Search

How does artificial intelligence monitor the eyesight of people with diabetes? And are surgeries more effective than modern obesity drugs? | Radiožurnál

Translated title. Original: Jak umělá inteligence hlídá zrak diabetiků? A jsou operace účinnější než moderní léky na obezitu? | Radiožurnál

The report describes how connected insulin pumps and AI-assisted retinal screening support diabetes care. In Czechia, the AI screening is covered by public health insurance and helps identify patients who need an eye specialist.

Diabetes: a simple photo of the eye could reveal what is damaging the kidneys - The Women’s Health

Translated title. Original: Diabete: una semplice foto dell'occhio potrebbe rivelare cosa sta danneggiando i reni - The Wom Healthy

A study suggests a retinal photograph analyzed with artificial intelligence can help distinguish diabetic kidney disease from other kidney conditions and may improve future risk predictions, though more diverse, larger studies are needed before clinical use.

Diet for diabetes: what to eat to curb glucose spikes

Translated title. Original: Dieta para la diabetes: qué comer para frenar los picos de glucosa

The document outlines dietary strategies for diabetes focusing on nutrient-dense, fiber-rich, low-glycemic foods; emphasizes how digestion rate affects glucose, suggests specific foods (eggplant, berries, avocado, legumes, barley), and recommends simple cooking tweaks (vinegar, spices) and yogurt for metabolic benefits.

Machine Learning Predicts Diabetes Risk in NAFLD - EMJ

A study builds a five-factor machine learning model to predict medium-term diabetes risk in NAFLD patients, validated externally with good discrimination but limited by sample scope; authors call for further multicenter validation before clinical use.

Samsung Health to Offer Free AI Diabetes Prevention Coaching for Seniors

Samsung Health will offer free AI diabetes prevention coaching for seniors in the US through a partnership with Lark Health, starting with the Medicare Diabetes Prevention Program and expanding to additional programs under the CMS ACCESS model; eligibility requires Medicare Part B and no active diabetes, with AI coaching supplementing but not replacing medical care.

Samsung Health now offers US users AI-backed diabetes prevention and support - SamMobile

Samsung Electronics America and Lark Health partner to bring AI-assisted prevention and chronic-care support for adults 65 and older within Samsung Health. A phased rollout includes CMS ACCESS Model-enabled programs for cardiometabolic and behavioral-health conditions, with eligibility-based options for broader access.

GitHub - Kolawole-Toheeb/diabetes-readmission-prediction · GitHub

This repository trains a machine-learning model to estimate 30-day hospital readmission risk for patients with diabetes and serves predictions through a FastAPI web service. Its best model, XGBoost, achieved a test AUC of 0.653.

ECHO implant: ETH Zurich regulates blood sugar in mice electrically

Translated title. Original: ECHO-Implantat: ETH Zürich reguliert den Blutzucker bei Mäusen elektrisch

ETH Zurich reports an implant called ECHO that electrically converts excess glucose to gluconate using a graphene-based catalyst, forming a closed-loop system with a CGM and smartphone interface. Preclinical mouse studies show rapid glucose reduction; significant questions remain about safety, human applicability, and device miniaturization before clinical use.

Artificial intelligence links LDL cholesterol to the risk of experiencing arrhythmias in diabetics - Libertad Digital

Translated title. Original: La inteligencia artificial vincula el colesterol LDL al riesgo de sufrir arritmias en diabéticos - Libertad Digital

An international study involving the UPM and the University of Ioannina shows LDL cholesterol as a key predictor of arrhythmias in type 2 diabetes patients, using six months of home-based biometric data and ML models to identify significant risk features and support personalized treatment strategies.

[2611.03601] DiabAgent: Toward Agentic Physiological Control

The paper introduces DiabAgent, a suite for agentic physiological control in diabetes, including APC for glucose management, DiabetSim for virtual patient testing, and SweetBench for benchmarking. Results show APC with Claude Opus 5.5 outperforms OpenAPS on COGI, with ablations highlighting the value of adaptation and context; a hybrid control approach balances performance and cost. The work offers a framework, simulator, and evaluation foundation for agent participation in physiological control.