylindrical disks (03x2mm) of deep zone cartilage were harvested from immature bovine femoral condyles. Explants were cultivated in high glucose DMEM (ImM Na pyruvate, 50ug.mL L-proline, 1% PS / AM antibiotic / antimycotic, 50ug / mL ascorbate-2-phosphate, 0.1% bovine serum albumin) ± dexamethasone (DEX; lOOnM) to induce a variable degree of tissue swelling and ± interleukin-l-alpha (IL-lo ; 50ng / mL) to induce a variable degree of GAG depletion. After 0, 6, 10, or 14 days of culture, explants were subjected to Raman, mechanical (compressive Young's modulus [EY]), and biochemical analysis (gravimetric H2O content and Dimethylmethylene blue [DMMB]-measured GAG content). Additional groups were cultured for 15 days without IL-lo, followed by IL-lo for subsequent 6 days in an attempt to induce GAG depletion after significant swelling.

[0176] To obtain the phantoms, GAG (bovine trachea chondroitin sulfate; Sigma) and hydrolyzed collagen (gelatin from bovine skin; Sigma) powders were mixed at prescribed ratios in PBS in Eppendorf tubes to produce 25 phantoms with every combination of 5 concentrations of GAG (0%, 2.5%, 5%, 7.5%, and 10%), 5 concentrations of collagen (0%, 5%, 10%, 15%, 20%), and corresponding H2O (H2O%=100%-GAG%-COL%). Tubes were mixed at 10-30 RPM (56°C) for 3 hours, followed by casting 50uL of each phantom into an aluminum sample pan for Raman analysis. Figure 18 shows the phantom production in more detail.

[0177] In figure 18, firstly, as shown on the left hand image, the mass of each analyte (water, GAG, collagen) is measured when added to the vessel. Then, as shown in the middle image, for phantoms with gelatin, they must be heated until the gelatin melts, and agitated gently to avoid forming a foam. A tube rotator in an over at 56C, at 0.5-5RPM was used for an hour or until all powder is dissolved.

[0178] Figure 18, right hand side shows that for the different explants obtainable there are two degrees of freedom - the amount of GAG and the amount of collagen, with the remainder being water. For example, if there is 10% GAG and 20% collagen, then for a three analyte explant the remaining 70% must be water. This means that the calibration set has 2 degrees of freedom and that a 2 dimensional array of calibration objects may be made, as shown in Figure 18, right hand side. In the image we space uniformly, but using some Bayesian adjustment to bias the set towards real tissue may result in better real world performance.

[0179] Raman conversion to absolute metrics via phantom calibration

[0180] Having obtained the explant model tissues and phantoms, we then measured Raman spectra of the cultured tissues, as shown in Figure 13 A-B-C. Using phantoms of known concentration of GAG, COL, and H2O, we developed a correction model that offers quantitative Raman measurements. First the Raman spectra were normalized and subjected to a multivariate least-squares linear regression model to calculate the relative contribution, or 'scores' of the predominant cartilage constituents (GAG, COL, and H2O) using equation: Cartilagespectra = GAGscore*GAGREF + COLscore*COI-REF + H2Oscore*H2OREF, where the derived regression coefficients (GAGscore, COLscore, H2Oscore) represent the relative contribution of the predominant constituents of cartilage ECM (GAG, COL, H2O) to the composite cartilage Raman spectra, calculated using the spectra of purified reference chemicals for each ECM constituent: GAGREF=chondroitin sulfate (shark cartilage; Sigma[C4384]); COI_REF=type-II collagen (chicken sternal cartilage; Sigma[C9301]); H2OREF= PBS [2]. For each constituent, the Raman score ranges from a value of zero to one based on the constituent's weighted contribution to the tissue spectra. The high wavenumber range (2700-3800cm-1) was preprocessed using a linear baseline subtraction, area-under- curve normalization, followed by measurement of the area under the organic-content- associated CH2 region (CHarea) and water-content-associated OH region (OHarea) .

[0181] The spectral contribution of cartilage constituents generally reflected their (relative) concentration in the tissue e.g., a lower GAG score observed for IL-1 a treated explants (Figure 13A-B-C) and higher OH area observed for -Dex explants (Figure 13). Due to spectral normalization, this regression analysis itself is not quantitative (R2=0.77 to 0.84; Figure 13A- B-C).

