# Predicting Blood Glucose with an LSTM and Bi-LSTM Based Deep Neural Network

arXiv: https://arxiv.org/abs/1809.03817 · HTML: https://mlsanity.com/paper/1809.03817/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Sep 2018.

## Built on by (most cited first, 9 of 9)

- [GlucoBench: Curated List of Continuous Glucose Monitoring Datasets with Prediction Benchmarks](https://mlsanity.com/paper/2410.05780/) (2024)
- [Diffusion Models for Time Series Forecasting: A Survey](https://mlsanity.com/paper/2507.14507/) (2025)
- [Stacked LSTM Based Deep Recurrent Neural Network with Kalman Smoothing for Blood Glucose Prediction](https://mlsanity.com/paper/2101.06850/) (2021)
- [Deep Personalized Glucose Level Forecasting Using Attention-based Recurrent Neural Networks](https://mlsanity.com/paper/2106.00884/) (2021)
- [Glycemic-Aware and Architecture-Agnostic Training Framework for Blood Glucose Forecasting in Type 1 Diabetes](https://mlsanity.com/paper/2502.14183/) (2025)
- [A Critical Review of the state-of-the-art on Deep Neural Networks for Blood Glucose Prediction in Patients with Diabetes](https://mlsanity.com/paper/2109.02178/) (2021)
- [MATURE-HEALTH: HEALTH Recommender System for MAndatory FeaTURE choices](https://mlsanity.com/paper/2304.09099/) (2023)
- [Learning Difference Equations with Structured Grammatical Evolution for Postprandial Glycaemia Prediction](https://mlsanity.com/paper/2307.01238/) (2023)
- [Evaluating Time-Series Foundation Models and Multimodal Dietary Context for CGM Forecasting](https://mlsanity.com/paper/2609.11872/) (2026)

## Abstract (arXiv)

A deep learning network was used to predict future blood glucose levels, as this can permit diabetes patients to take action before imminent hyperglycaemia and hypoglycaemia. A sequential model with one long-short-term memory (LSTM) layer, one bidirectional LSTM layer and several fully connected layers was used to predict blood glucose levels for different prediction horizons. The method was trained and tested on 26 datasets from 20 real patients. The proposed network outperforms the baseline methods in terms of all evaluation criteria.