Deep learning for blood glucose prediction: Reproducibility challenges and factors affecting differential performance

Baiying Lu, Biratal Wagle, Zhaohui Liang, Yanjun Cui, Temiloluwa Prioleau, Kezhi Li, Amara Tariq, Kezhi Li, Amara Tariq, Amara Tariq · 2026

Read Deep learning for blood glucose prediction: Reproducibility challenges and factors affecting differential performance by Baiying Lu, Biratal Wagle, Zhaohui Liang, Yanjun Cui, Temiloluwa Prioleau, Kezhi Li, Amara Tariq, Kezhi Li, Amara Tariq, Amara Tariq (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13541153.txt.

Description

Baiying Lu and co-authors tested whether deep learning glucose forecasts can be reproduced in 2026. Reliable blood sugar prediction underpins artificial pancreas systems. The team first reviewed 67 recent papers and logged the reporting gaps that block repeat work.

They then rebuilt eight leading methods under one standard frame. Tests used over 1.36 million sensor readings from 128 people with type 1 diabetes across three public data sets. The models repeated well on technical and statistical checks. Their concepts, though, transferred poorly to data sets with distinct care patterns. People with weaker sugar control drew larger forecast errors, which flags a fairness and safety concern.

Catalog record

CREATORBaiying Lu, Biratal Wagle, Zhaohui Liang, Yanjun Cui, Temiloluwa Prioleau, Kezhi Li, Amara Tariq, Kezhi Li, Amara Tariq, Amara Tariq
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE78 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13541153

Browse the full library · About the Tower of Babel