[0182] With respect to the details of the Raman spectroscopy, Raman spectra of explants and phantoms were acquired via a commercial RIP-RPB-785 probe (Ocean Insight), a 785nM fiber-coupled laser (Innovative Photonic Solutions), and a QEPro spectrometer (Ocean Optics). The spectral fingerprint range (800-1800cm-l) was preprocessed and subjected to a multivariate least-squares linear regression model to calculate the relative contribution, or 'scores' of the predominant cartilage constituents (GAG, COL, and H2O) using equation: Ca rti lagespectra = GAGscore*GAGREF + COLscore*COLREF + H2Oscore*H2OREF, Where GAGREF COLREF H OREF are the spectra of purified reference chemical [2] (Fig 13a, b). The areas under the OH and CH2 peaks in the Raman high-wavenumber region was measured to yield two additional metrics related to tissue hydration and organic content, respectively.

[0183] To correct for the non-linearity:

[0184] As discussed above, we first created a range of phantom calibration samples with known concentration of GAG, COL and water and measured the Raman spectra of these. We then converted the relative GAG / COL / water scores to absolute metrics using non-linear curve fitting. The concentration of each constituent was estimated using a 2D polynomial curve-fit of two Raman metrics using the Matlab curve fitting toolbox (sfit). A representative curve-fit is depicted in Figure 13D. Raman scores were selected by minimizing model error against known constituent concentrations in phantoms. After phantom conversion, the Raman score was able to predict real concentration with a high degree of accuracy (Figure 14D-E) (R2=0.89 for water and R2=0.82 for GAG). Confidence interval of H2O estimation was +2.2% / -3.1%. These results show that by correcting the Raman scores to absolute values using a non-linear transformation we can achieve accurate quantitative analysis.

[0185] From proof of concept to quantitative Raman histology

[0186] To achieve quantitative Raman histology, we aimed to utilize Raman spectroscopy as a nondestructive and spatially resolved technique for analyzing histological sections of connective tissues. This was accomplished by employing phantom correction to each Raman image acquired under a Raman microscope (Figures 15 and 19). Phantom correction or calibration, a critical step, compensates for the non-linearity in the Raman spectra of tissue samples. This non-linearity arises due to tissue light absorption and scattering, which can distort the Raman signals and make them less accurate for quantitative analysis. By using phantoms with known concentrations of glycosaminoglycans (GAG), collagen (COL), and water (H2O), the phantom correction model can accurately convert the relative Raman scores into absolute biochemical concentrations. As described above, this correction model involves creating a range of phantom calibration samples, measuring their Raman spectra, and then using a non-linear curve fitting approach to map the relative Raman scores to absolute metrics. This process ensures that the quantitative Raman measurements are highly accurate and reliable, compensating for any distortions caused by the tissue's optical properties.

[0187] Figures 16, 17, and 18 illustrate the method in more detail. Figure 15 shows the typical Raman setup required to obtain the Raman input data. A tissue sample 1536 is placed on a microscope stage 1538 that is capable of movement in at least the X and Y dimensions. A laser 1528 provides Raman excitation light via band pass filter 1530, dichroic mirror 1523, and objective lens 1534 to focus the Raman excitation light onto the tissue sample 1536. Raman light then generated in the tissue sample in response to the Raman excitation light then passes back through the objective lens 1534 and the dichroic mirror (the Raman light being at a different wavelength than the excitation light), where it then passes through a long pass filter 1526 and onto a tube lens 1524, which focuses the Raman light on to spectrometer 1522, which measures the spectrum of the Raman light to generate a Raman spectrum for the particular point on the sample being imaged, and passes the spectrum data to general purpose computer 1540 for processing. By moving the sample using the XYZ stage 1536, respective Raman spectra can be built up in a 2D Raman spectra map across the tissue, by taking a Raman spectrum at each respective 2D point on the tissue.

[0188] As shown in Figure 16, however, in the present embodiment the collected Raman data is then subject to a correction based on the calibration model that has been derived, for example as shown in Figure 17, and as explained previously. As shown in Figure 17, multiple phantoms 17.8 having known concentrations of GAG (17.4), collagen (17.2) and water (17.6) are used to generate a non-linear calibration model 17.10, which produces correction curves 17.12. When then producing a corrected image, Raman tissue data 17.14 for example obtained from the arrangement of Figure 15 is first processed to determined the relative concentrations of constituents, and then the relative concentrations are then subject to the correction curves 17.12, to give a corrected, absolute concentration measurement 16.16 at the imaging point. By repeating for different points across a scan region, then a 2D map of absolute values of biochemical components, such as GAG, can be obtained across the scanned region of interest.

[0189] In embodiments of the disclosure, processing of data is undertaken by a suitable programmed computer 1540, which receives the Raman spectrum data and is suitably programmed so as to be able to apply the processing steps and methodology described above to obtain first the relative concentrations of biochemical constituents, and then the absolute concentrations of biochemical constituents in the tissue.

