To evaluate whether a custom agentic artificial intelligence (AI) pipeline can overcome the limitations of general-purpose large language model tools, when compared with a generic commercial tool (Google NotebookLM [NBLM]), for generating podcast-style summaries of radiology research articles. Twenty-two PDF-format original research articles published in the April 2025 issue of Radiology were processed using our Programmable, Phoneme-Aware PDF-to-Podcast Pipeline (P5) and NBLM to generate 44 audio episodes. P5 utilizes a multi-agent workflow for script generation, quality assurance, pronunciation enhancement, and audio synthesis. Four radiologists from a pool of 25 (7 generalists and 18 specialists) were randomly assigned to evaluate each blinded audio episode, yielding 176 total evaluations. The primary outcomes were the number of hallucinations (factual errors) per episode and the percentage of hallucination-free episodes. Secondary outcomes included the number of inappropriate statements, mispronunciations, and flow disruptions; the composite quality score (Quality Assessment of Educational Podcasts [QAEP]); the key results coverage score; and overall listener preference. Data were analyzed using generalized linear mixed models. The P5 method produced significantly fewer hallucinations per episode compared with NBLM (mean, 0.32 vs. 0.93; P < 0.001) and a higher proportion of hallucination-free episodes (71.6% [63/88] vs. 56.8% [50/88]; P = 0.013), consistently across generalists and specialists. P5 demonstrated significantly fewer mispronunciations (mean, 0.11 vs. 1.62; P < 0.001) and flow disruptions (mean, 0.19 vs. 1.06; P < 0.001) per episode, as well as higher mean QAEP composite scores (4.62 vs. 4.28; P < 0.001), compared with NBLM, consistently across generalists and specialists. Raters preferred P5 over NBLM in 72.7% (64/88) of comparisons (P = 0.003). Our custom agentic AI pipeline generated podcast-style summaries of radiology research articles with significantly higher quality and greater listener preference than the generic commercial tool.
Selumetinib is approved for the treatment of inoperable plexiform neurofibromas (PN) in patients with neurofibromatosis type 1 (NF1). However, its efficacy in treating NF1-associated diffuse neurofibromas (NF1-DN) or optic pathway gliomas (NF1-OPG) remains unclear. We evaluated the efficacy and safety of selumetinib in these subgroups. This was a sub-analysis of a Korean phase II open-label trial focusing on non-target treatment effects on NF1-DN and NF1-OPG. A total of 88 pediatric and adult patients with NF1-PN (59 children and 29 adults) in this trial had been treated for at least 2 years (~ 26 cycles, 28-day cycle) with oral selumetinib (20 or 25 mg/m², or 50 mg/dose every 12 h). Tumor volume, quality of life (QoL), and visual acuity were assessed. Among the 88 included patients, NF1-DN was diagnosed in 25 (28%), and NF1-OPG in 3 (3%). All NF1-DN patients exhibited disfigurement, two experienced pain, and a partial response (PR; ≥20% tumor reduction at a single time) was achieved in 9 of these cases (36%). The median time to PR was 6 cycles (range, 6-12), and the median time to best response was 18 cycles (range, 6-26), with a median volume change of - 11.9% (range, - 55.4% to + 36.3%). Confirmed PR (cPR; PR sustained for > 6 cycles) was observed in 6 NF1-DN patients (24%), stable disease (SD) was observed in 9 of these patients (36%), and progressive disease (PD) in 10 cases (40%). In a paired comparison, cPR was significantly lower for NF1-DN than for NF1-PN (24% vs. 88%, P < 0.001), and the median best volume reduction was also smaller (- 11.9% vs. -42.1%, P < 0.001). For the 3 NF1-OPG patients, visual impairment was present in all cases at baseline. One patient achieved PR at cycle 12 (- 36.1% from baseline) and complete response (CR) at cycle 26 but the visual acuity did not improve in this case. SD was maintained in the other two patients. Within these subgroups, QoL improved in 16 patients (57%) by cycle 26. No serious adverse events or treatment discontinuations were reported. Selumetinib shows therapeutic potential against NF1-DN, although with a lower response than that observed against NF1-PN. The NF1-OPG findings herein for this drug were exploratory only because of the small number of cases, but they may provide an additional clinical context for evaluating the broader effects of selumetinib across NF1-associated lesions. KCT0003700, cris.nih.go.kr/cris/search/detailSearch.do/19,080, first submission March 4, 2019; first patient enrolled May 14, 2019.
