A5023858157- Advanced MRI Techniques and Applications
- Atomic and Subatomic Physics Research
- ECG Monitoring and Analysis
- EEG and Brain-Computer Interfaces
- melanin and skin pigmentation
- Advanced Steganography and Watermarking Techniques
- Image and Video Stabilization
- Advanced Vision and Imaging
- Multimedia Communication and Technology
- Non-Invasive Vital Sign Monitoring
- Anesthesia and Neurotoxicity Research
- Image Processing Techniques and Applications
- Medical Imaging Techniques and Applications
- Vehicle License Plate Recognition
- Face recognition and analysis
- Skin Protection and Aging
- Glioma Diagnosis and Treatment
- Digital Media Forensic Detection
- Autonomous Vehicle Technology and Safety
- Fire Detection and Safety Systems
- Video Surveillance and Tracking Methods
- Sympathectomy and Hyperhidrosis Treatments
- Wireless Body Area Networks
- Advanced Image Processing Techniques
- Biometric Identification and Security
Seoul National University
2024-2025
Seoul National University Hospital
2025
Soongsil University
2018
Chosun University
2008-2016
This study evaluated the effect of an accelerated three-dimensional (3D) T1-weighted pediatric brain MRI protocol using a deep learning (DL)-based reconstruction algorithm on scan time and image quality. retrospective included 46 patients who underwent conventional accelerated, pre- post-contrast, 3D 3T scanner (SIGNA Premier; GE HealthCare) at single tertiary referral center between March 1, 2023, April 30, 2023. Conventional scans were reconstructed intensity Filter A (Conv), whereas...
To assess the feasibility of ultrafast brain magnetic resonance imaging (MRI) in pediatric patients. We retrospectively reviewed 194 patients aged 0 to 19 years (median 10.2 years) who underwent both and conventional MRI between May 2019 August 2020. Ultrafast sequences included T1 T2-weighted images (T1WI T2WI), fluid-attenuated inversion recovery (FLAIR), T2*-weighted image (T2*WI), diffusion-weighted (DWI). Qualitative quality lesion evaluations were conducted on 5-point Likert scales by...
As medical images are created, displayed, transmitted or stored in a digital form, there has been growing interest protecting the against external/internal attackers. In this paper, we propose dual watermarking method(DWM) to protect transmission/storage. proposed DWM provides both robustness and fragileness with embedded watermarks, it can guarantee integrity of image and/or stored. DWM, watermarks carefully avoiding areas region interest(ROI) edge contents image. Based on experimental...
디지털 영상 확대를 위한 스케일링은 고품질의 영상이 요구될수록 많은 수행시간 및 하드웨어 자원량이 요구된다. 본 논문에서는 적은 연산량 자원으로 고품질 영상을 생성하는 이중 선형-3차회선 보간법을 제안한다. 제안한 보간법은 4번의 선형 보간법과 1번의 3차회선 보간법으로 이루어진 수평방향과 수직방향으로 각각 수행하는 구조이다. 실험결과, 제안하는 PSNR과 자원량 측면에서 비교했을 때, 양 보간법보다 우수한 PSNR을 제공했다.
Abstract BACKGROUND Hotspot mutations in the IDH genes serve as tumor-initiating events that define a subset of adult diffuse glioma can further be divided into either oligodendroglioma or astrocytoma subtypes. Despite standard treatment, IDH-mutant gliomas inevitably recur. Previous studies suggest therapeutic resistance may result from combination intratumoral cellular heterogeneity, epigenetic evolution, and acquired genetic alterations. However, how these multilayered molecular features...
The brain connectome helds promise to detect subtle changes in individuals at risk for Alzheimer's disease. We imaged using high resolution diffusion imaging 72 subjects enriched the APOE4 genotype reveal vulnerable networks associated with a composite AD factor including age, genotype, and sex. Sparse canonical correlation analysis (CCA) revealed weight subgraphs involving cuneus, temporal, cingulate cortex, cerebellum. Our results have identified structural weights several factors...
본 논문에서는 지능형 영상 감시시스템을 위한 LDA기반 원거리 얼굴인식 알고리즘을 제안한다. 기존 단일 거리 얼굴영상을 학습으로 사용한 알고리즘은 원거리로 갈수록 얼굴인식률이 떨어지는 문제점이 있다. 실제 거리별 방법은 얼굴인식률은 향상되지만 사용자가 직접 움직이며 학습용 취득해야하는 그러므로 거리에서 취득한 이용해 자동으로 생성하여 사용하는 방법을 제안하는 방법과 동일한 수준의 사용자 협조에서 등록영상을 생성할 수 있는 장점이 실험결과, 제안한 학습영상 기반 알고리즘에 비해 근거리에서 평균 16.3%, 원거리에서 18.0% 향상된 성능을 나타냈고, 방법보다 4.3%감소했지만, 원거리에서는 나타냈다.