[0190] By applying this methodology, therefore, we successfully demonstrated the ability to quantify key biochemical components, specifically glycosaminoglycans (GAGs), which are crucial for the structural and functional integrity of connective tissues. To visualize the distribution of GAGs within the tissue, we generated false-color maps from Raman spectral data. An example is shown in Figure 19. These maps provide a clear representation of the spatial variation in GAG concentrations across the histological section, enabling precise localization and quantification within the tissue architecture. This approach highlights the power of Raman spectroscopy to integrate molecular specificity with spatial context, which is essential for advancing the understanding of connective tissue composition and pathology. The results not only validate the feasibility of using Raman spectroscopy for quantitative analysis of GAGs but also demonstrate the potential for broader applications in quantitative Raman histology. This innovative technique allows for detailed mapping of biochemical components in situ, offering new avenues for studying tissue health, disease progression, and responses to therapeutic interventions. By highlighting quantitative Raman histology with false-color maps, we showcase the capability of this approach to deliver actionable molecular insights into the complex composition of connective tissues.

[0191] Part 2 Conclusion

[0192] This work advances a novel Raman assay platform for achieving non-destructive quantitative measures of cartilage biochemical composition. The work generalizes to an arbitrary number of components. Here we use only GAG, COL and water. Here we showed it in a relatively simple system of 3 components. Extension to more components requires a larger number of components to be included in the phantom development. In order for this approach to work it requires a priori knowledge of the make up of the biological systems. Connective tissues represent a prime example since +95 of the tissue wet weight can be attributed to GAG, COL and water.

[0193] The in vitro cultivation of cartilage explants serves as a model to explore the ability of phantom-trained Raman measures to accurately predict concentrations of biochemical constituents in cartilage for a wide range of degenerative states, where the relative proportions of GAG, COL, and H2O vary in the tissue. Our phantom-training set provides a mapping function that converts Raman scores to highly accurate measures of the concentration of biochemical constituents in cartilage, as evidenced by tight 95% confidence intervals when comparing predictions to direct assay measures of GAG and H2O.

[0194] This platform can serve as a standardized analytical tool to non-destructively quantitatively measure the composition of cartilage in response to degeneration and treatment across a range of preclinical and clinical R+D model systems, spanning from ex vivo explant studies, to in vivo animal studies, to RCT clinical trials. Clinically, quantitative Raman monitoring of patients can be performed arthroscopically for measures of cartilage composition in patients, allowing for earlier and more precise determination of tissue pathology and evaluation of tissue response to treatment. With respect to the interplay of the disclosures of parts 1 and 2 respectively of this disclosure, whilst the long focal length arrangement probe of the part 1 could be used for the Raman spectroscopy described in Part 2, this is not essential, and any type of Raman spectroscopy using any known type of Raman spectroscopy probe or microscope suitable for obtaining Raman light from a tissue sample may be used in the method and system of part 2. That is, it is not necessary that the Raman spectroscopy probe or microscope used in part 2 have the features of the Raman probe and system described in part 1.

[0195] Summary of potential instrumentation, methods and applications

[0196] Instrumentation

[0197] 1. Implementation of a long working distance lens or long focal length lens behind the band pass filter to focus the polarized (or non-polarized) laser light into the hollow needle (Used to create in vivo compatible needle probe).

[0198] 2. Use of an ultra-low working distance lens or short focal length lens to preserve polarisation from tissue (ensures polarisation is not scrambled and lost for tissues).

[0199] Method developments

[0200] 1. Use of tissue phantoms to convert relative Raman biomarkers from cartilage tissue specimens to absolute (quantitative) biochemical concentration values.

[0201] Applications

[0202] 1. Use of the thin needle and polarized Raman information for enhanced cancer diagnostics.

[0203] 2. Use of the thin Raman needle for prediction of mechanical properties in cartilage.

[0204] 3. Use of the Raman system for monitoring culture in a bioreactor.

[0205] Unless the context clearly requires otherwise, throughout the description and the claims, the words "comprise," "comprising," "include," "including," and the like are to be construed in an inclusive sense, as opposed to an exclusive or exhaustive sense; that is to say, in the sense of "including, but not limited to."

[0206] The words "coupled" or "connected" or "tied", as generally used herein, refer to two or more elements or nodes that may be either directly connected, or connected by way of one or more intermediate elements. Additionally, the words "herein,” "above," "below," and words of similar import, when used in this application, shall refer to this application as a whole and not to any particular portions of this application. Where the context permits, words in the Detailed Description using the singular or plural number may also include the plural or singular number, respectively. The words "or" in reference to a list of two or more items, is intended to cover all of the following interpretations of the word: any of the items in the list, all of the items in the list, and any combination of the items in the list.

