Skeletal Radiol DOI 10.1007/s00256-014-1852-3
REVIEW ARTICLE
T2* mapping for articular cartilage assessment: principles, current applications, and future prospects Tobias Hesper & Harish S. Hosalkar & Daniela Bittersohl & Götz H. Welsch & Rüdiger Krauspe & Christoph Zilkens & Bernd Bittersohl
Received: 4 December 2013 / Revised: 8 February 2014 / Accepted: 17 February 2014 # ISS 2014
Abstract With advances in joint preservation surgery that are intended to alter the course of osteoarthritis by early intervention, accurate and reliable assessment of the cartilage status is critical. Biochemically sensitive MRI techniques can add robust biomarkers for disease onset and progression, and therefore, could be meaningful assessment tools for the diagnosis and follow-up of cartilage abnormalities. T2* mapping could be a good alternative because it would combine the benefits of biochemical cartilage evaluation with remarkable features including short imaging time and the ability of highresolution three-dimensional cartilage evaluation—without the need for contrast media administration or special hardware. Several in vitro and in vivo studies, which have elaborated on the potential of cartilage T2* assessment in various cartilage disease patterns and grades of degeneration, have been reported. However, much remains to be understood and certain unresolved questions have become apparent with these studies that are crucial to the further application of this technique. This review summarizes the principles of the technique and current applications of T2* mapping for articular cartilage assessment. Limitations of recent studies are discussed and the potential implications for patient care are presented. T. Hesper : D. Bittersohl : R. Krauspe : C. Zilkens : B. Bittersohl Department of Orthopaedics Medical Faculty, University Düsseldorf, Düsseldorf, Germany H. S. Hosalkar Center of Hip Preservation and Children’s Orthopaedics, San Diego, CA, USA G. H. Welsch MR Center, Department of Radiology, Medical University of Vienna, Vienna, Austria B. Bittersohl (*) Department of Orthopaedics, Heinrich-Heine University, Medical School, Düsseldorf Moorenstrasse 5, 40225 Düsseldorf, Germany e-mail: [email protected]
Keywords Magnetic resonance imaging . T2* mapping . Cartilage . Review
Introduction Osteoarthritis (OA) is a multifactorial progressive degenerative joint disease that in its natural course eventually becomes symptomatically and functionally debilitating [1]. With advances in joint preservation surgery related to altering the course of the disease progression via early intervention, it is critical to accurately and objectively assess the cartilage status at various stages of the disease process in a reproducible manner [2]. Since the original conception of magnetic resonance (MR) as an imaging tool by Damadian in the early 1970s [3], MR imaging (MRI) has since evolved into being the most precious tool for assessing the degree and extent of degenerative changes in articular cartilage [4]. However, despite technical developments—such as the use of high MR field strengths and dedicated cartilage-specific sequences—that have considerably improved the accuracy of cartilage assessment in particular with regard to morphological changes, its ability to detect subtle cartilage matrix alterations that occur early in the course of OA (for example, disorder of the collagen fiber network, alterations in water content, glycosaminoglycan [GAG] depletion) [5] remains limited [6]. In recent years, several MRI techniques that are sensitive to biochemical changes within cartilage have been reported. This includes the techniques of delayed gadolinium-enhanced MRI of cartilage (dGEMRIC), which is sensitive to the charge density of cartilage contributed by the GAG content within cartilage) [7], T2/T2* mapping (sensitive to water content and collagen fiber network) [8, 9], and others [10–13]. These “biochemically sensitive” MRI techniques are able to add robust biomarkers for disease onset and progression with varying degrees of sensitivity and specificity, and therefore,
Skeletal Radiol
could be meaningful assessment tools for the diagnosis and follow-up of cartilage abnormalities. Compared with other biochemically sensitive MRI techniques (Table 1) [7, 9–19], T2* mapping has unique features including speed of imaging, high image resolution, and the ability to carry out isotropic three-dimensional (3D) cartilage evaluation [20]. It is also easy to implement on clinical MRI systems, as pulse sequences and inline processing software for generating quantitative T2* maps are available commercially. In addition, there is no need for contrast media administration or special hardware. This scientific review aims to outline the facts and current applications of T2* mapping for articular cartilage assessment. The basics of the technique and potential implications for patient care are outlined. Limitations of recent studies are discussed and future prospects are presented.
