Over 100 million people suffer from chronic pain with 26 million having back or leg pain [1]. The use of neuromodulatory devices such as spinal cord stimulation (SCS) and dorsal root ganglion (DRG) stimulation have been shown to be effective in treating chronic pain and are known to be less invasive than other surgical procedures [7,8]. Over the last 10 years, targeting the DRG has been shown advantageous in treating pain compared to traditional SCS [8]. In a 3-month follow-up study, the percentage of subjects receiving ≥50% pain relief was greater with DRG stimulation (81.2%) compared to the SCS (55.7%, P < 0.001). [[8], [9], [10], [11], [12], [13]]. Although effective, implantation of DRG stimulators are invasive and limit use of MRIs and electrocautery which may be needed for medical care after implantation.

Use of external ultrasound is gaining attention as a neuromodulatory device for the treatment of chronic pain and offers a non-invasive option. Though focused ultrasound (FUS) has been used in clinical practice for ablation and is in clinical trials for blood brain barrier opening, neuromodulatory low-intensity focused ultrasound (liFUS) treatment remains in pre-clinical models [[14], [15], [16], [17], [18], [19]]. One logistical reason for this is the inability of devices to reach nervous tissue at relevant depths in humans [20,21]. Our lab has pioneered work introducing liFUS for treatment of neuropathic pain by targeting the DRG in small and large animal models [22,23]. Most recently, we showed that liFUS of the DRG can alter pain behaviors in a swine model for up to 30 days after a single 3 min liFUS treatment [22]. In this study however, the device was not completely non-invasive. It required a thermocouple inserted to the DRG to document safety. Under ideal parameters, temperature changes at the DRG remained <2 °C [22].

In order to translate this device into the clinic, ideally it would be completely non-invasive. Here we examine how safety of parameters used for treatment could be assessed with magnetic resonance thermometry imaging (MRTI). We have shown in previous work that proton resonance frequency-based (PRFS) MRTI techniques that account for temporal B0 drift are able to detect changes in brain temperature associated with needle-based therapeutic ultrasound [24]. These include the corrected PRFS MRTI technique [25] where temperature maps derived from conventional PRFS MRTI are corrected with image phase difference maps interpolated from non-heated regions using a second order polynomial fit, and the referenceless PRFS MRTI technique [26].

Here we demonstrate feasibility of using MRTI to measure the minimal temperature changes (<2 °C) associated with liFUS at the DRG. We also ensure that the materials used in the device do not cause MRTI artifact that would affect temperature monitoring.