Current guidelines recommend surveillance or diagnostic lobectomy for indeterminate thyroid nodules. Although radiofrequency ablation (RFA) has emerged as a potential alternative, its role remains controversial. Functional imaging modalities, including 99mTc-methoxyisobutylisonitrile scintigraphy and ¹⁸F-fluorodeoxyglucose positron emission tomography, have been used to support risk stratification. This study evaluated the mid-term outcomes of RFA and their predictors in a functionally screened cohort of patients with indeterminate thyroid nodules. This retrospective cohort study included 109 patients with Bethesda category III or IV thyroid nodules who underwent RFA after functional imaging-based evaluation between January 2018 and November 2024. The primary outcome was local treatment failure (LTF), defined as the first occurrence of nodule regrowth, an increase in vital volume, or repeated indeterminate cytology. The secondary outcome was complete disappearance of the nodule. Supplementary outcomes included the volume reduction rate, symptom and cosmetic scores, and biochemical markers. During a median follow-up of 2.2 years (range, 0.7-6.8 years), LTF occurred in 15 of 109 patients (13.8%), whereas complete disappearance was observed in 9 (8.3%) patients. The 3-year LTF-free probability and cumulative incidence of complete disappearance were 85.4% and 6.5%, respectively. Significant reductions were observed in the nodule volume, symptom scores, cosmetic scores, and free T4 and thyroglobulin levels. A greater initial maximum nodule diameter was independently associated with a higher likelihood of LTF (hazard ratio [HR] per 1-cm increase, 2.35; 95% confidence interval [CI], 1.23-4.48; P = 0.010) and a lower likelihood of complete disappearance (HR, 0.16; 95% CI, 0.04-0.60; P = 0.006). At 3 years, 93% of the tumors with an initial diameter <3.0 cm remained free of LTF, whereas no tumors >2.1 cm achieved complete disappearance. In this functionally screened cohort of patients with indeterminate thyroid nodules, RFA demonstrated acceptable outcomes with volume reduction and symptom relief, particularly in patients with nodules <3 cm.
Cerebrospinal fluid (CSF) shunt devices are essential for managing hydrocephalus by diverting excess CSF from the ventricles and thereby reducing symptoms such as headaches, vision problems, and cognitive deficits. A typical shunt system includes a ventricular catheter, reservoir, valve, and distal catheter. These components are usually made of silastic and radiopaque materials for durability and imaging. The systems also include enhancements like programmable valves, anti-siphon devices, and gravitational units to improve function. Shunts are categorized by outflow destination (ventriculoperitoneal, ventriculoatrial, ventriculopleural, and lumboperitoneal) or by valve type (fixed or adjustable). While fixed valves have been the standard of care for years, adjustable valves allow non-invasive pressure changes, reducing the need for repeated surgeries. Radiographic evaluation of shunt placement and function is crucial in clinical care. Common complications associated with the use of such systems include mechanical failure (e.g., obstruction, fracture, and migration), infections, functional issues, and metastasis, which often require surgical revision. This paper offers an overview of CSF shunts, including their structure, functions, imaging characteristics, and potential complications. Additionally, we have summarized the imaging characteristics of currently available programmable shunt devices and briefly discussed other types of commonly used drainage devices along with their specific indications. A thorough understanding of these aspects is critical for effective diagnosis, treatment, and management of hydrocephalus and its related challenges.