Composition and structure of articular cartilage It is helpful to review the anatomy prior to embarking on any advanced cartilage MR assessment. Articular cartilage is a composite of chondrocytes (1–4 %) and extracellular elements including water (60–85 %), collagen fibers (10–20 %), and proteoglycans (5–10 %) [21, 22]. These components form an
interconnected lattice structure that regulates its biomechanical properties. The proteoglycan complexes, which consist of a hyaluronic backbone and proteoglycan side units of GAG chains that comprise a core protein and both chondroitin and keratin sulfate regions, bind collagen fibers and link them together (Fig. 1) [23]. The negatively charged hydrophilic GAG side-chains, which are constrained within the collagen matrix, contribute to the mechanical stiffness of articular cartilage by creating an osmotic pressure that is related to the attraction of positive ions and their surrounding fluid [21]. The tensile strength of the collagen network resists this osmotic swelling. With increasing cartilage compression, collagen fibers and GAG chains are squeezed together and fluid exudates from the cartilage. This compression of cartilage and water outflow, in turn, is limited by the osmotic pressure, friction forces between fluid and solid matrix, and the repulsive force generated by the negative charge density of the proteoglycans that increases as the compression progresses [24]. Based upon differences in collagen fiber orientation and biochemical composition, the articular cartilage can be divided into four zones: 1. The superficial or tangential zone (10–20 % of cartilage thickness; collagen fibers running parallel to the articular surface)
Table 1 Synopsis of selected magnetic resonance imaging (MRI) techniques utilized for compositional cartilage assessment MRI technique
Biochemical assessment
T2 mapping
Water content; collagen fiber network; zonal variation reflecting biochemical composition of cartilage Water content; collagen fiber network; zonal variation reflecting biochemical composition of cartilage Content of cartilage glycosaminoglycan; charge density of cartilage
Advantages
Disadvantages
Various studies on T2 cartilage evaluation Relatively long acquisition time; less sensitive published; commercially available on clinical in early stages of cartilage degeneration; T2 MRI scanners variations related to diurnal effects and imaging parameters; magic angle effect T2* mapping High image resolution; short acquisition time; Prone to susceptibility artifacts (e.g., post3D cartilage assessment; UTE-T2* mapping: surgical debris, tissue interfaces); magic high sensitivity to cartilage changes in deep angle effect; UTE-T2* mapping: technically cartilage zones demanding with long acquisition time dGEMRIC Various studies published; specifically for Need for contrast media administration with cartilage glycosaminoglycan content; new lengthening of total scan time for contrast techniques with ability to carry out 3D medium uptake; contrast media uptake cartilage assessment and short acquisition affected by patient factors (e.g., BMI); time requires consistent protocol T1rho imaging Water content; content of Sensitivity for early cartilage damage; reported Difficult to implement in clinical/research cartilage glycosaminoglycan correlation with radiographic cartilage routine; requires high field strengths and damage, joint pain, and functional high RF pulse energy levels; affected by impairment cartilage orientation within main magnetic field Sodium imaging Content of cartilage Specifically for cartilage glycosaminoglycan Technically demanding; requires high field glycosaminoglycan content strengths (≥ 3-T), high gradients and special RF coils; low SNR gagCEST imaging Content of cartilage Potentially specifically for glycosaminoglycan Robust gagCEST sequences have not yet been glycosaminoglycan implemented in clinical routine Diffusion imaging Water molecule mobility
Reported homogeneous results in MRI studies after joint preservation surgery
Technically demanding; low SNR and spatial resolution
UTE ultra-short echo time, 3D three-dimensional, dGEMRIC delayed gadolinium-enhanced MRI, BMI body mass index, RF radiofrequency, SNR signal-to-noise ratio, gagCEST glycosaminoglycan chemical exchange saturation transfer
Skeletal Radiol
Fig. 1 Schematic drawing illustrating the zonal anatomy and the structure and composition of articular cartilage. The extracellular components form an interconnected lattice structure in which proteoglycan aggregates of hyaluronic acid and side units of glycosaminoglycan (GAG) chains with hydrophilic chondroitin and keratin sulfate regions
bind collagen fibers and link them together. Both the high density of negatively charged GAG side chains, which attract positive ions and their surrounding fluid, and the tensile strength of the collagen network to resist this osmotic swelling, contributes to the mechanical stiffness of articular cartilage
2. The transitional or intermediate zone (~ 60 % of cartilage thickness; random collagen fiber orientation with collagen fibers bending to form arcades) 3. The radial or deep zone (~ 30 % of cartilage thickness; collagen fibers running perpendicular to the subchondral bone providing anchorage to the underlying calcified matrix) 4. The calcified zone (cartilage–bone interface) [21, 22]
those regions that are not exposed to habitual loading in, for example, peripheral joint regions where cartilage tissue is more prone to shear stress [16].
