Machine learning for earthquake engineering analysis: Comparing regression models to predict peak ground acceleration

 Corresponding author: Adekunle Job. Copyright © 2025 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution License 4.0. Proteoglycan deposition and Collagen Orientation in Tissue Engineered Cartilage constructs using Digital Densitometry (DD) Imaging and Polarized Light Microscopy (PLM) Techniques Adekunle Job * Department of Technical Physics, University of Eastern Finland, Yliopistonranta 8, 70210, Kuopio Finland. World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 Publication history: Received on 14 March 2025; revised on 03 May 2025; accepted on 05 May 2025 Article DOI: https://doi.org/10.30574/wjarr.2025.26.2.1344 Abstract Background: Articular cartilage is an essential connective tissue providing structural integrity and low-friction movement in joints. Due to its avascular nature, its ability to self-repair is limited, necessitating advancements in tissue engineering for cartilage regeneration. The extracellular matrix (ECM) components, primarily proteoglycans and collagen fibers, play a crucial role in determining the biomechanical properties of engineered cartilage. Understanding proteoglycan deposition and collagen orientation is vital for optimizing tissue engineering strategies. Objective: This study aims to quantify proteoglycan content in histological sections of tissue-engineered cartilage constructs using digital densitometry (DD) imaging and analyze collagen fiber orientation using polarized light microscopy (PLM). Constructs with different cell densities (2M, 5M, and 10M) were evaluated at four time points (Day 1, 7, 14, 21) to assess ECM maturation. Methods: Tissue-engineered cartilage constructs were cultured using gelatin methacrylic anhydride (GELMA), gellan gum (GG) hydrogel, and scaffold-free (SF) approaches. Histological sections were stained with Safranin-O, and DD imaging was performed using a calibrated Nikon microscope to quantify proteoglycan content. Collagen fiber orientation was analyzed using a Leitz Ortholux II POL microscope at 21 orientations, employing Michelson contrast and anisotropy index calculations for alignment assessment. Results: Proteoglycan Deposition (DD Imaging): Proteoglycan content increased significantly from Day 14 to Day 21, with scaffold-free constructs exhibiting the highest deposition. Constructs with higher cell densities (5M and 10M) demonstrated greater proteoglycan accumulation compared to lower-density constructs. Collagen Fiber Orientation (PLM Analysis): Polarized light microscopy revealed that constructs with higher cell densities (5M and 10M) had more organized collagen networks. Scaffold-free constructs exhibited superior collagen alignment compared to hydrogel-based scaffolds. The anisotropy index confirmed increased fiber organization over time, particularly in the 5M and 10M SF groups. Conclusion: The combination of DD imaging and PLM provided a comprehensive assessment of ECM maturation in tissue-engineered cartilage. The study demonstrated that scaffold-free constructs, particularly those with higher cell densities, exhibited enhanced proteoglycan deposition and superior collagen alignment. These findings support scaffold-free tissue engineering approaches for developing functional cartilage replacements in regenerative medicine. Future Directions: Further investigations will explore the molecular mechanisms driving ECM organization and optimize scaffold compositions to enhance cartilage regeneration potential for clinical applications. World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 843 Keywords: PLM; DD; Collagen; GELMA; GG; Proteoglycan 1. Introduction Articular cartilage is a specialized type of connective tissue that covers the ends of bones within a joint. Its primary function is to provide a smooth, low-friction surface that allows the bones to glide smoothly against each other during joint