Accurate classification of aortic dissection is essential for guiding clinical treatment and facilitating communication between radiologists and clinicians. Traditional classifications, such as the DeBakey and Stanford systems, categorize dissections according to their origin and extent, or by involvement of the ascending aorta. The Stanford classification has gained widespread use due to its clear correlation with the surgical and medical treatments available at the time. However, these traditional classifications provide limited guidance for contemporary management, particularly in the era of endovascular and hybrid surgical techniques for aortic repair. Modern classifications, such as the Type/Entry/Malperfusion (TEM) system and Type B dissection reporting standards proposed by the Society for Vascular Surgery and the Society of Thoracic Surgeons (SVS/STS), incorporate more detailed anatomic and clinical descriptors. The TEM system expands upon the Stanford classification by adding a "non-A non-B" category for arch dissections and incorporating entry tear and malperfusion descriptors. The SVS/STS classification introduces a zone-based model that defines the dissection type by the location and extent of the entry tear. Contemporary guidelines have identified several high-risk imaging features that predict poor outcomes in otherwise uncomplicated dissections. This pictorial review compares these classifications and highlights their implications for radiologic reporting, emphasizing how detailed anatomic and risk-based descriptors better align radiologic interpretations with contemporary management strategies.
The prognostic value of coronary artery calcium (CAC) volume and density was derived from an automated artificial intelligence (AI)-based analysis of non-electrocardiogram-gated chest CT. In this retrospective study, 7,552 asymptomatic adults who underwent chest CT as part of a national health screening program between 2007 and 2014 at two tertiary hospitals were examined for eligibility, of whom 1,109 with detectable CAC were analyzed. CAC density was derived by back-calculation from the Agatston score and CAC volume, both of which were obtained using AI software on chest CT. Differences in the probability of being free from major adverse cardiovascular events (MACE) across the four combined CAC volume-density groups were assessed using Kaplan-Meier curves and restricted mean survival time (RMST). Multivariable Cox proportional hazards models were used to assess the association between CAC volume and density and MACE. Among the 1,109 participants with nonzero CAC (median age, 60.3 years; 87% men), 207 experienced MACE during a median follow-up of 7.7 years. Ten-year RMSTs were 9.45 years in the low-volume-high-density group, 9.07 years in the low-volume-low-density group, 8.03 years in the high-volume-high-density group, and 7.68 years in the high-volume-low-density group. Differences in time to MACE were predominantly driven by CAC volume, with no significant density-related differences within the volume strata. CAC density demonstrated a significant, independent, inverse association after adjusting for CAC volume and clinical covariates (hazard ratio [HR] per increase by standard deviation [SD], 0.786; 95% confidence interval [CI], 0.659-0.936; P = 0.007). CAC volume also remained independently associated with an increased risk of MACE (HR per increase by SD, 2.608; 95% CI, 2.016-3.374; P < 0.001). CAC density derived from chest CT using automated AI quantification was independently and inversely associated with MACE, providing additional prognostic value when added to CAC volume.
This study evaluated the incidence and severity of the new onset of coronary artery calcification (CAC) related to clinical factors assessed at follow-up CAC scan among 6993 asymptomatic Korean adults (48.4 ± 7.2 years; 79.8% men) without underlying baseline CAC. During a median follow-up of 3.3 years, the risk of (a) the new onset of CAC (defined as coronary artery calcium score [CACS] > 0 at follow-up), (b) difference of ≥ 2.5 between baseline and follow-up square root (√) transformed CACS (Δ√transformed CACS), and (c) CACS > 100 at follow-up was evaluated. The incidence of the new onset of CAC, Δ√transformed CACS ≥ 2.5, and CACS > 100 was 19.3%, 13.0%, and 1.0%, respectively. After adjusting for traditional risk factors identified at initial CT scan and interscan period, the levels of systolic blood pressure, triglyceride glucose (TyG) index, hemoglobin A1C (HbA1C), and low-density lipoprotein cholesterol and current smoking were positively associated with the risk of the new onset of CAC and Δ√transformed CACS ≥ 2.5, respectively. TyG index, HbA1C, and current smoking were associated with the risk of CACS > 100 (all P < 0.05). An age-based subgroup analysis revealed that TyG index and current smoking were significantly associated with the new onset of CAC, Δ√transformed CACS ≥ 2.5, and CACS > 100 among participants aged < 55 years. In conclusion, both TyG index and smoking were independently associated with the new onset of CAC and its severity, especially in adults aged < 55 years without baseline CAC.
