A common treatment for low back pain involves fluoroscopy-guided spinal facet joint injections and/or medial branch nerve blocks. Unfortunately, fluoroscopy requires expensive equipment and personnel and exposes patients and healthcare providers to ionizing radiation. Ultrasound offers a safer, lower-cost alternative, but the traditional 2-dimensional (2D) ultrasound systems are limited due to poor image quality, particularly in patients with higher body mass index (BMI).

As an alternative, a novel Bone Enhanced Ultrasound (BEUS) technology uses artificial intelligence (AI) to create real-time 3-dimensional (3D) images of the spine to guide needle placement for these injections. The AI software is trained by overlaying computed tomography (CT) and ultrasound images from a patient dataset to recognize anatomical landmarks. BEUS aims to ultimately replace fluoroscopy for spinal injections, reducing radiation exposure, lowering healthcare costs, and improving accessibility, especially in rural settings where CT and fluoroscopy are unavailable.

A key limitation, however, is that the current AI system is trained based primarily on patients (mostly pediatric) undergoing perioperative assessment of scoliosis. To address this, the current study aims to develop a new, more clinically relevant training AI dataset by collecting spinal ultrasounds from up to 100 adult participants (most/all of whom are followed at the local chronic pain clinic for low back pain) with existing spinal CT or magnetic resonance imaging (MRI) scans. This dataset will be used to retrain the current AI model to enhance the accuracy of 3D spinal reconstructions, thereby improving the clinical relevance of the BEUS system.

Inclusion Criteria:

• Individuals who have had previous CT or MRI scans of their spine in the past 5 years

Exclusion Criteria:

• < 18 years old

Adults who have had previous CT or MRI scans of their spine in the past 5 years