movement. Articular cartilage is found in synovial joints, which are the most common type of joints in the human body and include joints like the knee, hip, shoulder, and elbow [1-2]. Articular cartilage consists of water (80%), articular cartilage cells (chondrocytes, 2%), and extra-cellular matrix, which includes proteoglycans (40% dry weight) and collagen (60% dry weight) [2]. Unlike other tissues, articular cartilage is avascular, meaning it lacks blood vessels, and aneural, meaning it lacks nerve fibers [2]. This limits its ability to heal and regenerate in response to injuries [2]. Collagen provides structural support, while proteoglycans attract water, giving the cartilage its compressibility and ability to withstand loads [2]. Articular cartilage is often divided into different zones based on its structure and function [2-3]. Matured articular cartilage can be categorized into four distinct zones based on the collagen fibril orientation; the superficial, middle, deep, and calcified cartilage zones [3]. In the superficial zone, the collagen fibrils (type II and type ІX) are oriented parallel to the articular surface. In the middle zone, the fibril orientation is irregular (random orientation) due to the transition from parallel orientation at the tissue surface to perpendicular orientation at the deep cartilage. In the deep zone, the orientation is perpendicular to the surface of the articular cartilage; while for the calcified cartilage zones, the cartilage connects to the subchondral bone [1-2]. In the extra cellular matrix (ECM), collagen is the most prevalent structural macromolecule; it accounts for around 60% of cartilage's dry weight. 90% to 95% of the collagen in the extracellular matrix (ECM) is type II collagen, which forms fibrils and fibers entwined with proteoglycan aggregates. Although present, collagen types I, IV, V, VI, IX, and XI make up a small part of the total. The type II collagen fibril network is formed and stabilized with the aid of the minor collagens [2]. Collagen type II forms the elementary component of the cross branded fibrils. While collagen type XI molecule bind covalently to collagen type II molecules and possibly become part of the interior structure of the cross banded fibrils. However, the functions of collagen type IX and type XI remain unclear. Nevertheless, it is assumed that they assist to stabilize and form the collagen fibrils collected fundamentally from collagen type II. The collagen type IX molecules extended portions can help bind the collagen fibril meshwork and interact with collagen meshwork with proteoglycans. Collagen type VI look to form a vital portion of the matrix, surrounding the chondrocytes and aids the attachment of the chondrocytes to the matrix. Collagen type X presence only close to the cells of the calcified zone of articular cartilage and the hypertrophic zone of growth plate that is the area where the longitudinal cartilage septa begin to mineralize, thereby suggests that type X has a part in cartilage militarization [5]. Proteoglycans play a crucial role in the structure and function of cartilage, particularly in articular cartilage found in joints. They are a type of macromolecule consisting of a core protein and long chains of carbohydrates called glycosaminoglycans (GAGs). Proteoglycans contribute to the unique properties of cartilage, such as its ability to resist compressive forces and retain water [7]. Here's how proteoglycan deposition works in cartilage. When pressure is applied to the joint, water within the proteoglycan-rich matrix is squeezed out, and as pressure is released, water is drawn back in, helping the cartilage return to its original shape [5]. This mechanism is crucial for shock absorption and joint lubrication [5]. In conditions like osteoarthritis, the balance between proteoglycan synthesis and degradation can be disrupted, leading to a loss of proteoglycans and deterioration of cartilage [5]. Researchers are investigating strategies