Clinical utility and dynamics of plasma biomarkers in early-onset dementia remain underexplored. To investigate plasma biomarker trajectories and their associations with clinical outcomes in early-onset Alzheimer disease (EOAD) and frontotemporal dementia (FTD). This multicenter, prospective cohort study analyzed participants in phase 1 of the Longitudinal Study of Early-onset Dementia and Family Members (LEAF), which was conducted from April 2021 through December 2023 in 34 centers across South Korea. Patients with β-amyloid-positive EOAD and FTD were included and underwent annual blood sampling and clinical assessment, within a follow-up period of approximately 2 years. Data were analyzed between June 2025 and March 2026. Levels of plasma phosphorylated tau 217 (p-tau217), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL) biomarkers were analyzed using assays. (1) Associations of baseline biomarkers with clinical outcomes (assessed with the Mini-Mental State Examination [MMSE] and the Clinical Dementia Rating-Sum of Boxes [CDR-SB] for the EOAD group, or the frontotemporal lobar degeneration [FTLD]-modified CDR-SB for the FTD group), (2) biomarker trajectories, and (3) association of biomarker level changes and clinical outcomes. A total of 322 participants with p-tau217, GFAP, and NfL analyses were stratified into the EOAD or FTD group based on their diagnosis. The EOAD group (n = 245) had a mean (SD) age of 61.8 (5.4) years and included 163 females (66.5%), while the FTD group (n = 77) had a mean (SD) age of 65.1 (7.3) years and included 45 females (62.3%). In the EOAD group, higher log2-transformed baseline p-tau217, GFAP, and NfL were each associated with faster decline in the MMSE score (association estimate [SE], -0.390 [0.127], P = .002; -0.775 [0.164], P < .001; and -0.679 [0.182], P < .001, respectively) and the CDR-SB score (estimate [SE], 0.401 [0.099], P < .001; 0.535 [0.126], P < .001; and 0.693 [0.122], P < .001, respectively). In the FTD group, GFAP and NfL were associated with MMSE decline (estimate [SE], -2.118 [0.566], P < .001 and -2.360 [0.428], P < .001, respectively), whereas p-tau217 was not (estimate [SE], 0.071 [0.418], P = .87). No biomarker was associated with FTLD-modified CDR-SB score change. Longitudinally, all mean (SD) biomarker levels increased in the EOAD group (p-tau217: 0.253 [0.077] pg/mL, P = .001; GFAP: 0.173 [0.040] pg/mL, P < .001; NfL: 0.149 [0.045] pg/mL, P = .001), whereas in the FTD group, only NfL level showed an upward pattern (0.251 [0.127] pg/mL, P = .05). Annualized biomarker changes were associated with worsening clinical outcomes in the EOAD group, but not in the FTD group. GFAP and NfL level increases were associated with MMSE score decline (estimate [SE], -0.005 [0.002], P = .007 and -0.010 [0.003], P = .001, respectively), while p-tau217 level increases were associated with CDR-SB score worsening (estimate [SE], 0.072 [0.024], P = .003) in the EOAD group. In this cohort study of patients with EOAD and FTD, baseline p-tau217, GFAP, and NfL were consistently associated with clinical outcomes in the EOAD group, whereas GFAP and NfL were associated with cognition only in the FTD group. These findings demonstrate distinct characteristics of plasma biomarkers in EOAD and FTD, supporting their potential utility for risk stratification.
