Objective: To prospectively evaluate the effect of accelerated deep learning-based reconstruction (Accel-DL) on improving brain magnetic resonance imaging (MRI) quality and reducing scan time compared to that in conventional MRI. Materials and Methods: This study included 150 participants (51 male; mean age 57.3 ± 16.2 years). Each group of 50 participants was scanned using one of three 3T scanners from three different vendors. Conventional and Accel-DL MRI images were obtained from each participant and compared using 2D T1- and T2-weighted and 3D gradient-echo sequences. Accel-DL acquisition was achieved using optimized scan parameters to reduce the scan time, with the acquired images reconstructed using U-Net-based software to transform low-quality, undersampled k-space data into high-quality images. The scan times of Accel-DL and conventional MRI methods were compared. Four neuroradiologists assessed the overall image quality, structural delineation, and artifacts using Likert scale (5- and 3-point scales). Inter-reader agreement was assessed using Fleiss’ kappa coefficient. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated, and volumetric quantification of regional structures and white matter hyperintensities (WMHs) was performed. Results: Accel-DL showed a mean scan time reduction of 39.4% (range, 24.2%–51.3%). Accel-DL improved overall image quality (3.78 ± 0.71 vs. 3.36 ± 0.61, P < 0.001), structure delineation (2.47 ± 0.61 vs. 2.35 ± 0.62, P < 0.001), and artifacts (3.73 ± 0.72 vs. 3.71 ± 0.69, P = 0.016). Inter-reader agreement was fair to substantial (κ = 0.34–0.50). SNR and CNR increased in Accel-DL (82.0 ± 23.1 vs. 31.4 ± 10.8, P = 0.02; 12.4 ± 4.1 vs. 4.4 ± 11.2, P = 0.02). Bland-Altman plots revealed no significant differences in the volumetric measurements of 98.2% of the relevant regions, except in the deep gray matter, including the thalamus. Five of the six lesion categories showed no significant differences in WMH segmentation, except for leukocortical lesions (r = 0.64 ± 0.29). Conclusion Accel-DL substantially reduced the scan time and improved the quality of brain MRI in both spin-echo and gradientecho sequences without compromising volumetry, including lesion quantification.

Purpose: This study aimed to project cancer incidence and mortality for 2025 to estimate Korea’s current cancer burden. Materials and Methods: Cancer incidence data from 1999 to 2022 were obtained from the Korea National Cancer Incidence Database, while cancer mortality data from 1993 to 2023 were acquired from Statistics Korea. Cancer incidence and mortality were projected by fitting a linear regression model to observed age-specific cancer rates against their respective years and then by multiplying the projected age-specific rates by the anticipated age-specific population for 2025. A joinpoint regression model was applied to identify significant changes in trends, using only the most recent trend data for predictions. Results: A total of 304,754 new cancer cases and 84,019 cancer deaths are expected in Korea in 2025. The most commonly diagnosed cancer is projected to be thyroid cancer, followed by the colorectal, lung, breast, prostate and stomach cancers. These six cancers are expected to account for 63.8% of the total cancer burden. Lung cancer is expected to be the leading cause of cancer-related deaths, followed by liver, colorectal, pancreatic, stomach, and gallbladder cancers, together comprising 66.6% of total cancer deaths. Conclusion The increasing incidence of female breast cancer and the rise in prostate and pancreatic cancers are expected to continue. As aging accelerates, cancer commonly found in older adults are projected to rise significantly.

Purpose: The current study provides national cancer statistics and their secular trends in Korea, including incidence, mortality, survival, and prevalence in 2022, with international comparisons. Materials and Methods: Cancer incidence, survival, and prevalence rates were calculated using the Korea National Cancer Incidence Database (1999-2022), with survival follow-up until December 31, 2023. Mortality data obtained from Statistics Korea, while international comparisons were based on GLOBOCAN data. Results: In 2022, 282,047 newly diagnosed cancer cases (age-standardized rate [ASR], 287.0 per 100,000) and 83,378 deaths from cancer (ASR, 65.7 per 100,000) were reported. The proportion of localized-stage cancers increased from 45.6% in 2005 to 50.9% in 2022. Stomach, colorectal, and breast cancer showed increased localized-stage diagnoses by 18.1, 18.5, and 9.9 percentage points, respectively. Compared to 2001-2005, the 5-year relative survival (2018-2022) increased by 20.4 percentage points for stomach cancer, 7.6 for colorectal cancer, and 5.6 for breast cancer. Korea had the lowest cancer mortality among countries with similar incidence rates and the lowest mortality-to-incidence (M/I) ratios for these cancers. The 5-year relative survival (2018-2022) was 72.9%, contributing to over 2.59 million prevalent cases in 2022. Conclusion Since the launch of the National Cancer Screening Program in 2002, early detection has improved, increasing the diagnosis of localized-stage cancers and survival rates. Korea recorded the lowest M/I ratio among major comparison countries, demonstrating the effectiveness of its National Cancer Control Program.