to promote proteoglycan synthesis and repair damaged cartilage, including tissue engineering and regenerative therapies. Tissue engineering is a multidisciplinary field that aims to create functional and viable tissues or organs using a combination of cells, biomaterials, and bioactive molecules [9]. The primary component of TE, scaffolds, promote tissue regeneration by fostering an environment that is conducive to cell anchoring, distribution, and functionalization in the presence of signaling molecules [10]. Cartilage tissue engineering aims to create functional cartilage replacements for damaged or degenerated cartilage in joints, such as in cases of osteoarthritis [9]. In tissue engineering, bio-material scaffolds are used to support cell growth and guide tissue formation. These scaffolds often mimic the natural extracellular matrix of the tissue being engineered. For cartilage tissue engineering, scaffolds are designed to support the deposition of proteoglycans, collagen, and other World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 844 components present in native cartilage [10]. While independent methods do not rely on scaffolds, scaffold-dependent procedures employ synthetic or natural biomaterials to create an environment that encourages cell development and interactions [4]. Scaffolds made of hydrogel are becoming a common therapy option for cartilage abnormalities. They are minimally invasive, replicate the natural environment, and may be injected to fill irregularly shaped lesions. Although initial mechanical stability is a key attribute of scaffolds, they come with a number of disadvantages, including cellcharacteristic changes, remodeling resistance, stress shielding, and toxicity associated with degradation [4]. When creating cartilage tissue using tissue engineering approaches, chondrocytes (cartilage cells) or stem cells are seeded onto the scaffolds. Over time, these cells proliferate and differentiate into chondrocyte-like cells that deposit proteoglycans and collagen, forming a tissue-engineered cartilage matrix [9]. Growth factors and bioactive molecules can be incorporated into tissue engineering approaches to enhance proteoglycan deposition and tissue maturation. These factors can stimulate chondrogenic differentiation and promote the synthesis of proteoglycans by the cultured cells [9]. The successful outcome of tissue engineering relies on the maturation of the engineered tissue to closely resemble native tissue [9]. Proteoglycan content and distribution are important markers for assessing the maturity and functionality of tissue-engineered cartilage [9]. Staining articular cartilage sections with Safranin O is a common method used to visualize the presence of glycosaminoglycans (GAGs), which are a major component of the cartilage extracellular matrix. Safranin O is a cationic dye that binds to the negatively charged GAGs, staining them red [11]. The intensity of red staining corresponds to the number of GAGs present in the cartilage matrix [11]. Regions with higher GAG content will appear darker red [11]. GAGs are abundant in the extracellular matrix of cartilage, particularly in areas rich in proteoglycans, such as the territorial matrix around chondrocytes and the inter-territorial matrix. You would observe red staining in these areas [11]. In cases of cartilage degeneration or damage, such as osteoarthritis, there might be a reduced presence of GAGs. This would result in reduced or absent Safranin O staining in the affected areas [12]. Digital densitometry is a technique used in various fields, including radiology, astronomy, and materials science, to measure the density or optical density of a material or substance. It involves the use of digital imaging technology and software to analyze the attenuation or absorption of light or other electromagnetic radiation as it passes through or interacts with a sample. The digital aspect of densitometry involves using digital sensors or