Hyeon Seo

With the growing use of virtual reality (VR), interaction techniques need to be evaluated across different task demands and in relation to real-world performance. This study compares four VR interaction techniques—Direct Controller, Hand Tracking, Raycasting, and Wrist Rotation—across discrete (text typing), spatial (tangram manipulation), and continuous (slider control) tasks. Twenty participants completed corresponding tasks in both real-world and VR environments. Task completion time and three questionnaires—the Reality-Comparison Questionnaire (RCQ), Technique-Appropriateness Assessment Questionnaire (TAAQ), and User Experience Questionnaire (UEQ)—were used to assess performance and user experience. Participant-level repeated-measures ANOVAs and Holm-adjusted paired tests were conducted. The results showed task-dependent differences among interaction techniques. Direct Controller was faster than Hand Tracking and Raycasting for tangram manipulation, while their typing-time differences were not significant after correction. Direct Controller also received higher agency ratings than Raycasting in typing and tangram. Raycasting had the lowest mean slider completion time, although its difference from Wrist Rotation was not significant after correction, while Wrist Rotation was also associated with higher physical fatigue. Hand Tracking received higher immersion and embodiment ratings than Raycasting in typing and tangram but showed longer completion times in the continuous task. Overall, the findings indicate that interaction performance and user experience vary with task characteristics, highlighting the importance of task-aware selection of VR interaction techniques.

Stroke causes motor and cognitive impairments. Neural stimulation techniques such as repetitive transcranial magnetic stimulation (rTMS) and transcranial direct current stimulation (tDCS) has shown promise in rehabilitating impaired neural functions. However, despite the reported efficacies, stimulus parameters (i.e., applied field amplitude, frequency, and duration) are loosely defined and carry the risk of non-localized heat dissipation, seizure, and scalp discomfort due to high field intensities and direct contact of the devices with the scalp. To advance stroke therapy and to minimize their associated risks, we are exploring an alternative neuromodulation strategy termed ultra high frequency-low intensity magnetic stimulation (UHF-LiMS) designed to enhance functional recovery. For preclinical assessment, focal infarction was induced via photothrombosis in the internal capsule area of the subcortex in male Sprague-Dawley rats, a region well-documented for suppressing spontaneous longitudinal recovery. Sinusoidal magnetic fields (3–10 mT peak amplitude, 400 kHz) were applied, and functional recovery was assessed behaviorally and via imaging. In addition, longitudinal [^18F]-FDG microPET scans and immunohistochemical analysis of c-Fos expression were used to evaluate metabolic activity and neural activation at cellular level. Magnetic stimulation induced substantial somatomotor and somatosensory recovery, with behavioral scores improving by up to ~70% of pre-infarct levels. PET imaging showed reduced stroke-induced diaschisis and increased cortical and subcortical metabolic activity with statistically significant correlations. Elevated c-Fos expression in both cortical and subcortical regions indicates enhanced neuronal activation, serving as a transcriptional signature of neural plasticity. Our study demonstrates that UHF-LiMS effectively restores motor and sensory functions in a white matter stroke model, highlighting its therapeutic potential as a neurostimulation technique.

Transcranial alternating current stimulation (tACS) is a promising noninvasive technique for modulating disrupted neural oscillations in psychiatric disorders and enhancing cognitive functions. However, its efficacy remains debated, partly because neuronal morphology and other microscopic factors critically affect the response to external electric fields. To address this issue and investigate cellular- and network-level mechanisms underlying tACS-induced neural entrainment, we developed a cortical microcircuit model integrating realistic neuronal morphologies, synaptic connectivity, and intrinsic oscillatory dynamics. Using the NEURON simulation environment, we constructed a microcircuit comprising five distinct biophysical cortical neuron models. Neural responses to a range of tACS intensities were assessed with metrics based on cross-correlation, phase coherence, and phase-locking value. While tACS modulates spike timing without significantly altering firing rates, pyramidal neurons are particularly sensitive to external fields compared to interneurons. In addition, tACS can either disrupt or enhance synchronization depending on the endogenous oscillation and stimulation intensity. Our computational study reveals that tACS effects arise from a complex interplay between intrinsic neuronal properties and network dynamics. These findings underscore the importance of neuronal morphology in determining tACS responses and provide insights that may help optimize stimulation parameters for precise neuromodulation in both clinical and research settings.

Deep learning-based analysis of brain magnetic resonance imaging (MRI) plays a crucial role in the early diagnosis of Alzheimer’s disease (AD). However, data scarcity and racial bias present significant challenges to the generalization of diagnostic models. Large-scale public datasets, which are predominantly composed of Caucasian individuals, often lead to performance degradation when applied to other ethnic groups owing to domain shifts. To address these issues, this study proposes a two-stage transfer learning framework. Initially, a 3D ResNet model was pretrained on a large-scale Alzheimer’s disease neuroimaging initiative (ADNI) dataset to learn structural brain features. Subsequently, the pretrained weights were transferred and fine-tuned on a small-scale Korean dataset utilizing only 30% of the data for training. The proposed model achieved superior performance in classifying mild cognitive impairment (MCI), which is crucial for early diagnosis, compared with a model trained from scratch using 70% of the Korean data. Furthermore, it effectively mitigated the significant performance degradation observed when directly applying the pretrained model, demonstrating its ability to resolve the domain-shift issue. This study explored the feasibility of transfer learning to address data scarcity and domain shift issues in AD classification, underscoring its potential for developing AI-based diagnostic systems tailored to specific ethnic populations.

최근 뇌전도(EEG, Electroencephalogram) 전처리 적용이 신호 분석 및 분류 성능에 큰 영향을 미칠 수 있으나, 청각 자극과 청각 상상 발화 데이터에서 독립성분분석(ICA, Independent Component Analysis)이 미치는 영향과 이를 평가하기 위해 딥러닝을 활용한 연구는 아직 명확히 이루어지지 않았다. 본 연구에서는 EEG 신호의 특성 중 하나인 사건유발전위(ERP, Event-Related Potential)를 활용하여 청각 자극에 대한 실제 발화와 상상 발화 데이터에서 ICA 적용 전후의 아티팩트 제거 효과를 분석하고, 딥러닝 모델을 통해 ICA 적용이 모델 성능에 미치는 영향을 평가한다. 본 연구는 ICA가 청각 자극 기반 상상 발화 데이터에서 아티팩트 제거에는 효과적일 수 있으나, ERP 신호의 시간적 일관성 유지와 딥러닝 모델의 성능 향상, 그리고 유용한 특징 추출에는 제한점이 있음을 시사한다.

경두개 저강도 집속 초음파(tFUS, transcranial Focused UltraSound)는 에너지를 뇌의 특정 영역에 집중시켜 신경을 조절하는 기술이다. 이때 두개골에 의해 유도 음압의 정확한 결과 예측이 어려워 의료 영상을 활용한 시뮬레이션이 사용된다. 특히, 두개골 중간에 위치하는 해면골의 존재가 중요한 요인으로 보고되었으나 그 영향은 아직 체계적으로 분석하지 못했다. 이에 본 연구팀은 다른 요인의 영향을 배제하고 해면골의 영향에 집중하여 분석하기 위해 두개골을 단순 구형으로 모델링하고 해면골 비율을 변화시키며 결과를 분석했다. 그 결과, 최대 압력은 해면골 비율이 증가함에 따라 감소하다가 다시 증가하는 경향을 보였으며, 집속 영역은 전반적으로 감소했다. 향후 연구에서 해면골 비율에 따른 유도 음압을 예측하는 인공지능 모델을 개발하고 개인 맞춤형 결과를 제공할 수 있다고 기대한다.