Imaging modalities which contribute to a better understanding of osteoarthritis (OA) are radiographs, magnetic resonance imaging (MRI), X-ray computed tomography (CT), ultrasound, dual energy absorptiometry and positron emission tomography. Among these modalities, MRI and CT provide three-dimensional images, MRI is classically used in clinical routine to visualize cartilage, joint effusion, ligaments, tendons, meniscus, osteophytes, and bone marrow lesions or subchondral bone cysts (SBCs). In research, semi-quantitative scoring and quantitative assessment of tissue dimension (volume, thickness) are possible on MRI images [1]. Three dimensional sequences recently developed allow thin continuous slices and reduced the partial volume effect [2]. Compositional MRI based on sequences such as T2, T2*, and T1 ρ, relaxation times and glycosamino-glycanes chemical exchange saturation transfert (gagCEST) and sodium imaging can give additional information about cartilage degeneration without contrast agent [1]. Using contrast agent, delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) can give information about the proteoglycanes content [1]. Subchondral bone cysts are associated with an increase risk of knee replacement [3], and may predict symptoms [4]. Based on a micro-CT study ex-vivo, they are formed to concentric (re)arrangement of trabeculae around an enlarged marrow space [5].
CT is known to enable a good visualization of bone and calcified tissue and has been proposed to evaluate calcium deposits which play a role in initiation and disease progression [6]. Dual energy CT provides good diagnostic accuracy for the detection of monosodium urate crystals [7] but requires further investigation in case of calcium pyrophosphate deposition diseases [8].
The progress of OA management requires the development of non-invasive diagnostic methods able to quantify cartilage and subchondral bone changes [9]. Recently, spectral photon-counting computed tomography (SPCCT) has been introduced in the clinical field. They are characterized by new detectors, i.e., energy-resolving photon-counting detector, that simultaneously count photons and resolve their energy [10]. In living patients, SPCCT has, for instance, been already used to characterize kidney stones [11], coronary artery calcium scoring [12], coronary stent [13], lung cancers [14], brain [15] and anatomical structures of temporal bone and wrist [16,17]. SPCCT extends the dual energy CT approach by a multispectral approach [18].
Photon-counting detectors compared to energy-integrating integrator (EI) detectors usually found in conventional clinical CT were demonstrated to reduce the image noise and the dose in describing the fine anatomy of cadaveric wrists [19], and for the quantitative assessment of trabecular bone micro-architecture on vertebral phantoms [20]. The direct converters are based on semi-conductors, the signal generated by a single X-ray photon is short enough to decay before the next photon arrival, then the electronic signal is proportional to the energy of the incoming X-ray and compare to a threshold voltage, each threshold determinining separated energy bins [10, 21].
One of the main application of SPCCT is in K-edge imaging, the principle being that the linear attenuation coefficient of contrast agents such as iodine, gadolinium, gold or bismuth presents discontinuity when crossing their K-ray energy [22]. K-edge imaging was suggested to give potential additional information on cartilage glycosamino-glycanes content in excised osteoarthritic knees [23]. Using a dual contrast agent method, it was possible to assess the proteoglycanes and water content in human knee cartilage [24]. This approach was confirmed in a study using SPCCT arthrography in mono-iodo-acetate injected knees of living pigs [25]. Moreover, SPCCT is able to provide material decomposition maps since the attenuation of each material being dependent of the energy, it becomes possible to obtain biochemical information visualized as color overlay images [10]. Based on this method, the differenciation of calcium pyrophosphate and hydroxyapatite deposits were suggested in in-vitro studies [26,27].
In addition to the improvement of the compromise between the image quality and the dose delivery, photon-counting detectors are able to exploit spectral information and generate virtual mono-energetic images with a potential to improve the contrast between tissues with intrinsic low attenuation. In a previous study, we have demonstrated that virtual mono-energetic images at 60 keV based on SPCCT scanners were the best choice for both bone and cartilage visualization [28].
We present, as a proof of concept, results of the quantitative assessment of cartilage without any contrast agent, and subchondral bone cysts obtained from virtual mono-energetic images of knees specimens acquired with a whole body helical SPCCT. Cartilage assessment will be compared with high resolution mono-energetic synchrotron radiation CT images with phase contrast (which improve the soft tissue contrast) used as benchmark.