Assessment of Spinal Support Function in Infectious Spondylitis by Artificial Intelligence Algorithms

The objective : using artificial intelligence algorithms to develop a method for assessing the support function of the spine in infectious spondylitis based on sagittal balance and segmental instability parameters. Subjects and Methods . 249 patients with infectious spondylitis were included in a retrospective study, all of them were admitted to National Medical Research Center of Phthisiopulmonology and Infectious Diseases from 2020 to 2022. The etiology of the spinal inflammation was determined by microbiologic and pathomorphologic tests of biological specimens from the focus of spinal destruction. Results . Patients were divided into the groups of those with tuberculous spondylitis (n=114) and those with non-specific spondylitis (n= 35). The frequency of neurological complications and severity of destruction according to Pola classification and Asia-IMSOP scale were assessed in the groups; statistical and correlation analysis of the parameters of pathologic instability of vertebral segments and global sagittal balance was performed. Severe types of support dysfunction taking into account the summarized data of sagittal balance deviations and segmental instability, were diagnosed in 45.7%. A classification has been developed that details and specifies using mathematical parameters, types and degrees of impairment of the support function, taking into account both global sagittal balance data and segmental disorders.