cameras to capture images, followed by computer-based analysis to calculate the optical density values. Polarized light microscopy is a traditional technique for visualizing the collagen network structure of articular cartilage. Articular cartilage repair and tissue engineering researches have evolved new request for techniques with capacity of quantitative characterization of the scar and repair tissues, as well as properties of the collagen network. Modern polarized light microscopy can be used to analyze parallelism,birefringence and collagen fibril orientation. Modern instruments are computer based and the measurements are more flexible to perform. Nevertheless, on many numerous occasions the presentation of results brings about difficulties, even errors, due to the demand in the theoretical aspects of the technique [20]. Prior to each PLM measurement session, automated alignment and calibration procedures are run to guarantee measurement reliability. Following alignment, the microscope is calibrated. Calculating the sample's orientation-dependent birefringence required this calibration technique. For impartial study of fibril parallelism or orientation, it is not required. After the polarizer has been rotated step-by-step, grayscale photos are taken. The device will continue to capture photos until the camera's saturation point is achieved [20]. 1.1. Objective of the study The objective of this project work is to get familiar with Quantifying proteoglycan (PG) content concentration in histology sections of three types of tissue engineered cartilage constructs with digital densitometry (DD) imaging techniques, and observing the difference in these biomarkers to demonstrate the influence of these gels and scaffold free on the developing neo-tissue. While the polarized light microscopy (PLM) will be used to study collagen fibers orientation, where the polarization direction of the light is affected by the collagen fibers. 2. Material and Method In this project, there were three types of tissue engineered cartilage constructs which include number of cells (2M, 5M and 10M), and each construct were cultured for four different durations (Day 1, 7, 14 and 21). These tissue constructs were prepared using bone marrow mesenchymal stem cells (BMSCs) encapsulated in gelatin methacrylic anhydride World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 845 (GELMA) (n=16) and gellan gum (GG) hydrogel (n=16) or suspended in the media as pellets (scaffold-free constructs; n=16). These constructs were recovered at four time points (Day 1, 7, 14, 21; n=4/gel/time point) and frozen till the beginning of the project. A set of histological sections were prepared to study PG deposition from frozen samples. The samples were thawed and used to prepare histology sections. This set of histological section were stained using Safranin-o (n=2/gel/time point), a substance that make the proteoglycan to become visible with a reddish color and this was used for Digital Densitometry (DD), where the concentration of safranin-o correlates to the deposition of PG in the samples. Digital densitometry imaging was acquired with a light microscope (Nikon Microphot FXA, Tokyo, Japan) equipped with a CCD-cooled camera (Hamamatsu photonics K.K, Hamamatsu City, Japan, pixel size = 0.14 lm) and 4x and 10x magnification. Images were calibrated against neutral density filters (optical density values 0.0, 0.3, 0.6, 1.0, 1.3, 1.6, 2.0, 2.6, and 3.0) (Schott, Mainz, Germany). Calibration images were taken at every day before DD measurement for different configuration. Each sample were indicated with a symbol M, SP, SF which represent GELMA, gellan gum and scaffold free respectively, and each of the samples were seeded with cells in the range from cell free to, 2 million, 5 million, and 10 million cells. Three slices per sample were measured. The data obtained from these measurements provided information on the tissue-engineered cartilage. The image data obtained