The superficial zone, which is the most cellular zone, has a high collagen and water content, whereas the content of proteoglycan is low. The transitional zone has a higher proteoglycan content and a lower collagen and water content than the superficial zone. The radial zone has a high proteoglycan content (proteoglycan content is highest in the upper sector of the radial zone) while the collagen and water content is low. Of note, besides these zonal variations, there are also regional differences in the composition and ultra-structure of articular cartilage [25–27]. This is well portrayed by a thicker radial zone and a thinner transitional zone in weight-bearing cartilage regions with a higher GAG content compared with
Cartilage changes in osteoarthritis The alterations in osteoarthritic cartilage, including compositional and structural changes, are manifold; multiple phases and grades of degeneration may co-exist in different areas of the same joint at a given time [5, 21, 28, 29]. In simplistic terms, cartilage degeneration in OA is morphologically characterized by the fraying or fibrillation of the superficial zone in the initial stages followed by fissures extending down to the subchondral bone, cleft formation, the release of fibrillated cartilage debris decreasing the cartilage thickness, and eventually full-thickness cartilage loss as the disease progresses [21]. At a molecular level, there is degradation of the collagen–proteoglycan complex that includes alterations of the collagen fibrils, a disorganization of the collagen network, collagen depletion (advanced stage of degeneration),
Skeletal Radiol
proteoglycan matrix breakdown, and proteoglycan depletion [5, 28]. The initial increase in cartilage water content is attributed to: 1. The exposure of water binding molecules on the collage fibrils 2. The proteoglycan complex breakdown with a higher rate of non-aggregated proteoglycan complexes leading to a more penetrable extracellular matrix 3. The reduced tensile stiffness of the degraded collagen network to resist the osmotic swelling, and potentially 4. The temporary reactive increase in the net rate of the proteoglycan synthesis (chondrocyte response to tissue damage) These are common findings in the early stages of cartilage degeneration. With progression of collagen network degradation and proteoglycan depletion, the cartilage water content eventually decreases [5].