This study aimed to investigate changes in candidemia incidence, species distribution, antifungal susceptibility, initial antifungal use, and mortality trends in Korea before and during the COVID-19 pandemic. A retrospective analysis was conducted on candidemia cases from two tertiary care hospitals in Korea between 2017 and 2022. Data were compared between the pre-pandemic (2017-2019) and pandemic (2020-2022) periods. Statistical methods included incidence rate ratios (IRRs) and multivariate Cox regression to assess 30-day mortality risk factors. A total of 470 candidemia cases were identified, with 48.7% occurring pre-pandemic and 51.3% during the pandemic. While the overall incidence of candidemia remained similar across the two periods (IRR 1.15;p=0.13), the incidence in intensive care units (ICUs) significantly increased during the pandemic (IRR 1.50; p<0.01). The distribution of Candida species did not differ significantly between the two periods. Fluconazole non-susceptibility in C. albicans markedly decreased (10.0% vs. 0.9%, p<0.01), whereas C. glabrata exhibited a significant rise in caspofungin non-susceptibility during the pandemic (0% vs. 22.4%, p<0.01).Echinocandin use increased (21.8% vs. 34.4%; p<0.01), while fluconazole use declined (48.0% vs. 32.8%; p<0.01). Although the 30-day mortality rate was higher during the pandemic (60.2% vs. 57.2%), the difference was not statistically significant (p=0.57).The findings highlight the need for region-specific surveillance and tailored management strategies to improve candidemia outcomes, especially during healthcare disruptions like the COVID-19 pandemic.

Objective: To prospectively evaluate the effect of accelerated deep learning-based reconstruction (Accel-DL) on improving brain magnetic resonance imaging (MRI) quality and reducing scan time compared to that in conventional MRI. Materials and Methods: This study included 150 participants (51 male; mean age 57.3 ± 16.2 years). Each group of 50 participants was scanned using one of three 3T scanners from three different vendors. Conventional and Accel-DL MRI images were obtained from each participant and compared using 2D T1- and T2-weighted and 3D gradient-echo sequences. Accel-DL acquisition was achieved using optimized scan parameters to reduce the scan time, with the acquired images reconstructed using U-Net-based software to transform low-quality, undersampled k-space data into high-quality images. The scan times of Accel-DL and conventional MRI methods were compared. Four neuroradiologists assessed the overall image quality, structural delineation, and artifacts using Likert scale (5- and 3-point scales). Inter-reader agreement was assessed using Fleiss’ kappa coefficient. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated, and volumetric quantification of regional structures and white matter hyperintensities (WMHs) was performed. Results: Accel-DL showed a mean scan time reduction of 39.4% (range, 24.2%–51.3%). Accel-DL improved overall image quality (3.78 ± 0.71 vs. 3.36 ± 0.61, P < 0.001), structure delineation (2.47 ± 0.61 vs. 2.35 ± 0.62, P < 0.001), and artifacts (3.73 ± 0.72 vs. 3.71 ± 0.69, P = 0.016). Inter-reader agreement was fair to substantial (κ = 0.34–0.50). SNR and CNR increased in Accel-DL (82.0 ± 23.1 vs. 31.4 ± 10.8, P = 0.02; 12.4 ± 4.1 vs. 4.4 ± 11.2, P = 0.02). Bland-Altman plots revealed no significant differences in the volumetric measurements of 98.2% of the relevant regions, except in the deep gray matter, including the thalamus. Five of the six lesion categories showed no significant differences in WMH segmentation, except for leukocortical lesions (r = 0.64 ± 0.29). Conclusion Accel-DL substantially reduced the scan time and improved the quality of brain MRI in both spin-echo and gradientecho sequences without compromising volumetry, including lesion quantification.

