
Textural characteristics of subchondral bone in osteoarthritis
Author(s) -
М. А. Кабалык,
Кабалык Максим Александрович
Publication year - 2016
Publication title -
kazanskij medicinskij žurnal
Language(s) - English
Resource type - Journals
eISSN - 2587-9359
pISSN - 0368-4814
DOI - 10.17750/kmj2016-518
Subject(s) - osteoarthritis , medicine , subchondral bone , radiography , histogram , gray level , nuclear medicine , orthodontics , pixel , radiology , articular cartilage , pathology , artificial intelligence , image (mathematics) , alternative medicine , computer science
Aim. To assess the relationship between textural characteristics of the subchondral bone and standard X-ray data, to determine markers of subchondral bone remodeling in gonarthrosis.Methods. The studied group included 92 patients aged 66.1±10.5 years with I-IV grades osteoarthritis by the Kellgren, in the comparison group - 24 volunteers aged 29.6±5.96 years without clinical or radiological signs of gonarthrosis. Standard digital X-ray of the knee joint was performed. On the image, the area of interest was chosen, including a portion of the subchondral bone of 48±2×90±4 pixels of size. According to the area texture, the gray-level histogram and 3D graph of the pixels intensity distribution in area were made.Results. The distribution of individual pixel values relating to the average gray-level values showed an inverse correlation with the disease stage (r=-0.52, p=0.00004) and the presence of large osteophytes (r=-0.40, p=0.002). Extremum of 3D histogram minimum value directly correlated with radiographic stage of gonarthrosis (r=0.42, p=0.0009), patients’ age (r=0.33, p=0.01) and the osteophytes number (r=0.43, p=0.0007). This figure was higher in the group of patients with osteoarthritis (p=0.009) and significantly decreased with the disease progression (p=0.04).Conclusion. For the first time the analysis of 3D surface reconstruction depending on the gray-level pixel values was used, which showed good characteristics on the distinguishing groups of patients with osteoarthritis, and comparability with standard radiographic protocol data; the best results demonstrated the minimum value at 3D histogram that had significant variation depending on the disease stage.