Relaxometry and the T2* signal Relaxation-time mapping is based on the physical characteristics of nuclei relaxation. During a 90° radio frequency (RF) pulse-driven excitation, the precessing nuclei are tilted in the main magnetic field and synchronized to precess in-phase [30]. This tilted in-phase nuclei precession is the basis for successive transverse magnetization and the creation of an RF pulse that induces a voltage in the receiver coil, which is the MR signal that is measured. Nuclei relaxation occurs immediately after the RF pulse. This relaxation results from the exchange of energy between the nuclei and their surroundings (spin–lattice or T1 relaxation) and from nuclei de-phasing due to variations in the precessing frequencies of the nuclei that arise from random interactions between adjacent nuclei (spin–spin or T2 relaxation) [30]. An additional de-phasing effect comes into play if gradient-echo (GRE) MRI is performed. This characteristic effect, which results from local field in-homogeneity due to differences in the magnetic susceptibility among various tissues, chemical shifts, gradients applied to perform spatial encoding, and main magnetic field heterogeneity, is referred to as T2* relaxation [31]. T2* relaxation is unique for GRE imaging because in spin-echo MRI this de-phasing effect is eliminated by the applied 180° refocusing pulse. The relationship between T2 and T2* relaxation can be expressed by the following equation: 1/T2*=1/T2+γΔBinhom where γ represents the gyromagnetic ratio and ΔBinhom is the magnetic field heterogeneity across a voxel [31]. In order to generate a map of relaxation time (T2 or T2*), consecutive images with varying echo times and signal levels are required [32]. Therefore, multi-echo sequences are
implemented where the setting of echo times is selected to target the characteristic value of T2 or T2* in the region of interest (ROI). Subsequently, an exponential function is applied to the measured signal levels providing the relaxation time, which is defined as the decay of the signal by 1/e (37 %) [31]. These relaxation times can be presented pixel-wise either in gray-scale or as color-scale maps. T2 and T2* relaxation are sensitive to water content and interactions between water molecules and collagen fibers [31, 33, 34]. High T2 or T2* values indicate a high water content and superior water molecule mobility and vice versa. In normal articular cartilage, a decrease in T2 and T2* is noted toward the deep cartilage zones where the uniform perpendicular collagen fiber orientation and the high proteoglycan content promote water molecule restriction and T2/T2* decay [7, 35]. Although few studies have noted a correlation between T2 and T2* mapping [8, 36–38], there are significant differences between the two imaging modalities that have led to diverging T2 and T2* values in various grades of cartilage degeneration [20, 39–45]. Because T2* is influenced by both the T2 relaxation and by coherent de-phasing effects, T2* mapping will be influenced by local susceptibility fields, which operate at a macroscopic or microscopic level [31]. The characteristically lower spectrum of T2* values reflects the additional contribution of these coherent de-phasing effects. In addition, T2 mapping spin-echo sequences utilize echo times of~10– 100 ms. Therefore, T2 mapping techniques capture T2 relaxation, which to a large extent is related to bulk water, while they are rather insensitive to T2 signals that decay more rapidly (T2 relaxation
REVIEW ARTICLE
T2* mapping for articular cartilage assessment: principles, current applications, and future prospects Tobias Hesper & Harish S. Hosalkar & Daniela Bittersohl & Götz H. Welsch & Rüdiger Krauspe & Christoph Zilkens & Bernd Bittersohl
Received: 4 December 2013 / Revised: 8 February 2014 / Accepted: 17 February 2014 # ISS 2014
Abstract With advances in joint preservation surgery that are intended to alter the course of osteoarthritis by early intervention, accurate and reliable assessment of the cartilage status is critical. Biochemically sensitive MRI techniques can add robust biomarkers for disease onset and progression, and therefore, could be meaningful assessment tools for the diagnosis and follow-up of cartilage abnormalities. T2* mapping could be a good alternative because it would combine the benefits of biochemical cartilage evaluation with remarkable features including short imaging time and the ability of highresolution three-dimensional cartilage evaluation—without the need for contrast media administration or special hardware. Several in vitro and in vivo studies, which have elaborated on the potential of cartilage T2* assessment in various cartilage disease patterns and grades of degeneration, have been reported. However, much remains to be understood and certain unresolved questions have become apparent with these studies that are crucial to the further application of this technique. This review summarizes the principles of the technique and current applications of T2* mapping for articular cartilage assessment. Limitations of recent studies are discussed and the potential implications for patient care are presented. T. Hesper : D. Bittersohl : R. Krauspe : C. Zilkens : B. Bittersohl Department of Orthopaedics Medical Faculty, University Düsseldorf, Düsseldorf, Germany H. S. Hosalkar Center of Hip Preservation and Children’s Orthopaedics, San Diego, CA, USA G. H. Welsch MR Center, Department of Radiology, Medical University of Vienna, Vienna, Austria B. Bittersohl (*) Department of Orthopaedics, Heinrich-Heine University, Medical School, Düsseldorf Moorenstrasse 5, 40225 Düsseldorf, Germany e-mail: [email protected]