from DD is processed using MATLAB. The results obtained will also be used as preliminary data for future studies. The Leitz Ortholux II POL (Leitz Wetzlar, Wetzlar, Germany) microscope body was used for this study's enhanced polarised light microscopy (PLM) instrument, which was calibrated in accordance with Mehta et al. [18]. It was equipped with a monochromatic light source ( λ = 630 nm, Edmund Optics Inc., Barrington, NJ, USA), crossed polarizers (Techspec optics A sample is positioned at a 90-degree angle between a polarizer and an analyzer in the set-up. Samples were measured at 21 orientations spanning 180 degrees, spaced 9 degrees apart. MATLAB program was used for all the analyses to determine the orientation of collagen fibers (Matlab R2023a, Mathworks Inc., Natick, USA). The fibrous collagen network structure of articular cartilage imposes an angle dependence on the intensity of the observed light. Thus, by measuring the sample at several angles, a parallelism index (i.e., anisotropy) can be defined using Michelson contrast [19]. This type of linear polarization enabled the use of Stokes parameters (S0, S1, and S2) in the calculation of the mean fiber angle in each pixel. 𝑆0= 𝐼(0°)+𝐼(90°),𝑆1= 𝐼(0°)−𝐼(90°),𝑆2= 2×𝐼(45°)−𝑆0 ………… (1) Where I: light/signal intensity in each pixel at the specified polarization angle, S0: total intensity of incident light, S1: the amount of linear/horizontal polarization, and S2: the amount of +45° or -45° polarization. The angle of collagen fibers (𝜓) is the orientation of the birefringent structures (i.e., collagen fibers): 𝜓 = 𝑎𝑟𝑐𝑡𝑎𝑛(𝑆2 𝑆1) 2 ……………… (2) Where 0° ≤𝜓≤90° (3). After acquiring the PLM image of the sections, the collagen orientation was estimated on a pixel-by-pixel basis. 3. Result 3.1. Digital Densitiometry (DD) Result In this result proteoglycan deposition in tissue engineered cartilage constructs were analyzed. This was done in histology sections of three types of tissue engineered cartilage contracts with digital densitometry (DD) imaging techniques using magnification 4 and 10. The proteoglycan content based of these magnifications and the days of culture which are day 14 and 21 were compared. Data for day 1 and 7 were inaccessible due to the small sizes of our samples and difficultly to prepare histology slides which make it difficult to have a reasonable result on those days. World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 846 Table 1 and 5 (fig. 1 and fig. 5) are used to show day 14 and 21 for comparison for all amount of cells in both magnification 4 and 10 respectively and while table 2 and 6 (fig. 2 and fig. 6), table 3 and 7 (fig 3 and fig.7) and table 4 and 8 (fig. 4 and fig. 8) are used for day 14 and 21 for comparison with amount of cells 2M, 5M and 10M respectively in both magnification 4 and 10. Table 1 Average values of proteoglycan for day 14 and 21 (Mag. 4) 4X Day 14 average 4X Day 21 average 0M_DAY14_B1 0.1725 0M_DAY21_A1 0.1142 10M_DAY14A_A1 0.2878 0M_DAY21_B1 0.4554 10M_DAY14B_A1 0.2012 0M_DAY21_C1 0.1946 10SF_DAY14_A1 0.1789 10M_DAY21_A1 0.1562 10SF_DAY14_B1 0.1334 10M_DAY21A_B1 0.2275 10SP_DAY14_A1 0.0960 10M_DAY21B_B1 0.2446 10SP_DAY14_B1 0.0893 10SF_DAY21_A1 0.3774 2M_DAY14A_A1 0.1909 10SF_DAY21_B1 0.2325 2M_DAY14A_B1 0.1137 10SP_DAY21A_A1 0.0572 2M_DAY14B_A1 0.0822 10SP_DAY21B_A1 0.0606 2M_DAY14B_B1 0.1244 2M_DAY21A_A1 0.1049 2SF_DAY14_A1 0.1965 2M_DAY21B_A1 0.1545 2SF_DAY14_B1 0.1944 2M_DAY21B_B1 0.2734 5M_DAY14_A1 0.1559 2SF_DAY21_A1 0.2339 5M_DAY14A_B1 0.1156 5M_DAY21_A1 0.1695 5SF_DAY14_A1 0.1412 5M_DAY21A_B1 0.1665 5SF_DAY14_B1 0.2681 5M_DAY21B_B1 0.1606 5SF_DAY21_A1 0.4866 5SF_DAY21_B1 0.2865 Figure 1 The plot of day 21 and day 14 (mag. 4) World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 847 The plot simple revealed the abundance of proteoglycan in day 21 in comparison with day 14 and the mid values in day 21 is thicker than that of day 14, the mid values as seen still confirmed the greater concentration of proteoglycan