BACKGROUND: Although the use of computed tomography (CT) imaging during pregnancy is increasing, the decision to use CT imaging remains challenging due to the conflicting considerations of the potential risks of fetal radiation and the medical benefits to the mother, with inconsistent evidence on offspring outcomes. METHODS: Nationwide retrospective cohort study based on the Korean National Health Insurance Service database, including 5,457,153 live births between 2005 and 2019. After 1:4 propensity score matching by maternal characteristics and diagnosis at the same gestational age, 13,307 CT-exposed and 42,946 unexposed pregnancies were included. CT exposure during pregnancy was identified using claims data from the last menstrual period to delivery. Estimated fetal radiation exposure was approximated based on anatomical scan location and categorized into lower and higher exposure groups, corresponding to non-abdominopelvic and abdominopelvic CT, respectively. Outcomes included major congenital malformations, congenital heart disease, and cerebral palsy identified using ICD-10 codes. Cox and logistic regression models were used, accounting for clustering by mother. RESULTS: Among CT-exposed pregnancies, 37.0% occurred within 14 days post-last menstrual period (LMP), and no increased risk of malformations or cerebral palsy was observed in this group. However, CT exposure during the first trimester was associated with a higher risk of major congenital malformations (Odds ratio (OR) 1.14, 95% confidence interval (CI) 1.02–1.30) and cerebral palsy (Hazard ratio (HR) 2.10, 95% CI 1.48–2.97). In a sensitivity analysis, non-abdominopelvic CT was associated with congenital heart disease (OR 1.23, 95% CI 1.03–1.47), but not with cerebral palsy, while abdominopelvic CT was associated with both congenital heart disease (OR 1.47, 95% CI 1.12–1.92) and cerebral palsy (HR 2.04, 95% CI 1.08–3.84). CONCLUSIONS: Even diagnostic-level CT radiation exposure during pregnancy may be associated with increased risks of congenital and neurodevelopmental disorders. Current radiation safety thresholds based solely on carcinogenic risk may underestimate fetal developmental vulnerability. However, this study has limitations, particularly the potential for residual confounding related to clinical acuity at presentation, which may not be fully addressed despite restriction to similar diagnostic indications and extensive propensity score matching.
Adherence to subsequent lung cancer screening (LCS) is essential but not well characterized in biennial national programs. To assess whether baseline LCS results and their management are associated with subsequent screening adherence and clinical outcomes. We analyzed participants in the Korean national LCS program who underwent baseline low-dose computed tomography (CT) between 2019 and 2021 and remained lung cancer-free for 2 years. Baseline results were classified as true negative or false positive (FP); FPs were grouped by management (invasive diagnostic procedures, chest CT surveillance, or no further evaluation). Adherence to the next biennial screening was modeled using multivariable logistic regression. Lung cancer incidence and all-cause mortality were assessed using Cox models with a 2-year landmark design. Among 235,753 participants, 54.4% returned for subsequent screening. Compared with true-negative results, adherence was lower among those with FP results who underwent invasive procedures (adjusted odds ratio [aOR], 0.64; 95% confidence interval [CI], 0.56-0.72) or CT surveillance (aOR, 0.77; 95% CI, 0.74-0.80) but was similar among those with FP results without further evaluation (aOR, 1.00; 95% CI, 0.95-1.04). In the landmark analysis, FPs were associated with higher risks of lung cancer incidence (adjusted hazard ratios [aHRs], 2.95-2.49 across management groups) and higher all-cause mortality among those who underwent invasive procedures (aHR, 1.57; 95% CI, 1.10-2.24) or CT surveillance (aHR, 1.23; 95% CI, 1.07-1.41). Downstream management after baseline FP findings, rather than FP status alone, was associated with subsequent LCS adherence, highlighting the need to support return screening in higher risk FP groups.