Keywords Magnetic resonance imaging . T2* mapping . Cartilage . Review
Introduction Osteoarthritis (OA) is a multifactorial progressive degenerative joint disease that in its natural course eventually becomes symptomatically and functionally debilitating [1]. With advances in joint preservation surgery related to altering the course of the disease progression via early intervention, it is critical to accurately and objectively assess the cartilage status at various stages of the disease process in a reproducible manner [2]. Since the original conception of magnetic resonance (MR) as an imaging tool by Damadian in the early 1970s [3], MR imaging (MRI) has since evolved into being the most precious tool for assessing the degree and extent of degenerative changes in articular cartilage [4]. However, despite technical developments—such as the use of high MR field strengths and dedicated cartilage-specific sequences—that have considerably improved the accuracy of cartilage assessment in particular with regard to morphological changes, its ability to detect subtle cartilage matrix alterations that occur early in the course of OA (for example, disorder of the collagen fiber network, alterations in water content, glycosaminoglycan [GAG] depletion) [5] remains limited [6]. In recent years, several MRI techniques that are sensitive to biochemical changes within cartilage have been reported. This includes the techniques of delayed gadolinium-enhanced MRI of cartilage (dGEMRIC), which is sensitive to the charge density of cartilage contributed by the GAG content within cartilage) [7], T2/T2* mapping (sensitive to water content and collagen fiber network) [8, 9], and others [10–13]. These “biochemically sensitive” MRI techniques are able to add robust biomarkers for disease onset and progression with varying degrees of sensitivity and specificity, and therefore,
Skeletal Radiol
could be meaningful assessment tools for the diagnosis and follow-up of cartilage abnormalities. Compared with other biochemically sensitive MRI techniques (Table 1) [7, 9–19], T2* mapping has unique features including speed of imaging, high image resolution, and the ability to carry out isotropic three-dimensional (3D) cartilage evaluation [20]. It is also easy to implement on clinical MRI systems, as pulse sequences and inline processing software for generating quantitative T2* maps are available commercially. In addition, there is no need for contrast media administration or special hardware. This scientific review aims to outline the facts and current applications of T2* mapping for articular cartilage assessment. The basics of the technique and potential implications for patient care are outlined. Limitations of recent studies are discussed and future prospects are presented.
Composition and structure of articular cartilage It is helpful to review the anatomy prior to embarking on any advanced cartilage MR assessment. Articular cartilage is a composite of chondrocytes (1–4 %) and extracellular elements including water (60–85 %), collagen fibers (10–20 %), and proteoglycans (5–10 %) [21, 22]. These components form an
interconnected lattice structure that regulates its biomechanical properties. The proteoglycan complexes, which consist of a hyaluronic backbone and proteoglycan side units of GAG chains that comprise a core protein and both chondroitin and keratin sulfate regions, bind collagen fibers and link them together (Fig. 1) [23]. The negatively charged hydrophilic GAG side-chains, which are constrained within the collagen matrix, contribute to the mechanical stiffness of articular cartilage by creating an osmotic pressure that is related to the attraction of positive ions and their surrounding fluid [21]. The tensile strength of the collagen network resists this osmotic swelling. With increasing cartilage compression, collagen fibers and GAG chains are squeezed together and fluid exudates from the cartilage. This compression of cartilage and water outflow, in turn, is limited by the osmotic pressure, friction forces between fluid and solid matrix, and the repulsive force generated by the negative charge density of the proteoglycans that increases as the compression progresses [24]. Based upon differences in collagen fiber orientation and biochemical composition, the articular cartilage can be divided into four zones: 1. The superficial or tangential zone (10–20 % of cartilage thickness; collagen fibers running parallel to the articular surface)
Table 1 Synopsis of selected magnetic resonance imaging (MRI) techniques utilized for compositional cartilage assessment MRI technique
Biochemical assessment
T2 mapping
Water content; collagen fiber network; zonal variation reflecting biochemical composition of cartilage Water content; collagen fiber network; zonal variation reflecting biochemical composition of cartilage Content of cartilage glycosaminoglycan; charge density of cartilage
Advantages
Disadvantages
