Low-back pain (LBP) is a leading cause of disability worldwide. When symptoms persist beyond 4–6 weeks, when red flags are suspected, or when precise patient selection for procedures is needed, imaging—primarily MRI (Magnetic Resonance Imaging)—becomes pivotal. The purpose is to provide a pragmatic, radiology-first roadmap that aligns an imaging phenotype with anatomical targets and appropriate image-guided interventions, integrating MRI-based phenotyping with image-guided interventions for subacute and chronic LBP. In this narrative review, we define operational MRI criteria to distinguish radicular from non-radicular phenotypes and to contextualize endplate/Modic and facet/sacroiliac degenerative changes. We then summarize selection and technique for major procedures: epidural and periradicular injections (including selective nerve root blocks), facet interventions with medial branch radiofrequency ablation (RFA), sacroiliac joint injections and lateral branch RFA, basivertebral nerve ablation (BVNA) for vertebrogenic pain, percutaneous disc decompression, minimally invasive lumbar decompression (MILD), and vertebral augmentation for painful fractures. For each target, we outline preferred and alternative guidance modalities (fluoroscopy, CT, or ultrasound), key safety checks, and realistic effect sizes and durability, emphasizing when to avoid low-value or poorly indicated procedures. This review proposes a phenotype-driven reporting template and a care-pathway table linking MRI patterns to diagnostic blocks and definitive image-guided treatments, with the aim of reducing cascade testing and therapeutic ambiguity. A standardized phenotype → target → tool approach can make MRI reports more actionable and help clinicians choose the right image-guided intervention for the right patient, improving outcomes while prioritizing safety and value.

Background/Objectives: Low back pain (LBP) is a clinically relevant symptom in adolescents and adults with scoliosis, but its relationship with curve magnitude is inconsistent, and curve measurement alone therefore cannot explain symptom burden or direct pain management. This narrative review examines imaging findings and clinically plausible mechanisms of scoliosis-related LBP, considered separately for adolescent idiopathic scoliosis, adult idiopathic scoliosis, adult degenerative scoliosis and adult spinal deformity, and their implications for management. Methods: This narrative review was conducted and reported with reference to the SANRA quality domains. PubMed and the Cochrane Library were searched between June and July 2026, with database-specific search strings, search dates and record yields reported in full. Study selection was performed by the first author and reviewed by the co-authors, without independent duplicate screening or formal risk-of-bias assessment. Results: Curve magnitude is not a consistent independent predictor of pain burden, and pain distribution does not reliably follow curve laterality in adult degenerative scoliosis. We propose an author-derived and prospectively unvalidated seven-component framework comprising five candidate pain generators—facet-mediated, vertebrogenic/endplate, foraminal or lateral recess compression, sagittal malalignment with paraspinal overload, and sacroiliac overload—together with biopsychosocial pain modifiers and associated neural-axis abnormalities. Imaging may identify clinically plausible contributors but cannot establish pain causation in isolation; clinical concordance and, for selected interventions, supportive diagnostic procedures remain essential. Standing full-spine radiography, or low-dose biplanar acquisition where serial monitoring is required in adolescents, provides structural and alignment assessment, whereas MRI should be tailored to the clinical scenario and differs in purpose between adolescents and adults. Evidence supporting mechanism-directed interventions derives largely from non-scoliotic chronic LBP cohorts and is identified as such throughout. Conclusions: Imaging in scoliosis-related LBP should identify clinically concordant candidate pain sources rather than merely quantify deformity. Structural abnormalities are not synonymous with pain generators, and treatment selection should integrate imaging with clinical assessment and, when appropriate, supportive diagnostic procedures. The proposed framework and management pathway are author-derived and require prospective validation.