Haruhiko Kishima

Background: Neuropathic pain (NP) remains difficult to manage because conventional pharmacological therapies often have limited efficacy and intolerable side effects. Noninvasive neuromodulation techniques have emerged as potential alternatives with fewer adverse effects. Meta-analyses of randomized controlled trials suggest that high-frequency repetitive transcranial magnetic stimulation over the primary motor cortex has analgesic effects; however, its efficacy is modest and restricted by the inability to stimulate deeper brain regions. Low-intensity transcranial focused ultrasound stimulation (TUS) is a novel approach that enables precise targeting of deep brain structures and modulation of neural activity. Although promising in preclinical and preliminary human studies, its long-term therapeutic efficacy for NP remains unknown. Objective: This trial aims to explore the therapeutic effects and safety of repeated TUS sessions in patients with NP. Methods: This randomized, participant-blinded, placebo-controlled, 3-arm, parallel-group clinical trial will enroll 39 participants with NP (pain duration ≥3 months; pain intensity ≥4 on the numerical rating scale). Participants will be randomly allocated to one of three groups: (1) TUS targeting the primary motor cortex, (2) TUS targeting the posterior superior insula, or (3) sham stimulation. Each participant will undergo weekly sessions for 8 weeks. The primary outcome is change in weekly average pain intensity scores recorded in a pain diary. Secondary outcomes include additional pain scales, quality of life measures, psychological assessments, quantitative sensory testing, motor cortical excitability, and adverse events. Results: The first participant was enrolled on April 30, 2025, and 20 participants had been enrolled as of February 28, 2026, with enrollment ongoing. The final analysis will be conducted after completion of follow-up and data verification. Conclusions: To our knowledge, this will be the first randomized controlled trial to examine the long-term therapeutic effects of repeated low-intensity TUS for NP. It will provide exploratory proof-of-concept data regarding its clinical efficacy and safety.

BACKGROUND AND OBJECTIVES: Placement of stereotactic electroencephalography (SEEG) electrodes involves insertion of an obturator toward the target under image guidance without direct vision. Thus, tactile feedback during obturator manipulation is key to avoiding complications. Resistance encountered during advancement of the obturator that hinders planned electrode placement has prompted us to investigate the factors contributing to this resistance. METHODS: We analyzed 168 trajectories in 24 consecutive patients who underwent SEEG electrode placement between May 2019 and March 2023: 135 were placed using a 1.5-mm obturator and 33 were placed using a 0.86-mm obturator. The depth at which resistance was encountered was documented intraoperatively and compared with the preplanned trajectory to determine the structure where resistance occurred. Factors that may contribute to resistance were analyzed. RESULTS: Resistance was encountered in a total of 22 trajectories: 19 from the 1.5-mm obturator group and 3 from the 0.86-mm obturator group. Sulcal bank penetration was observed in 15 trajectories (68%) with resistance and in 42/146 trajectories (29%) without resistance. Sulcal bank penetration contributed most to resistance, more than lesion presence or trajectory length. In the 1.5-mm obturator group, insertion angles differed between trajectories with and without resistance (37.1° ± 19.0 vs 58.8° ± 15.5, 95% CI = 11.8-31.6, P = .000055). The difference remained significant when considering all pia penetrations, including at the cortical surface (37.6° ± 18.5 vs 66.1° ± 13.3, 95% CI = 21.3-35.5, P = 2.19 × 10 ⁻¹³ ). CONCLUSION: Sulcal bank penetration is the main factor contributing to resistance; resistance risk increases as the insertion angle decreases. Our findings guide forecasting resistance during sulci-traversing in SEEG electrode implantation, which may be unavoidable for optimal gray matter coverage, although forceful advancement against resistance should be avoided to minimize hemorrhagic risks.