Various studies on T2 cartilage evaluation Relatively long acquisition time; less sensitive published; commercially available on clinical in early stages of cartilage degeneration; T2 MRI scanners variations related to diurnal effects and imaging parameters; magic angle effect T2* mapping High image resolution; short acquisition time; Prone to susceptibility artifacts (e.g., post3D cartilage assessment; UTE-T2* mapping: surgical debris, tissue interfaces); magic high sensitivity to cartilage changes in deep angle effect; UTE-T2* mapping: technically cartilage zones demanding with long acquisition time dGEMRIC Various studies published; specifically for Need for contrast media administration with cartilage glycosaminoglycan content; new lengthening of total scan time for contrast techniques with ability to carry out 3D medium uptake; contrast media uptake cartilage assessment and short acquisition affected by patient factors (e.g., BMI); time requires consistent protocol T1rho imaging Water content; content of Sensitivity for early cartilage damage; reported Difficult to implement in clinical/research cartilage glycosaminoglycan correlation with radiographic cartilage routine; requires high field strengths and damage, joint pain, and functional high RF pulse energy levels; affected by impairment cartilage orientation within main magnetic field Sodium imaging Content of cartilage Specifically for cartilage glycosaminoglycan Technically demanding; requires high field glycosaminoglycan content strengths (≥ 3-T), high gradients and special RF coils; low SNR gagCEST imaging Content of cartilage Potentially specifically for glycosaminoglycan Robust gagCEST sequences have not yet been glycosaminoglycan implemented in clinical routine Diffusion imaging Water molecule mobility
Reported homogeneous results in MRI studies after joint preservation surgery
Technically demanding; low SNR and spatial resolution
UTE ultra-short echo time, 3D three-dimensional, dGEMRIC delayed gadolinium-enhanced MRI, BMI body mass index, RF radiofrequency, SNR signal-to-noise ratio, gagCEST glycosaminoglycan chemical exchange saturation transfer
Skeletal Radiol
Fig. 1 Schematic drawing illustrating the zonal anatomy and the structure and composition of articular cartilage. The extracellular components form an interconnected lattice structure in which proteoglycan aggregates of hyaluronic acid and side units of glycosaminoglycan (GAG) chains with hydrophilic chondroitin and keratin sulfate regions
bind collagen fibers and link them together. Both the high density of negatively charged GAG side chains, which attract positive ions and their surrounding fluid, and the tensile strength of the collagen network to resist this osmotic swelling, contributes to the mechanical stiffness of articular cartilage
2. The transitional or intermediate zone (~ 60 % of cartilage thickness; random collagen fiber orientation with collagen fibers bending to form arcades) 3. The radial or deep zone (~ 30 % of cartilage thickness; collagen fibers running perpendicular to the subchondral bone providing anchorage to the underlying calcified matrix) 4. The calcified zone (cartilage–bone interface) [21, 22]
those regions that are not exposed to habitual loading in, for example, peripheral joint regions where cartilage tissue is more prone to shear stress [16].
The superficial zone, which is the most cellular zone, has a high collagen and water content, whereas the content of proteoglycan is low. The transitional zone has a higher proteoglycan content and a lower collagen and water content than the superficial zone. The radial zone has a high proteoglycan content (proteoglycan content is highest in the upper sector of the radial zone) while the collagen and water content is low. Of note, besides these zonal variations, there are also regional differences in the composition and ultra-structure of articular cartilage [25–27]. This is well portrayed by a thicker radial zone and a thinner transitional zone in weight-bearing cartilage regions with a higher GAG content compared with
Cartilage changes in osteoarthritis The alterations in osteoarthritic cartilage, including compositional and structural changes, are manifold; multiple phases and grades of degeneration may co-exist in different areas of the same joint at a given time [5, 21, 28, 29]. In simplistic terms, cartilage degeneration in OA is morphologically characterized by the fraying or fibrillation of the superficial zone in the initial stages followed by fissures extending down to the subchondral bone, cleft formation, the release of fibrillated cartilage debris decreasing the cartilage thickness, and eventually full-thickness cartilage loss as the disease progresses [21]. At a molecular level, there is degradation of the collagen–proteoglycan complex that includes alterations of the collagen fibrils, a disorganization of the collagen network, collagen depletion (advanced stage of degeneration),
Skeletal Radiol
proteoglycan matrix breakdown, and proteoglycan depletion [5, 28]. The initial increase in cartilage water content is attributed to: 1. The exposure of water binding molecules on the collage fibrils 2. The proteoglycan complex breakdown with a higher rate of non-aggregated proteoglycan complexes leading to a more penetrable extracellular matrix 3. The reduced tensile stiffness of the degraded collagen network to resist the osmotic swelling, and potentially 4. The temporary reactive increase in the net rate of the proteoglycan synthesis (chondrocyte response to tissue damage) These are common findings in the early stages of cartilage degeneration. With progression of collagen network degradation and proteoglycan depletion, the cartilage water content eventually decreases [5].
