{"ok":true,"nota":{"id":"fd0f82244d9ec0dd","slug":"un-sistema-con-inteligencia-artificial-anticipa-mejor-la-glucosa-en-sa-fd0f82","titulo":"Un sistema con inteligencia artificial anticipa mejor la glucosa en sangre y corrige sus errores solo","resumen":"Investigadores combinaron un modelo de predicción de datos de glucosa con un agente de aprendizaje por refuerzo que aplica pequeños ajustes en tiempo real. El trabajo es un preprint sin revisión por pares y los resultados se evaluaron en simulación.","fuente":"medrxiv","fuenteNombre":"medRxiv · preprints","color":"#1f7a4d","autor":"Cao, X., Wang, Q.","url":"https://www.medrxiv.org/content/10.64898/2026.10.23.26366163v1?rss=1","imagen":"/hoy/portada/fd0f82244d9ec0dd","tema":"ciencia","subtemas":["diabetes","inteligencia artificial","monitoreo de glucosa","gemelos digitales","aprendizaje por refuerzo"],"entidades":["medRxiv"],"formato":"dato","entretenimiento":false,"pubTs":"2026-10-24T00:00:00.000Z","indexTs":"2026-10-24T16:44:51.482Z","hash":"d8307c24f25d514d58c3b8a8fb58dcd6d93ecd9796931aca3c1c799600501139","tsa":{"time":"Oct 24 16:45:14 2026 GMT","provider":"freetsa.org","sobre":"sello"},"stats":{"views":1,"likes":0,"clicks":0,"dwell":0},"tipo":"motor","registro":"hallazgo","porQueImporta":"Para quien convive con diabetes, un sistema que se adapta solo a cada cuerpo podría evitar subas y bajas peligrosas sin esperar a que alguien recalibre el modelo a mano. Hoy esas correcciones suelen llegar tarde. Si funciona fuera del laboratorio, el control diario de la glucosa podría dejar de depender de recordar cada ajuste. Pero todavía no se sabe si esta mejora se mantiene con pacientes reales y durante semanas.","quienFirma":"medRxiv (preprint; autores no identificados en el material recibido)","confianza":"preprint","consulta":{"pregunta":"¿Confiarías en una app que ajusta sola tu control de glucosa?"}},"sello":{"hash":"d8307c24f25d514d58c3b8a8fb58dcd6d93ecd9796931aca3c1c799600501139","sello":"21bbb54caf48a9df61ea11657fb9f41ed8f218d46ca2a6fa9014d8091e273b88","beacon":{"drand":{"cadena":"8990e7a9aaed2ffed73dbd7092123d6f289930540d7651336225dc172e51b2ce","ronda":6580975,"aleatorio":"7cce22f6f8f0e18a1e57d3c0d55b5004bf079b568a9a9f143dc9251a504825bd","tiempo":"2026-10-24T16:44:30.000Z"},"nist":{"cadena":2,"pulso":1997681,"salida":"C5248B3EE24B422962809C61551C0ED22A554FF83DF586AF6189AC5F408A5DD6B8269834835CE193FDEC10C180159B7695EE635C161F0BDB8944F4CE7E0AE962","tiempo":"2026-10-24T16:44:00.000Z"},"noAntes":"2026-10-24T16:44:30.000Z"},"verificacion":{"drand":true,"nist":true},"tsa":{"time":"Oct 24 16:45:14 2026 GMT","provider":"freetsa.org","sobre":"sello"},"original":{"titulo":"Digital Twin Workflow for Continuous Glucose Forecasting: Real-Time Data Assimilation Using Reinforcement Learning","descripcion":"Digital twins for physiological forecasting require robust mechanisms for continuously assimilating streaming data and adapting to patient-specific dynamics. Threshold-triggered model updates can restore predictive performance after substantial drift, but their reactive nature limits continuous adaptation. We propose a proactive digital twin framework that couples a pretrained PatchTST foundation model with an Actor-Critic reinforcement learning (RL) agent for residual correction. Using a 30-minute look-back window and recent residual history, the agent generates bounded corrections ({+/-} 0.05) to the first-step baseline forecast, with corrected predictions recursively propagated over a 15-","pubDate":"2026-10-24","url":"https://www.medrxiv.org/content/10.64898/2026.10.23.26366163v1?rss=1"},"indexado":"2026-10-24T16:44:51.482Z"}}