Pain and working memory interact bidirectionally, yet most paradigms treat pain as an extraneous distractor rather than task-relevant content. This study investigated whether thermal sensations can be encoded and updated in working memory in an n-back paradigm. Nineteen healthy adults completed visual (location) and thermal (temperature) single n-back tasks, cross-modal conditions with task-irrelevant distractors, and a dual n-back task across three load levels (1-, 2-, 3-back). Results showed that thermal cues consistently yielded lower accuracy and longer response times (RTs) compared to visual cues (p < 0.001, q < 0.001). While task-irrelevant thermal input tended to prolong RTs in the visual task under low load (p = 0.033, q = 0.099), task-irrelevant visual input showed a trend-level improvement in thermal-task accuracy under high load (3-back; p = 0.020, q = 0.060), consistent with a potential cross-modal effect. Qualitative data indicated that participants strategically transcoded thermal sensations into word or numerical labels to support maintenance. These findings demonstrate that pain can be operationalized as mnemonic content, though its processing incurs significant executive costs due to transcoding demands and resource competition. By shifting focus from “pain-as-interference” to “pain-as-content”, this paradigm offers a principled approach for the mechanistic study of nociceptive working memory updating and provides a foundation for quantifying cognitive interference in clinical pain populations.

The development of algorithms to accurately decode neural information has long been a research focus in the field of neuroscience. Brain decoding typically involves training machine learning models to map neural data onto a preestablished vector representation of stimulus features. These vectors are usually derived from image- and/or text-based feature spaces. Nonetheless, the intrinsic characteristics of these vectors might fundamentally differ from those that are encoded by the brain, limiting the ability of decoders to accurately learn this mapping. To address this issue, we propose a framework, called brain-aligning of semantic vectors, that fine-tunes pretrained feature vectors to better align with the structure of neural representations of visual stimuli in the brain. We trained this model with functional magnetic resonance imaging (fMRI) and then performed zero-shot brain decoding on fMRI, magnetoencephalography (MEG), and electrocorticography (ECoG) data. fMRI-based brain-aligned vectors improved decoding performance across all three neuroimaging datasets when accuracy was determined by calculating the correlation coefficients between true and predicted vectors. Additionally, when decoding accuracy was determined via stimulus identification, this accuracy increased in specific category types; improvements varied depending on the original vector space that was used for brain-alignment, and consistent improvements were observed across all neuroimaging modalities.

This study explored radiomic features that help identify non-contrast-enhancing tumors (nCET) by analyzing regions where contrast-enhancing tumors (CET) transformed into nCET after Bevacizumab (BEV) treatment. The BEV cohort included 24 recurrent GBM (rGBM) patients treated with BEV, showing reduced contrast-enhancement on gadolinium-enhanced T1-weighted imaging (T1Gd) imaging. The 11C-methionine positron emission tomography (Met-PET) cohort consisted of 24 newly diagnosed GBM (nGBM) patients with available Met-PET data. VOIs were created from T2WI, FLAIR, T1Gd, and Met-PET to analyze nCET and T2/FLAIR lesions. After significant radiomic features were identified, a prediction model for nCET was developed in the BEV cohort and subsequently evaluated in the Met-PET cohort. A total of 37 and 46 significant radiomic features were found in the BEV and Met-PET cohorts, respectively. The key feature, T2WI_whole_GLCMcorrelation_1, was selected for predictive modeling. The model demonstrated high accuracy (AUC = 0.93, p < 0.0001) in the BEV cohort, with sensitivity and specificity of 0.91, while the Met-PET cohort showed moderate accuracy (AUC = 0.74, p = 0.0053). Image reconstruction using these features also effectively visualized nCET in nGBM. These findings suggest that radiomic features in CET regions transforming to nCET after BEV treatment harbors valuable information for identifying nCET in GBM.

We present a unique dataset of chronic wireless electrocorticography (ECoG) recordings obtained from two fully implanted nonhuman primates (adult Japanese macaques, Macaca fuscata) spanning hundreds of days post implantation. Each animal was equipped with bilateral subdural ECoG arrays targeting the sensorimotor cortices and a fully implantable wireless transmission unit. The dataset involves multiple tasks, including resting-state measurements, auditory listening paradigms, voluntary button presses, reaching movements, and somatosensory evoked potentials, providing a broad range of behavioural and stimulus conditions. All raw signals, event annotations, and metadata are organized according to the Brain Imaging Data Structure (BIDS) extension for intracranial electrophysiology, ensuring ease of reuse and interoperability with common neurophysiological software. We verified the data quality and stability through impedance monitoring, power spectral analyses, and task-specific event-related measures across the recording period, confirming the reliability and consistency of the ECoG signals. By offering open access to these longitudinal wireless ECoG data, we aim to facilitate the acquisition of new insights into long-term cortical dynamics and advance brain–computer interface (BCI) research.