Brain Scans Predict TMS Success for Tinnitus

A two-week course of repetitive transcranial magnetic stimulation (rTMS) successfully reduced tinnitus symptoms for just over half of patients in a new study. The critical factor for success, according to research published in *Frontiers in Neurology*, was not the severity of a patient’s tinnitus, but the specific structure of their brain before treatment began. Machine learning analysis of pre-treatment brain scans identified a single region whose size strongly predicted who would respond to the neuromodulation therapy.

Predicting Treatment Response with Pre-Treatment Brain Scans

Researchers Zhongling Ding, Bo Peng, and Mengfang Gong sought to address a central problem in tinnitus neuromodulation: the inconsistent effectiveness of rTMS. While some patients experience significant relief, others see no benefit. The team hypothesized that pre-existing differences in brain structure could explain this variability, given that tinnitus is associated with structural brain changes and rTMS works by inducing neuroplasticity.

They enrolled 64 patients with chronic subjective tinnitus and 18 healthy controls. All patients underwent a standard 2-week rTMS protocol targeting the left temporoparietal cortex. Response to treatment was defined as a reduction of at least 5 points on the Tinnitus Handicap Inventory (THI) score. Using this criterion, 36 patients (56.25%) were classified as responders.

Before treatment began, each participant received a high-resolution structural MRI scan. The researchers extracted 242 distinct measurements of brain morphology, analyzing the volume and thickness of regions across the entire cortex. Their goal was to find a structural signature that separated future responders from non-responders before any treatment was delivered.

A Single Brain Region Emerges as the Top Predictor

Initial analysis identified ten regional features that differed between the responder and non-responder groups. These features spanned networks involved in attention, emotion, and sensory processing. To determine which were most important for prediction, the team built a machine learning model using the ExtraTrees algorithm.

The model’s performance was strong, achieving an area under the curve (AUC) of 0.85 and an accuracy of 77%. It was particularly good at identifying responders, with a recall rate of 97%. To interpret the model, the researchers used SHAP analysis, a method that ranks the contribution of each feature to the prediction.

One feature stood out: the gray matter volume of the right pars triangularis. This region, part of the right inferior frontal gyrus, was the top predictor of a positive rTMS outcome. Patients destined to respond to treatment had a significantly larger volume in this area compared to those who would not.

The Right Pars Triangularis: A Marker of Neuroplastic Potential

The significance of this finding became clearer when the researchers compared all three groups: responders, non-responders, and healthy controls. The volume of the right pars triangularis was significantly larger in responders (0.90 ± 0.08) than in both healthy controls (0.86 ± 0.06) and non-responders (0.86 ± 0.07). This pattern suggests that a larger volume in this specific region is not simply a marker of having tinnitus, but a specific signature associated with a brain that is primed to respond to rTMS.

Interestingly, the size of this region did not correlate with the degree of symptom improvement (ΔTHI or ΔVAS scores), nor was it related to baseline tinnitus severity. This indicates the relationship is not linear but may represent a threshold effect. The authors propose that a larger right pars triangularis may indicate a greater “neuroplastic reserve,” meaning the brain’s inherent capacity to reorganize in response to the magnetic stimulation is higher.

This work aligns with a growing focus on brain biomarkers for predicting tinnitus treatment success, moving the field toward more personalized medicine. It also underscores the principle of integrated auditory health from cochlea to cortex, where effective treatment requires understanding central brain mechanisms, not just peripheral hearing.

Toward Precision Neuromodulation for Tinnitus

The practical implication of this study is direct. A pre-treatment structural MRI scan, analyzed for the volume of the right pars triangularis, could help clinicians identify which patients are most likely to benefit from a course of rTMS. This stratification could prevent unnecessary cost and time for probable non-responders and allow clinicians to recommend alternative therapies earlier. For those predicted to respond, it could provide confidence in pursuing the treatment.

This approach represents a shift toward precision neuromodulation. Instead of applying a one-size-fits-all protocol, treatment can be guided by individual neurobiology. The finding that brain structure predicts outcome better than clinical symptoms is a powerful reminder that conditions like tinnitus, misophonia, and hyperacusis are fundamentally brain-based disorders.

Further research is needed to confirm this biomarker in larger, independent patient groups. Future studies could also explore whether this structural feature predicts response to other neuromodulation therapies for auditory disorders. For now, this study provides a clear, data-driven path to making rTMS for tinnitus a more targeted and efficient treatment.

The source study, “Pre-treatment brain structural biomarkers for predicting repetitive transcranial magnetic stimulation efficacy in subjective tinnitus,” by Zhongling Ding, Bo Peng, and Mengfang Gong, is available in Frontiers in Neurology (DOI: 10.3389/fneur.2026.1808769).