Physical activity reduces the risk of colon cancer, but its prognostic impact after cancer diagnosis remains unclear. To evaluate the association between post-diagnosis activity and cause-specific mortality, we reconstructed a colorectal cancer patient cohort from the 2009–16 Korean National Health Insurance Service (NHIS) database. Subgroup analyses were performed by treatment group. In total, 27,143 colon cancer patients and 16,453 rectal cancer patients were included in the analysis (mean follow-up, 4.3 years; median 4.0 years). In the surgically treated group, a high level of activity (the weighted sum of the frequencies for walking, moderate, and vigorous activity greater than or equal to 3 times/week) was inversely associated with all-cause mortality (colon cancer: HR, 0.79; 95% CI, 0.72 to 0.88; rectal cancer: HR, 0.75; 95% CI, 0.66 to 0.86) and colorectal cancer-specific mortality (colon cancer: HR, 0.85; 95% CI, 0.76 to 0.97; rectal cancer: HR, 0.77; 95% CI, 0.66 to 0.90). No significant results were shown for cardiovascular disease-specific mortality. No association was shown in patients who received chemoradiotherapy without surgery. The present study may provide evidence for post-diagnosis physical activity as a prognostic factor in colorectal cancer, particularly in surgically treated early-stage patients.
Background/Objectives: To develop an automated deep learning-based bone age prediction model using the Tanner–Whitehouse (TW3) method and evaluate its feasibility by comparing its performance with that of pediatric radiologists. Methods: The hand and wrist radiographs of 560 Korean children and adolescents (280 female, 280 male, mean age 9.43 ± 2.92 years) were evaluated using the TW3-based model and three pediatric radiologists. Images with bony destruction, congenital anomalies, or non-diagnostic quality were excluded. A commercialized AI solution built upon the Rotated Single Shot MultiBox Detector (SSD) and EfficientNet-B0 was used. Bone age measurements from the model and radiologists were compared using the paired t-tests. Linear regression analysis was performed and the coefficient of determination (r²), mean absolute error (MAE), and root mean square error (RMSE) were measured. A Bland–Altman analysis was conducted and the proportion of bone age predictions within 0.6 years of the radiologists’ assessments was calculated. Results: The TW3-based model demonstrated no significant differences between bone age measurements and radiologists, except for participants 13 years old (overall, p = 0.874; 6–8 years, p = 0.737; 8–9 years, p = 0.093; 9–10 years, p = 0.301; 10–11 years, p = 0.584; 11–13 years, p = 0.976; 13 years, p < 0.001). There was a strong linear correlation between the model prediction and radiologist assessments (r2 = 0.977). The RMSE and MAE values of the model were 0.529 (95% CI, 0.482–0.575) and 0.388 (95% CI, 0.361–0.417) years. Overall, 82.3% of bone age model predictions were within 0.6 years of the radiologists’ interpretation. Conclusions: Automated deep learning-based bone age assessment has the potential to reduce radiologists’ workload and provide standardized measurements for clinical decision making.
Successful human–large language model collaboration is associated with model confidence and reader expertise in chest imaging, with rationale quality facilitating uptake of correct advice but increasing overreliance on incorrect suggestions.
Purpose: This study analyzed the malignancy rate of lymph nodes (LNs) according to the concordance of ultrasound (US) and computed tomography (CT) imaging features in patients with thyroid cancer.Methods: A retrospective analysis was performed of 277 cytopathologically confirmed cervical LNs (53.1% malignant) from 228 patients with thyroid cancer who underwent US-guided biopsy and preoperative US and CT. Malignancy rates were calculated according to the concordance of US and CT imaging features by matching cortical hyperechogenicity or abnormal vascularity on US (US-echo-vascularity) with strong or heterogeneous enhancement on CT (CT-enhancement), cystic changes on US (US-cystic) with cystic changes on CT (CT-cystic), and echogenic foci on US (US-EF) with calcifications on CT (CT-calcification). Results: Malignancy rates were significantly higher in positively concordant cases across all imaging features compared to negatively concordant cases (US-echo-vascularity/CT-enhancement: 93.5% vs. 8.2%, P<0.001; US-cystic/CT-cystic: 94.4% vs. 41.3%, P<0.001; and US-EF/CT-calcification: 84.4% vs. 34.5%, P<0.001). Discordance in US-cystic/CT-cystic (96.3% vs. 94.4%, P=0.735) or US-EF/CT-calcification (86.0% vs. 84.4%, P=0.873) findings showed malignancy rates comparable to those for positively concordant cases. However, discordance in US-echo-vascularity/CT-enhancement resulted in a significantly lower malignancy rate than in positively concordant cases (66.8% vs. 93.5%, P<0.001). In this discordant group, CT demonstrated a higher diagnostic odds ratio (3.896 vs. 0.257, P=0.002) and higher sensitivity (81.1% vs. 18.9%, P<0.001) than US.Conclusion: Positive US/CT concordance across all imaging features, as well as discordance in US-cystic/CT-cystic or US-EF/CT-calcification, strongly suggested LN metastasis. CT-enhancement was particularly reliable when discordant with US-echo-vascularity.