[0207] Various modifications, whether by addition, substitution, or deletion will be apparent to the intended reader to provide further embodiments of the present disclosure, any and all of which are intended to be encompassed by the appended claims.

Claims

Claims1. A system for obtaining structural information relating to tissue, the system comprising: a needle probe projecting from a probe housing body, a probe housing body comprising a long focal length lens arrangement, the long focal length lens arrangement positioned within the probe housing body, distal from the junction of the needle probe and the probe housing body; the needle probe being configured to direct light to the tissue and collect reflected Raman scattered light; and a spectrometer configured to image the reflected Raman scattered light to produce at least one Raman spectrum.

2. The system of claim 1, wherein the long focal length lens arrangement has a focal length substantially equal to the distance between the long focal length lens arrangement and the junction of the needle probe and the probe housing body.

3. The system of claim 1, wherein the needle probe comprises a short focal length lens arrangement, the short focal length lens arrangement being positioned at a proximal end of the needle probe.

4. The system of claim 2, wherein the length of the needle probe is substantially equal to the focal length of the long focal length lens arrangement.

5. The system of claim 1, wherein the probe housing body further comprises a band pass filter, the band pass filter being positioned between the long focal length lens arrangement and the junction of the needle probe and the probe housing body.

6. The system of claim 1, wherein the needle probe is configured to direct polarized light to the tissue.

7. The system of claim 6, the system further comprising: a beam splitter configured to split the Raman scattering into two polarized components; and the spectrometer configured to simultaneously and separately image the two polarized components to produce two Raman spectra.

8. The system of claim 1, wherein the light is provided by a multi-mode laser.

9. The system of claim 1, wherein the needle probe has a diameter of less than 5mm.

10. The system of claim 1, wherein the needle probe has a diameter of less than 3mm.

11. The system of claim 1, wherein the needle probe has a diameter of less than 1mm.

12. A method for obtaining structural information relating to tissue, the method comprising: directing light to the tissue using a needle probe; focusing the light onto the tissue using a long focal length lens arrangement such that the light is reflected from the tissue producing Raman scattered light, wherein the long focal length lens arrangement is positioned within a probe housing body, distal from the junction of the needle probe and the probe housing body; collecting the Raman scattered light using the needle probe; imaging the Raman scattered light using a spectrometer to produce at least one Raman spectrum; and processing and analysing the at least one Raman spectra to obtain structural information relating to the tissue.

13. The method of claim 12, wherein the focusing of the light onto the tissue is at least in part dependent upon the long focal length lens arrangement having a focal length substantially equal to the distance between the long focal length lens arrangement and the junction of the needle probe and the probe housing body.

14. The method of claim 12, the method further comprising: focusing the light from the long focal length lens arrangement onto the tissue via a band pass filter, the band pass filter being positioned within the probe housing body between the long focal length lens arrangement and the needle probe.

15. The method of claim 12, wherein the directing of the light to the tissue is achieved by the needle probe and a short focal length lens arrangement, the short focal length lens arrangement being positioned at a proximal end of the needle probe.

16. The method of claim 12, the method further comprising: directing light from a multi-mode laser to the tissue using a needle probe.

17. The method of claim 11, the method further comprising: directing polarized light to the tissue using needle probe.

18. The method of claim 17, the method further comprising:splitting the Raman scattered light into two polarised components using a beam splitter; simultaneously and separately imaging the two polarised components using a spectrometer to produce two Raman spectra; and processing and analysing the two produced Raman spectra to obtain structural information relating to the tissue.

19. A method of determining the biochemical composition of tissue, comprising: obtaining at least one Raman spectrum of a tissue sample, the biochemical composition of which is to be found; processing the Raman spectrum to obtain relative tissue constituent composition data indicating the relative biochemical tissue constituent proportions in the tissue sample; and then applying a non-linear fitted curve model to the relative tissue constituent composition data to determine quantitative absolute measurements of the biochemical tissue constituents.

20. A method according to claim 19, wherein the relative tissue constituent composition data is obtained using a multivariate least-squares linear regression model.

21. A method according to claims 19 or 20, wherein the non-linear fitted curve model is obtained using a a 2D polynomial curve-fit of a plurality of selected Raman data.

22. A method according to claim 21, wherein the Raman data were selected by minimizing model error against known constituent concentrations in a plurality of phantoms used to obtain test data to calibrate the non-linear fitted curve model.

