J Clin Neurosci. 2025 Jul 12;140:111467. doi: 10.1016/j.jocn.2025.111467. Online ahead of print.
ABSTRACT
Prolapsed intervertebral disc (PIVD) of the lumbar region is a major cause of low back pain, accounting for a large proportion of morbidity and healthcare expenditure. While MRI is the gold standard for diagnosis, its unavailability and high cost in developing nations require a clinical method for the identification of PIVD. Artificial intelligence (AI) based diagnostic systems provide an alternative, but current models are based largely on radiological rather than clinical parameters. Therefore, this study aims to identify key clinical determinants for diagnosing lumbar PIVD, forming the basis for an AI-driven diagnostic model. Prospective cross-sectional research was performed between October 2023 and January 2024 at a Haryana-based tertiary care hospital. The three-stage methodology adopted for the study included: (1) thorough review of the literature, (2) patient interviews (n = 12) with established lumbar PIVD, and (3) a survey of expert opinion (n = 12) among physiotherapists, neurologists, and neurosurgeons with special interest in spinal disorders. The data were analyzed based on frequency distribution and descriptive statistics. Clinical determinants were grouped into four categories: demographic (age 25-50 years), anthropometric (height, Body Mass Index > 25 kg/m2), symptomatic (low back pain, radiating pain, neurological deficits, abnormal posture, limited lumbar range of motion), and occupational (sitting > 6 h, heavy lifting). Expert verification attested to their relevance in PIVD diagnosis. The identification of these clinical determinants allows for a transition from MRI-dependent diagnosis to AI-assisted clinical evaluation. Incorporating these tested parameters within AI algorithms can improve diagnostic accuracy, maximize patient management, and decrease the dependency on expensive imaging methods.
PMID:40652579 | DOI:10.1016/j.jocn.2025.111467