Relaxometry and the T2* signal Relaxation-time mapping is based on the physical characteristics of nuclei relaxation. During a 90° radio frequency (RF) pulse-driven excitation, the precessing nuclei are tilted in the main magnetic field and synchronized to precess in-phase [30]. This tilted in-phase nuclei precession is the basis for successive transverse magnetization and the creation of an RF pulse that induces a voltage in the receiver coil, which is the MR signal that is measured. Nuclei relaxation occurs immediately after the RF pulse. This relaxation results from the exchange of energy between the nuclei and their surroundings (spin–lattice or T1 relaxation) and from nuclei de-phasing due to variations in the precessing frequencies of the nuclei that arise from random interactions between adjacent nuclei (spin–spin or T2 relaxation) [30]. An additional de-phasing effect comes into play if gradient-echo (GRE) MRI is performed. This characteristic effect, which results from local field in-homogeneity due to differences in the magnetic susceptibility among various tissues, chemical shifts, gradients applied to perform spatial encoding, and main magnetic field heterogeneity, is referred to as T2* relaxation [31]. T2* relaxation is unique for GRE imaging because in spin-echo MRI this de-phasing effect is eliminated by the applied 180° refocusing pulse. The relationship between T2 and T2* relaxation can be expressed by the following equation: 1/T2*=1/T2+γΔBinhom where γ represents the gyromagnetic ratio and ΔBinhom is the magnetic field heterogeneity across a voxel [31]. In order to generate a map of relaxation time (T2 or T2*), consecutive images with varying echo times and signal levels are required [32]. Therefore, multi-echo sequences are
implemented where the setting of echo times is selected to target the characteristic value of T2 or T2* in the region of interest (ROI). Subsequently, an exponential function is applied to the measured signal levels providing the relaxation time, which is defined as the decay of the signal by 1/e (37 %) [31]. These relaxation times can be presented pixel-wise either in gray-scale or as color-scale maps. T2 and T2* relaxation are sensitive to water content and interactions between water molecules and collagen fibers [31, 33, 34]. High T2 or T2* values indicate a high water content and superior water molecule mobility and vice versa. In normal articular cartilage, a decrease in T2 and T2* is noted toward the deep cartilage zones where the uniform perpendicular collagen fiber orientation and the high proteoglycan content promote water molecule restriction and T2/T2* decay [7, 35]. Although few studies have noted a correlation between T2 and T2* mapping [8, 36–38], there are significant differences between the two imaging modalities that have led to diverging T2 and T2* values in various grades of cartilage degeneration [20, 39–45]. Because T2* is influenced by both the T2 relaxation and by coherent de-phasing effects, T2* mapping will be influenced by local susceptibility fields, which operate at a macroscopic or microscopic level [31]. The characteristically lower spectrum of T2* values reflects the additional contribution of these coherent de-phasing effects. In addition, T2 mapping spin-echo sequences utilize echo times of~10– 100 ms. Therefore, T2 mapping techniques capture T2 relaxation, which to a large extent is related to bulk water, while they are rather insensitive to T2 signals that decay more rapidly (T2 relaxation