Breast imaging audits are essential for measuring performance, ensuring regulatory compliance, and improving patient outcomes. The American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) v2025 Manual refines established auditing principles and introduces several important additions to the Auditing and Outcomes Monitoring chapter. This review details fundamental auditing principles and describes the methodology and rationale for three key updates. First, routine auditing of BI-RADS category 3 assessments is now recommended to promote appropriate use of this category and ensure that it is applied correctly to findings with a ≤2% likelihood of malignancy. Second, a new framework has been introduced for auditing preoperative breast magnetic resonance imaging performed for staging, which incorporates both examination-level and finding-level metrics. Third, prospective assignment of the initial method of detection is encouraged to link cancer outcomes directly to screening and clinical presentation. Together, these updates expand breast imaging audits beyond traditional metrics and lay the foundation for future benchmark developments and global standardization.
This study aimed to develop a deep learning (DL) framework for automated detection and burden assessment of paramagnetic rim lesions (PRLs) by using quantitative susceptibility mapping (QSM) in multiple sclerosis (MS), compare QSM-only and QSM + fluid-attenuated inversion recovery (FLAIR) configurations, and evaluate the clinical relevance of the automated PRL burden. Brain magnetic resonance imaging data, including QSM and three-dimensional (3D) FLAIR data, obtained between January and December 2025, were retrospectively collected from 106 patients suspected of having MS. The dataset consisted of an exploration set (n = 84)-divided into training (n = 54) and internal test (n = 30) sets-and a temporal test set (n = 22). The PRLs were manually segmented to establish the ground truth. A 3D nnU-Net framework was trained using the QSM-only and QSM + FLAIR configurations for patient-level classification, lesion-level detection, and PRL burden assessment. A separate clinical implication cohort (n = 117) was used to assess the associations between the automated PRL burden and clinical outcomes. In the exploration set (median age: 40.0 years; 67.9% female), 69 (82.1%) patients were PRL-positive, with a total of 705 PRLs. In the internal test set, the QSM-only model showed higher lesion-level sensitivity (72.4% [147/203] vs. 62.1% [126/203], P = 0.004) and precision (78.2% [147/188] vs. 65.6% [126/192], P = 0.009). In the temporal test set, the lesion-level sensitivity was 45.3% (29/64) for the QSM-only model and 54.7% (35/64) for the QSM + FLAIR model (P = 0.181), whereas both models achieved 100% (8/8) patient-level sensitivity. No significant differences were observed in patient-level classification in either test set. A higher automated PRL burden was associated with poorer cognition (P = 0.002). QSM-based DL models enabled automated detection and burden assessment of PRLs in MS, with the QSM-only model performing comparably to QSM + FLAIR while offering a simplified single-sequence pipeline. The association between the automated PRL burden and cognitive impairment highlights its potential as a biomarker for MS assessment.
Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal malignancy. Imaging plays a pivotal role in the diagnosis and management of patients with PDAC, with treatment strategies largely guided by the tumor stage and resectability at initial diagnosis. The pancreatic CT protocol is the preferred first-line imaging modality, while MRI and PET/CT serve as problem-solving tools. In non-metastatic PDAC, anatomical resectability is mainly determined by the extent of tumor-vessel contact and is categorized as resectable, borderline resectable, or locally advanced disease. Although treatment decisions have traditionally relied on anatomical resectability, a growing body of evidence has highlighted the limitations of this approach. Consequently, an expanded concept of resectability that incorporates anatomical, biological, and host-related conditional factors has emerged. Potential biomarkers that reflect tumor biology include carbohydrate antigen 19-9 levels, imaging tumor phenotype, lymph node status, and metabolic activity on PET/CT. Neoadjuvant therapy (NAT) is widely used to treat borderline resectable or locally advanced disease. However, radiological restaging after NAT remains challenging and tends to overestimate vascular invasion, probably because imaging cannot reliably differentiate viable tumors from post-treatment fibrosis. In this context, biological factors may provide incremental value for treatment-response assessment. This review provides an updated overview of imaging techniques, assessments of anatomical resectability, challenges in radiological restaging after NAT, and biological and conditional factors that may enable more refined risk stratification in patients with PDAC.