23. A method according to claim 22, wherein the calibration of the non-linear fitted curved model comprises: obtaining a plurality of Raman spectra relating respectively to a plurality of phantom specimens having known biochemical compositions; processing the Raman spectra to obtain relative biochemical constituent concentrations; and then finding a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions; wherein the found non-linear transformation curve is then used as the non-linear fitted curve model.

24. A method according to any of claims 19 to 23, wherein the tissue is animal connective tissue, and the biochemical tissue constituents comprise at least GAG, collagen, and water.

25. A method according to claim 24 when dependent on claim 23, wherein the phantom specimens include specimens having different known concentrations of GAG, collagen, and water.

26. A method of calibrating a non-linear transformation model for the quantification of tissue subject to Raman spectroscopy, comprising: obtaining a plurality of phantom specimens having known biochemical compositions; obtaining a plurality of Raman spectra relating respectively to the plurality of phantom specimens; processing the Raman spectra to obtain relative biochemical constituent concentrations; and then finding a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions.

27. A method according to claim 26, wherein the relative biochemical constituent concentrations are obtained using a multivariate least-squares linear regression model.

28. A method according to claims 26 or 27, wherein the non-linear transformation curve is found using a 2D polynomial curve-fit of a plurality of selected Raman data.

29. A method according to claim 28, wherein the Raman data were selected by minimizing model error against known constituent concentrations in a plurality of phantoms used to obtain test data to calibrate the non-linear fitted curve model.

30. A method according to any of claims 19 to 23, wherein the phantom specimens comprise at least GAG, collagen, and water.

31. A method according to claim 30, wherein the phantom specimens include specimens having different known concentrations of GAG, collagen, and water.

32. A system, comprising: a Raman spectrometer; and a processor arranged to receive and process Raman spectrum data obtained by the Raman spectrometer;the system further comprising processor readable instructions that when executed by the processor cause the processor to: receive at least one Raman spectrum of a tissue sample, the biochemical composition of which is to be found; process the Raman spectrum to obtain relative tissue constituent composition data indicating the relative biochemical tissue constituent proportions in the tissue sample; and then apply a non-linear fitted curve model to the relative tissue constituent composition data to determine quantitative absolute measurements of the biochemical tissue constituents.

33. A system according to claim 32, wherein the relative tissue constituent composition data is obtained using a multivariate least-squares linear regression model.

34. A system according to claims 32 or 33, wherein the non-linear fitted curve model is obtained using a a 2D polynomial curve-fit of a plurality of selected Raman data.

35. A system according to claim 34, wherein the Raman data were selected by minimizing model error against known constituent concentrations in a plurality of phantoms used to obtain test data to calibrate the non-linear fitted curve model.

36. A system according to claim 35, wherein the processor is further caused to undertake calibration of the non-linear fitted curved model, the calibration comprising: obtaining a plurality of Raman spectra relating respectively to a plurality of phantom specimens having known biochemical compositions; processing the Raman spectra to obtain relative biochemical constituent concentrations; and then finding a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions; wherein the found non-linear transformation curve is then used as the non-linear fitted curve model.

37. A system according to any of claims 32 to 36, wherein the tissue is animal connective tissue, and the biochemical tissue constituents comprise at least GAG, collagen, and water.

38. A system according to claim 37 when dependent on claim 36, wherein the phantom specimens include specimens having different known concentrations of GAG, collagen, and water.

39. A system, comprising: a Raman spectrometer; and a processor arranged to receive and process Raman spectrum data obtained by the Raman spectrometer; the system further comprising processor readable instructions that when executed by the processor cause the processor to: obtain a plurality of Raman spectra relating respectively to a plurality of phantom specimens having known biochemical compositions; process the Raman spectra to obtain relative biochemical constituent concentrations; and then find a non-linear transformation curve that converts the relative constituent concentrations to the known biochemical compositions.

40. A system according to claim 39, wherein the relative biochemical constituent concentrations are obtained using a multivariate least-squares linear regression model.

41. A system according to claims 39 or 40, wherein the non-linear transformation curve is found using a 2D polynomial curve-fit of a plurality of selected Raman data.

42. A system according to claim 41, wherein the Raman data were selected by minimizing model error against known constituent concentrations in a plurality of phantoms used to obtain test data to calibrate the non-linear fitted curve model.

43. A system according to any of claims 39 to 42, wherein the phantom specimens comprise at least GAG, collagen, and water.

44. A system according to claim 43, wherein the phantom specimens include specimens having different known concentrations of GAG, collagen, and water.

Images