Pushing the boundaries of deep brain stimulation imaging: Role of computer modeling in patient-specific MRI hardware development and surgical lead management

---

During the past decade, deep brain stimulation (DBS) of the human brain has developed into a remarkable treatment for several major disabling neurological and psychiatric disorders. Despite general effectiveness of DBS, its underlying mechanism of action are unclear and its optimal therapeutic protocols remain controversial. Due to its unparalleled soft-tissue contrast, magnetic resonance imaging (MRI) is excellently poised for DBS target verification and evaluation of treatment-induced changes in function of affected brain networks. Today however, high-field MRI (>1.5 T) is largely inaccessible to patients with DBS implants due to safety concerns, and postoperative MRI at 1.5 T is practiced in only few centers and under strict safety guidelines that severely limit image quality.  Improvements in MRI hardware and DBS device management strategies that enable safe and high-resolution imaging of implanted electrodes and target structures can dramatically improve our ability to monitor dynamic changes induced by neurostimulation, in deep brain nuclei, and combine this information with high spatial resolution anatomical images. Such a leap could revolutionize image-guided neurostimulation as it provides MRI as a crucial diagnostic tool for the development of novel therapeutics.

 

The major safety concern in MRI of patients with wire implants is the so-called “antenna effect”. Here, the electric field of the MRI transmitter couples with long implanted DBS leads and causes the specific absorption rate (SAR) of the radiofrequency (RF) energy to significantly amplify at the implant’s tip. This talk introduces two recent approaches that we recently developed at Harvard Medical School to address this issue; First, we have demonstrated that the SAR problem during 3.0 T MRI can be significantly reduced by applying a clinical lead management strategy that optimizes the routing of extracranial leads. Second, we introduced the first generation of patient-specific reconfigurable MRI coils that demonstrated great promise to reduce SAR and image artifacts during DBS imaging at 1.5 T. I will talk about the challenges that we face to propagate these novel methodologies into widespread clinical practice, and the crucial role of computational modeling to obtain a reliable measure of tissue heating and safety margins in realistic patient populations.

---

---

---

---