pipette
ENEnglish

SSMCGM-Stream: A Streaming State-Space Model for Personalized Diabetes Management

Shakson Isaac, Chirag Patel

PreprintAfirmaciones fuertes, leer con cuidadoUso en el mundo real

En palabras de los autores

Continuous glucose monitoring supports diabetes management but requires forecasting models that are accurate, interpretable, and efficient enough for real-time deployment. We present SSMCGM-Stream, a streaming state-space model that maintains a personalized hidden state updated with each incoming glucose reading to provide constant-time forecasting, physiologically structured counterfactual reasoning, and interpretable forecasts through hidden-state attribution. Built on this foundation, SSMCGM-Agent delivers risk-aware, natural-language decision support by translating forecasts, explanations, and counterfactuals into grounded recommendations through a guarded language-model interface. Evaluated on the T1DEXI cohort, SSMCGM-Stream achieves forecasting performance comparable to strong neural baselines while substantially reducing computational cost, enabling continuous on-device diabetes decision support.

Resultado principalEl resumen no menciona limitaciones.

Apareció: jueves, 22 de octubre. arXiv. Preprint, todavía sin revisión por pares.

Comentario de los autores: Accepted to AI4Science and GenAI4Health NeurIPS 2026 workshops