RESUMEN
This study addresses challenges related to privacy issues in utilizing medical data, particularly the protection of personal information. To overcome this obstacle, the research focuses on data synthesis using real-world time-series generative adversarial networks (RTSGAN). A total of 53,005 data were synthesized using the dataset of 15,799 patients with colorectal cancer. The results of the quantitative evaluation of the synthetic data's quality are as follows: the Hellinger distance ranged from 0 to 0.25; the train on synthetic, test on real (TSTR) and train on real, test on synthetic (TRTS) results showed an average area under the curve of 0.99 and 0.98; a propensity mean squared error was 0.223. The synthetic and real data were similar in the qualitative methods including t-SNE and histogram analyses. The application of synthetic data in predicting five-year survival in colorectal cancer patients demonstrates comparable performance to models based on real data. This study employs distance to closest records and membership inference test to assess potential privacy exposure, revealing minimal risk. This study demonstrated that it is feasible to synthesize medical data, including time-series data, using the RTSGAN, and the synthetic data can be evaluated to accurately reflect the characteristics of real data through quantitative and qualitative methods as well as by utilizing real-world artificial intelligence models.
Asunto(s)
Neoplasias Colorrectales , Humanos , Redes Neurales de la ComputaciónRESUMEN
Objective: Cancer cachexia is progressive weight loss due to muscle/adipose tissue wasting and inadequate intake that occurs in response to malignancy. It is an independent predictor of disease recurrence and reduced survival in several cancers. However, cachexia's relationship with gynecologic malignancy outcomes has only been examined in small studies with limited follow-up and inconsistent definitions of cachexia. This study investigated the impact of cachexia on disease recurrence and overall survival in high-risk endometrial carcinoma patients. Methods: This retrospective cohort study examined data from patients with high-risk non-metastatic primary endometrial carcinoma treated at a single institution from 2015 to 2020. Treatment for all subjects included total hysterectomy, surgical staging, pelvic external beam radiotherapy with or without adjuvant chemotherapy. Radiation planning CT datasets were used to measure skeletal musculature at the L3 vertebral level. Skeletal muscle index (SMI) was defined as total L3 skeletal muscle cross sectional area (cm2)/height2 (m2), and cachexia was defined based on SMI. Results: 55 patients were eligible for analysis. Several SMI thresholds were used to define cachexia, and analysis was performed for each definition. Kaplan-Meier and Cox-proportional hazards regression analysis yielded no significant reduction in overall survival (OS) or progression-free survival (PFS) in patients with cachexia, regardless of threshold chosen. However, 4 of 13 definitions of cachexia showed significantly improved OS in patients without cachexia, relative to those with cachexia. There were no significant differences in disease recurrence. Conclusions: Cachexia as defined in this study was not associated with poor outcomes in endometrial carcinoma patients based on OS, PFS, or disease recurrence.