To evaluate the performance of the left atrial time-to-peak (TTP) derived from time-resolved magnetic resonance angiography (TR-MRA) as a noninvasive indicator of left atrial pressure (LAP) in patients with atrial fibrillation (AF). This retrospective study included 92 patients who underwent cardiac TR-MRA and catheter-based LAP measurements prior to catheter ablation for AF between January 2021 and December 2022. TR-MRA-derived TTP was measured in the left atrium using contrast-enhanced time-signal intensity curves. Catheter-based LAP measured during sinus rhythm served as the reference standard, with high LAP defined as ≥15 mmHg. Univariable and multivariable linear regression analyses were performed, with LAP as the dependent variable and echocardiographic parameters as covariates. The diagnostic performance of TTP for high LAP was evaluated using receiver operating characteristic (ROC) curve analysis. Among the 92 patients (mean age ± standard deviation, 60 ± 10 years; 70 male), 70 (76.0%) had low LAP, and 22 (23.9%) had high LAP. Mean TTP was significantly longer in patients with high LAP than in those with low LAP (25.1 ± 6.9 s vs. 16.9 ± 5.7 s, P < 0.001). In multivariable linear regression analysis, TTP remained independently associated with LAP (β = 0.46, P < 0.001). ROC analysis demonstrated good discriminative performance of TTP for identifying high LAP, with an area under the curve of 0.834 (95% confidence interval, 0.742-0.904). Using a TTP cutoff of >19.6 seconds, the sensitivity and specificity for identifying elevated LAP levels were 81.8% (18/22) and 75.7% (53/70), respectively. TR-MRA-derived TTP showed a significant independent association with invasively measured LAP. It may serve as a reliable noninvasive imaging marker for identifying high LAP in patients with AF undergoing catheter ablation.
Spontaneous intracranial hypotension is a neurologic condition that is caused by a spinal cerebrospinal fluid (CSF) leak. The resulting CSF hypovolemia can manifest as a variety of clinical symptoms, with orthostatic headache being the most common. Although this disease has been recognized for decades, modern understanding of the types of causative spinal CSF leaks, diagnostic imaging tests to localize these leaks, and treatment options has evolved substantially in recent years. In this focused review article, we will provide an overview of the current diagnosis and treatment of spontaneous intracranial hypotension. We will emphasize recent improvements in understanding the pathophysiology of spinal leaks, developments in myelographic techniques to localize CSF leaks, and new treatment options for each type of leak.
暂无摘要（点击查看详情）
This corrects the article on p. 214 in vol. 27, PMID: 41735814.
This study aimed to evaluate the accuracy of predicting final adult height (FAH) in Korean girls with central precocious puberty (CPP) using artificial intelligence (AI)-derived bone age assessments integrated into the Bayley-Pinneau (BP) or Korean National Growth Chart (KGC) prediction models. This single-center, retrospective study included 122 Korean girls with CPP who received gonadotropin-releasing hormone agonist (GnRHa) treatment for at least two years between January 2000 and November 2022. We assessed bone age and predicted adult height at the initiation and completion of GnRHa treatment. We used three bone age assessment methods: human expert assessment based on the Greulich-Pyle (GP) atlas (Human-GP), AI-derived GP (AI-GP), and AI-weighted GP scoring (AI-GPw). We calculated predicted adult heights (PAHs) using both the BP and KGC models, generating 12 PAH estimates per patient (2 time points × 3 bone-age methods × 2 height-prediction models). We assessed prediction accuracy and agreement with FAH using linear regression analysis and Bland-Altman plots and performed multivariable analysis to identify significant predictors of FAH. Human-GP, AI-GP, and AI-GPw demonstrated comparable overall performance in predicting FAH (R²: 0.470-0.646 and 0.691-0.822 for treatment initiation and completion, respectively). AI-GPw combined with BP yielded slightly better point estimates but showed no statistically significant differences. At both time points, the BP model demonstrated consistently narrower 95% limits of agreement (LoA) than the KGC model. Multivariable analysis identified AI-GPw-BP and height percentile score as significant predictors of FAH at both time points; mid-parental height was significant only at treatment initiation. Human-GP, AI-GP, and AI-GPw demonstrated comparable accuracy in predicting FAH. The BP model demonstrated consistently narrower 95% LoA than did the KGC model. AI-GPw-BP was an independent predictor of FAH. These findings support the clinical utility of AI-derived bone age assessments for individualized FAH prediction in patients with CPP.