Objective: To prospectively evaluate the effect of accelerated deep learning-based reconstruction (Accel-DL) on improving brain magnetic resonance imaging (MRI) quality and reducing scan time compared to that in conventional MRI. Materials and Methods: This study included 150 participants (51 male; mean age 57.3 ± 16.2 years). Each group of 50 participants was scanned using one of three 3T scanners from three different vendors. Conventional and Accel-DL MRI images were obtained from each participant and compared using 2D T1- and T2-weighted and 3D gradient-echo sequences. Accel-DL acquisition was achieved using optimized scan parameters to reduce the scan time, with the acquired images reconstructed using U-Net-based software to transform low-quality, undersampled k-space data into high-quality images. The scan times of Accel-DL and conventional MRI methods were compared. Four neuroradiologists assessed the overall image quality, structural delineation, and artifacts using Likert scale (5- and 3-point scales). Inter-reader agreement was assessed using Fleiss’ kappa coefficient. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated, and volumetric quantification of regional structures and white matter hyperintensities (WMHs) was performed. Results: Accel-DL showed a mean scan time reduction of 39.4% (range, 24.2%–51.3%). Accel-DL improved overall image quality (3.78 ± 0.71 vs. 3.36 ± 0.61, P < 0.001), structure delineation (2.47 ± 0.61 vs. 2.35 ± 0.62, P < 0.001), and artifacts (3.73 ± 0.72 vs. 3.71 ± 0.69, P = 0.016). Inter-reader agreement was fair to substantial (κ = 0.34–0.50). SNR and CNR increased in Accel-DL (82.0 ± 23.1 vs. 31.4 ± 10.8, P = 0.02; 12.4 ± 4.1 vs. 4.4 ± 11.2, P = 0.02). Bland-Altman plots revealed no significant differences in the volumetric measurements of 98.2% of the relevant regions, except in the deep gray matter, including the thalamus. Five of the six lesion categories showed no significant differences in WMH segmentation, except for leukocortical lesions (r = 0.64 ± 0.29). Conclusion Accel-DL substantially reduced the scan time and improved the quality of brain MRI in both spin-echo and gradientecho sequences without compromising volumetry, including lesion quantification.

Breast cancer is the most common cancer in women worldwide, and its early detection is critical for improving survival outcomes. As a diagnostic and screening tool, mammography can be less effective owing to the masking effect of fibroglandular tissue, but breast US has good sensitivity even in dense breasts. However, breast US is highly operator dependent, highlighting the need for artificial intelligence (AI)-driven solutions. Unlike other modalities, US is performed using a handheld device that produces a continuous real-time video stream, yielding 12000–48000 frames per examination. This can be significantly challenging for AI development and requires real-time AI inference capabilities. In this review, we classified AI solutions as computer-aided diagnosis and computer-aided detection to facilitate a functional understanding and review commercial software supported by clinical evidence.In addition, to bridge healthcare gaps and enhance patient outcomes in geographically under resourced areas, we propose a novel framework by reviewing the existing AI-based triage workflows including mobile ultrasound.

Gastric cancer is one of the most common cancers in both Korea and worldwide. Since 2004, the Korean Practice Guidelines for Gastric Cancer have been regularly updated, with the 4th edition published in 2022. The 4th edition was the result of a collaborative work by an interdisciplinary team, including experts in gastric surgery, gastroenterology, endoscopy, medical oncology, abdominal radiology, pathology, nuclear medicine, radiation oncology, and guideline development methodology. The current guideline is the 5th version, an updated version of the 4th edition. In this guideline, 6 key questions (KQs) were updated or proposed after a collaborative review by the working group, and 7 statements were developed, or revised, or discussed based on a systematic review using the MEDLINE, Embase, Cochrane Library, and KoreaMed database. Over the past 2 years, there have been significant changes in systemic treatment, leading to major updates and revisions focused on this area.Additionally, minor modifications have been made in other sections, incorporating recent research findings. The level of evidence and grading of recommendations were categorized according to the Grading of Recommendations, Assessment, Development and Evaluation system. Key factors for recommendation included the level of evidence, benefit, harm, and clinical applicability. The working group reviewed and discussed the recommendations to reach a consensus. The structure of this guideline remains similar to the 2022 version.Earlier sections cover general considerations, such as screening, diagnosis, and staging of endoscopy, pathology, radiology, and nuclear medicine. In the latter sections, statements are provided for each KQ based on clinical evidence, with flowcharts supporting these statements through meta-analysis and references. This multidisciplinary, evidence-based gastric cancer guideline aims to support clinicians in providing optimal care for gastric cancer patients.