Kao-Lang Liu

To evaluate whether personalized lung mass improves pre-treatment lung dose estimation accuracy compared to the 1000-g lung model, using post-treatment ⁹⁰Y PET/CT as the reference. Sex-specific anatomical differences and dosimetric factors associated with radiation pneumonitis (RP) were also explored. A retrospective analysis was conducted on 100 patients treated with ⁹⁰Y radioembolization from 2018 to 2025. Pre-treatment lung dose estimates were derived using both the standard 1000-g lung mass model and CT-based personalized lung mass. Post-treatment lung absorbed doses were quantified via ⁹⁰Y PET/CT. Dosimetric and clinical variables were compared between patients with and without RP, and sex-specific differences in lung anatomy and dose distribution were evaluated. The mean lung mass (760.11 ± 144.89 g) was significantly lower than the 1000-g model (P < 0.001). Two patients (2%) developed RP, and both had baseline lung metastases (28% overall). Although limited by the small number of events, the RP group showed a trend toward higher median administered activity (4425 MBq [range, 3000-5800 MBq] vs. 1300 MBq [range, 400-4900 MBq]) and lung absorbed dose, whereas no similar trend was observed for lung shunt fraction (LSF) (7.6% [range, 5.1-10.0%] vs. 4.3% [range, 1.0-14.5%]). Personalized lung dose showed a stronger correlation with ⁹⁰Y PET/CT than the 1000-g model (r = 0.678 vs. r = 0.612; P = 0.009), while LSF showed no correlation (r = 0.150). Sex-based dose differences seen with the 1000-g model were absent with personalized lung mass, which showed significantly lower lung mass in females (810.78 ± 126.21 g vs. 636.05 ± 109.33 g, P < 0.001). Personalized dosimetry improves predictive accuracy and mitigates sex-related dosimetric bias, suggesting CT-derived lung mass should be considered to optimize ⁹⁰Y radioembolization precision. Given our limited sample size, the association between pre-existing lung metastases and RP warrants further investigation.

Background: Pancreatic cancer (PC) is frequently missed on unenhanced CT examinations performed for unrelated clinical indications, where the pancreas is included incidentally and clinical suspicion is low. Purpose: To develop and validate a deep learning-based tool for PC diagnosis and risk stratification on unenhanced CT. Materials and Methods: This retrospective study included 3080 unenhanced CT studies of Taiwanese patients with PC, other pancreatic diseases and normal pancreas between 2004 and 2019 from a tertiary hospital, randomly divided into training, validation, and internal test sets. Unenhanced CT studies from United States institutions were used for external testing. A hybrid convolutional neural network–transformer model was trained for PC diagnosis and risk stratification. Performance was evaluated using sensitivity, specificity, and area under the curve (AUC), with comparisons to 2 radiologists by McNemar’s test and exploratory decision curve analysis. Results: The internal dataset included 713 PCs (mean age, 64.6 ± 12.0 years; 384 men), 1661 normal pancreas and 706 other pancreatic diseases. In an exploratory comparison restricted to unenhanced CT (29 PCs, 31 controls), the sensitivity of computer-aided diagnosis (CAD) tool (89.7%, 72.6-97.8) seemed comparable with that of 1 radiologist (86.2%, 68.3-96.1) and higher than another (41.4%, 23.5-61.1); but wide confidence intervals and inter-radiologist variability warrant cautious interpretation. In the internal test set (142 PCs, 474 controls), sensitivity was 90.8% (84.9-95.0) and specificity 93.0% (90.4-95.2) (AUC: 0.98), with sensitivity comparable to radiologist reports based on enhanced and unenhanced CT (95.4%, 90.2-98.3; P = .21). In the external set (42 PCs, 22 controls), sensitivity was 76.2% (60.5-87.9) and specificity 86.4% (65.1-97.1) (AUC: 0.89). The tool stratified cases into 7 risk levels with likelihood ratios ranging from <0.01 to 173.46. Exploratory decision curve analysis suggested potential net benefit across threshold probabilities. Conclusion: This tool may assist in the opportunistic detection and risk stratification of PC on unenhanced CT.

Background: Sarcopenia, characterized by degenerative skeletal muscle loss, is increasingly linked to poor surgical outcomes. Glutamine, an immune-modulating formula, may stimulate muscle protein synthesis and inhibit degradation. We used the psoas major muscle area (PMMA) at the third lumbar vertebra, normalized for height (PMMA index), as a skeletal muscle indicator. This study investigates whether perioperative glutamine supplementation mitigates psoas muscle atrophy. Methods: We enrolled gastric adenocarcinoma (GA) patients undergoing gastrectomy. Computed tomography assessed the psoas muscle short axis. Muscle atrophy was estimated by changes between preoperative and three-month post-gastrectomy scans. Perioperative glutamine supplementation (PGS) comprised five-day parenteral plus one-month oral use. Propensity score matching minimized potential bias. A linear regression model predicted the association. Results: Of 516 patients analyzed (2016–2019), 100 (19.4%) received PGS. After propensity score matching, each group contained 97 cases. The PGS group showed a significantly higher median PMMA index change than the non-PGS group (0.3 vs. −0.3 cm²/m², p = 0.004). Multivariate analysis revealed that PGS was significantly associated with increased PMMA index (coefficient = 0.60; 95% CI: 0.19–1.01; p = 0.005). Conclusions: PGS may help restore psoas muscle atrophy in GA patients undergoing gastrectomy. The underlying mechanisms likely relate to glutamine’s role in protein metabolism and immune function. Further studies are needed to elucidate these mechanisms fully.