Advanced colorectal neoplasia (CRN) and coronary artery disease (CAD) are major health problems with common underlying pathogenic mechanisms. This study sought to investigate the association between the presence of advanced CRN and subclinical coronary atherosclerosis through a large cohort of asymptomatic Korean individuals who voluntarily underwent colonoscopy and coronary computed tomography angiography (CCTA) as routine health screening tests. A total of 6,044 Korean individuals aged ≥ 20 years who underwent a general health examination at the Health Promotion Center of Ulsan University Hospital between January 2009 and March 2020 were retrospectively analyzed. Advanced CRN was defined as the presence of invasive cancer or adenoma with a villous component, high-grade dysplasia, and/or a size of ≥ 1 cm. Subclinical coronary atherosclerosis was evaluated by CCTA. Participants with any CRN (n = 1,916, 31.7%) or advanced CRN (n = 240, 4.0%) had a higher coronary artery calcium score and a higher prevalence of any coronary, calcified, mixed, and non-calcified plaques, and obstructive CAD (≥ 50% diameter stenosis) on CCTA compared with participants with non-CRN (n = 4,128, 68.3%) (all P < 0.05). In the multivariable analysis to evaluate the association between CRN and CCTA findings, the advanced CRN group showed a statistically significant association with obstructive CAD (odds ratio, 1.65; 95% confidence interval, 1.09-2.50; P = 0.019). Asymptomatic subjects with CRN showed a higher prevalence of subclinical coronary atherosclerosis compared to those with non-CRN. Advanced CRN was independently associated with obstructive CAD on CCTA.
Focal cerebral arteriopathy-inflammatory type (FCA-i) is a leading cause of pediatric arterial ischemic stroke, but diagnostic challenges persist, particularly in East Asian populations where moyamoya disease (MMD) prevalence is high. The focal cerebral arteriopathy severity score quantifies arteriopathy severity but has not been validated in East Asian cohorts. We aimed to validate the focal cerebral arteriopathy severity score in Korean pediatric patients with FCA-i and compare temporal progression patterns with unilateral MMD. We conducted a retrospective cohort study of children with arterial ischemic stroke presenting to Seoul National University Hospital between January 2002 and December 2024. Patients were classified according to the Childhood Arterial Ischemic Stroke Standardized Classification and Diagnostic Evaluation criteria. The focal cerebral arteriopathy severity score was applied to serial magnetic resonance angiograms at baseline, peak severity, and final follow-up. Among 216 children with arterial ischemic stroke, 132 patients (61.1%) demonstrated arteriopathy, including 49 with FCA-i (median age, 8.6 [interquartile range (IQR), 6.4-11.3] years; 55% male), 60 with MMD (median age, 5.7 [IQR, 3.1-9.2] years; 43% male), and 13 with arterial dissection (median age, 7.0 [IQR, 3.1-10.4] years; 62% male). In FCA-i patients, the severity score correlated significantly with baseline infarct burden (ρ=0.42; P=0.0069) and exhibited characteristic monophasic evolution with early peak at 2 months followed by gradual recovery reaching lowest values at 11 months. Patients with unilateral MMD demonstrated consistently higher severity scores at all timepoints compared with FCA-i (baseline: 6.0 versus 2.0; final: 8.0 versus 3.0; P<0.001) without radiographic recovery. A baseline severity score ≥8.0 predicted contralateral progression in unilateral MMD with an area under the curve of 0.962 (sensitivity, 0.83; specificity, 0.91). The focal cerebral arteriopathy severity score demonstrates validity as a dynamic biomarker for monitoring FCA-i in Korean pediatric patients, exhibiting characteristic monophasic recovery patterns that distinguish it from progressive unilateral MMD.