Nicolas Amoretti

Musculoskeletal image-guided interventions are integral to modern radiologic practices, offering precise, minimally invasive solutions for a range of diagnostic and therapeutic needs. Despite their clinical utility, these interventions are often limited by operator dependency, procedural variability, and a lack of workflow standardization. As artificial intelligence technologies continue to evolve, they offer a significant opportunity to enhance the precision, safety, and efficiency of musculoskeletal interventions. This review explores the current and emerging applications of artificial intelligence across the procedural continuum of musculoskeletal image-guided interventions. We examine its role in preprocedural planning—including lesion detection, segmentation, and personalized risk stratification—through to intraprocedural guidance with artificial intelligence–assisted navigation, needle trajectory optimization, and robotics, and finally postprocedural evaluation using radiomics, complication prediction, and natural language processing for structured reporting. We address the critical challenges that limit clinical translation, including the need for external validation, generalizability across diverse patient populations, regulatory clearance, along with ethical considerations such as bias, data governance, and equitable access. We highlight the absence of multicenter trials and open-access datasets tailored to interventional practice. Particular attention is given to the role of artificial intelligence as a decision-support tool rather than a replacement for operator expertise, with emphasis on its potential to enhance training, standardize practice, and expand access to high-quality care. By highlighting both the practical applications and limitations of current technologies and with health care increasingly shaped by digital innovation, radiologists must engage critically with artificial intelligence technologies to ensure safe, effective, and inclusive deployment. This review synthesizes the current state of artificial intelligence in musculoskeletal interventions and outlines future directions, including the need for interventional artificial intelligence registries, foundation model development, and collaborative translational research. By focusing on practical integration, we aim to support radiologists in leveraging artificial intelligence to enhance procedural accuracy, consistency, and patient outcomes.

To highlight the expanding role of interventional radiology in managing high-level athletes by presenting advanced minimally invasive procedures that extend beyond traditional infiltration techniques, aiming to accelerate recovery and preserve athletic performance. This review explores three main interventional procedures currently implemented in elite sports medicine: (1) percutaneous computed tomography–guided screw fixation for isthmic spondylolysis, (2) vertebroplasty for posttraumatic vertebral fractures, and (3) percutaneous lumbar diskectomy under computed tomography and fluoroscopic guidance. Each technique is detailed from patient selection and procedural setup to postoperative outcomes, with emphasis on imaging guidance, anatomical precision, and rapid rehabilitation. In high-level athletes, computed tomography–guided percutaneous screw fixation for isthmic spondylolysis provides immediate biomechanical stabilization, pain relief, and early return to sport, with the option of hardware removal. Vertebroplasty in selected A1-type fractures enables full weight-bearing within 48 hours and eliminates the need for bracing. Percutaneous lumbar diskectomy offers precise decompression with minimal disruption, yielding symptom resolution within days and preserving spinal integrity. Across all procedures, interventional radiology techniques have shown low complication rates, excellent tolerance under local anesthesia, and consistent return to preinjury performance levels. Interventional radiology now plays a central role in elite sports medicine, offering targeted, image-guided, and minimally invasive alternatives to traditional surgery. These techniques align with the demands of high-performance athletes, ensuring fast recovery, sustained functionality, and long-term musculoskeletal health. As interventional radiology continues to evolve, will become an essential pillar in the multidisciplinary care of professional athletes.