in day 21 to day 14. Table 2 Average values of proteoglycan for 2million cells scaffold (Mag. 4) day 14 average day 21 averag e 2M_DAY14A_A1 0.1909 2M_DAY21A_A1 0.1049 2M_DAY14A_B1 0.1137 2M_DAY21B_A1 0.1545 2M_DAY14B_A1 0.0822 2M_DAY21B_B1 0.2734 2M_DAY14B_B1 0.1244 2SF_DAY21_A1 0.2339 2SF_DAY14_A1 0.1965 2SF_DAY14_B1 0.1944 Figure 2 The plot of day 21 and day 14 in 2million cells scaffold (mag. 4) The plot still revealed the superiority of day21 with the content of proteoglycan side by side to day 14 in 2 million cells scaffold and the mid values in day 21 still thicker than that in day 14, this also confirmed the level of proteoglycan content higher in day 21. Table 3 Average values of proteoglycan for 5million cells scaffold (Mag. 4) day 14 average day 21 average 5M_DAY14_A1 0.1559 5M_DAY21_A1 0.1695 5M_DAY14A_B1 0.1156 5M_DAY21A_B1 0.1665 5SF_DAY14_A1 0.1412 5M_DAY21B_B1 0.1606 5SF_DAY14_B1 0.2681 5SF_DAY21_A1 0.4866 5SF_DAY21_B1 0.2865 World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 848 Figure 3 The plot of 5million cells scaffold (mag. 4) Day 21 with the content of proteoglycan high than day 14 in 5 million cells scaffold and the mid values in day 21 still thicker, this also confirmed the level of proteoglycan content higher in day 21. Table 4 Average values of proteoglycan for 10million cells scaffold (Mag 4) day 14 average day 21 average 10M_DAY14A_A1 0.2878 10M_DAY21_A1 0.1562 10M_DAY14B_A1 0.2012 10M_DAY21A_B1 0.2275 10SF_DAY14_A1 0.1789 10M_DAY21B_B1 0.2447 10SF_DAY14_B1 0.1334 10SF_DAY21_A1 0.3774 10SP_DAY14_A1 0.0960 10SF_DAY21_B1 0.2325 10SP_DAY14_B1 0.0893 10SP_DAY21A_A1 0.0572 10SP_DAY21B_A1 0.0606 Figure 4 The plot of 5 million cells scaffold (magnification 4) This plot remains consistent with others showing the consistency of day 21 having high proteoglycan compared to day 14. World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 849 Table 5 Average values of proteoglycan for day 14 and 21 (Mag.10) 10X DAY 14 average 10X DAY 21 averag e 10 M_DAY14A_A1 0.2173 0M_DAY21_B1 0.1526 10 M_DAY14A_B1 0.1775 0M_DAY21_C1 0.2364 10 M_DAY14B_A1 0.2420 0M_DAY21A_AI 0.2300 10 M_DAY14B_B1 0.1633 0M_DAY21B_A1 0.2074 10SF_DAY14_A1 0.1970 0SP_DAY21_A1 0.6755 10SF_DAY14_B1 0.3785 10M_DAY21_A1 0.1279 10SP_DAY14_A1 0.1143 10M_DAY21A_B1 0.2496 10SP_DAY14_B1 0.2170 10M_DAY21B_B1 0.2100 2M_DAY14_A1 0.1778 10SF_DAY21_B1 0.2430 2M_DAY14_B1 0.1744 10SF_DAY21A_A1 0.3678 2SF_DAY14_A1 0.2053 10SF_DAY21B_A1 0.3598 2SF_DAY14_B1 0.2121 10SP_DAY21_A1 0.0955 5M_DAY14_A1 0.1834 2M_DAY21_B1 0.1596 5M_DAY14_B1 0.1782 2M_DAY21A_A1 0.2752 5SF_DAY14_A1 0.1497 2M_DAY21B_A1 0.1622 5SF_DAY14_B1 0.2914 2SF_DAY21_A1 0.2456 5M_DAY21_A1 0.4860 5M_DAY21A_B1 0.2189 5M_DAY21B_B1 0.2238 5SF_DAY21_A1 0.4014 5SF_DAY21_B1 0.2415 Figure 5 The plot of day 21 and day 14 (mag.10) World Journal of Advanced Research and Reviews, 2025, 26(02), 842-855 850 The plot revealed the abundance of proteoglycan on day 21 in comparison with day 14 and the mid values in day 21 and day 14, however look similar in thickness, but the day 21 remain high in proteoglycan content. Table 6 Average values of proteoglycan for 2-day million cells scaffold (Mag.10) day 14 average day 21 averag e 2M_DAY14_A1 0.1778 2M_DAY21_B1 0.1596 2M_DAY14_B1 0.1744 2M_DAY21A_A1 0.2752 2SF_DAY14_A1 0.2053 2M_DAY21B_A1 0.1622 2SF_DAY14_B1 0.2121 2SF_DAY21_A1 0.2456 Figure 6 The plot of day 21 and day 14 in 2million cells scaffold (mag.10) The plot still revealed the superiority of day21 with the content of proteoglycan side by side to day 14 in 2million cells scaffold and the mid values in day 21 still thicker than that in day 14, this also confirmed the level of proteoglycan content higher in day 21. Table 7 Average values of proteoglycan for 5million cells scaffold (Mag. 10) day 14 average day 21 average 5M_DAY14_A1 0.1834 5M_DAY21_A1 0.4860 5M_DAY14_B1 0.1782 5M_DAY21A_B1 0.2189 5SF_DAY14_A1 0.1497 5M_DAY21B_B1 0.2238 5SF_DAY14_B1 0.2914 5SF_DAY21_A1 0.4014 5SF_DAY21_B1 0.2415