Purpose To evaluate clinical and radiological outcomes of a series of patients treated with a removable percutaneous interspinous process spacer (IPS) (LobsterProject® Techlamed®) for symptomatic degenerative lumbar spinal stenosis (DLSS). Methods All patients treated in two centres with this IPS during 2019 were retrospectively reviewed. Procedures were performed under deep sedation or general anaesthesia by two interventional radiologists. Patients were clinically evaluated before intervention and at 3-month follow-up with Visual Analog Scales for pain (VAS), Oswestry Disability Index (ODI) and radiologically with MRI or CT scans. Neural foramina were independently measured for each patient on pre- and post-procedural CT scans by two radiologists. Results Fifty-nine patients were treated in the selected period of which fifty-eight had complete documentation (mean age 71.2 ± 9.2 years [55–92], 32 males, 26 females). Forty-eight interventions were performed under deep sedation and ten under general anaesthesia, without procedural complications. Clinical follow-up at 3 months showed a significant reduction of pain (VAS from 83 ± 9 to 29 ± 19, − 65%; p < 0.001) and an improvement in functional outcomes (ODI from 31 ± 12 to 13 ± 10%, − 58%; p < 0.001). There was one case of unsatisfactory positioning post procedure, two cases of posterior migration at 3-month follow-up and one case of spinous process fracture. Mean neural foramina area increased from 77 ± 23 to 95 ± 27 mm² (+ 26%; p < 0.001) with very good inter-observer reliability (Cronbach’s alpha = 0.899). Conclusion Percutaneous minimally invasive insertion of a removable IPS device demonstrates a favourable safety profile, good clinical outcomes at 3 months, and apparent anatomical increase in foraminal dimensions. Trial registration ClinicalTrials.gov Identifier: NCT05203666—Release Date: 21^st January 2022, retrospectively registered.

Purpose To evaluate the technical efficacy, safety, and reproducibility of automated percutaneous lumbar discectomy (APLD) under CT and fluoroscopic guidance, for treating radiculopathy caused by lumbar disc herniation in patients impervious to conservative treatment. Methods A total of 77 patients with symptomatic lumbar disc herniation were treated with APLD in a prospective multicentric study performed in four centers across three countries. Magnetic resonance imaging and/or computed tomography was used to evaluate the disc herniation before and after the procedure. Only local anesthesia was used during these procedures. Clinical outcomes were measured with the visual analog scale (VAS) for pain at one and 6 months after the procedure. Results Technical success rate was 100% with a mean intervention duration of 30 min (15–45 min). No complications occurred during the procedure. Post-lumbar puncture syndrome occurred in three patients who were successfully treated with blood patches. VAS decreased from a mean of 8 before the intervention to 3 1 month after (p value = 0.001). The requirement for analgesia decreased from 100 to 27%. No statistically significant differences in outcomes were found between the centers. Conclusion APLD with dual imaging guidance under local anesthesia is a safe, feasible, and reproducible technique to treat symptomatic lumbar disc herniation.

Approximately 20% of patients have persistent unexplained pain after total knee arthroplasty (TKA). Currently available treatments are unsatisfactory. The present report describes four patients in whom transcatheter arterial embolization had a remarkable effect on pain after TKA. Abnormal neovessels were identified in all patients. For 48 h, one patient experienced remarkable postprocedural pain at the inner side of the knee that was subsided by level 1 analgesics and another patient development of a spontaneous skin ulceration resolving within 8 days. The mean Knee injury and Osteoarthritis Outcome Score pain subtotal had increased from 39 to 82 one month after treatment. Endovascular occlusion of neovascularization, decreasing chronic inflammation and the growth of unmyelinated sensory nerves may be treatment options for persistent unexplained pain following TKA. Level of Evidence IV, Case report.

Background and objectives: Tumor-related vertebral compression fractures often result in severe back pain as well as progressive neurologic impairment and additional morbidities. The fixation of these fractures is essential to obtain good pain relief and to improve the patients’ quality of life. Thus far, several spine implants have been developed and studied. The aims of this review were to describe the implants and the techniques proposed to treat cancer-related vertebral compression fractures and to compile their safety and efficacy results. Materials and Methods: A systematic MEDLINE/PubMed literature search was performed, time period included articles published between January 2000 and March 2019. Original articles were selected based on their clinical relevance. Results: Four studies of interest and other cited references were analyzed. These studies reported significant pain and function improvement as well as kyphotic angle and vertebral height restoration and maintain for every implant and technique investigated. Conclusions: Although good clinical performance is reported on these devices, the small numbers of studies and patients investigated draw the need for further larger evaluation before drawing a definitive treatment decision tree to guide physicians managing patients presenting with neoplastic vertebral compression fracture.

To assess the feasibility and clinical outcome of combined CT and fluoroscopy-guided percutaneous vertebroplasty in the management of painful Schmorl’s nodes (SN).