Prediction of polyneuropathy in recent-onset diabetes: A machine learning algorithm using blood-based protein biomarkers and standard demographic and clinical features
Bibliographic record
Abstract
Background Blood-based protein biomarkers may be an attractive diagnostic tool to detect diabetic sensorimotor polyneuropathy (DSPN) in routine clinical examinations. We aimed to compare a protein-based model with a model using traditional risk factors in detecting the presence of DSPN in individuals recently diagnosed with diabetes from the German Diabetes Study. Methods A total of 135 inflammatory and neuronal protein biomarkers were measured using proximity extension assay in blood samples of 66 and 357 individuals with and without DSPN, respectively, based on the Toronto Consensus Criteria. We constructed (i) a protein-based prediction model using lasso logistic regression, (ii) an optimized traditional risk model with age, sex, waist circumference, and diabetes type as demographic/clinical attributes selected a priori, and (iii) a model combining both. The AUC (95% CI) and robust bootstrapping assessed predictive performances. Results Lasso logistic regression selected the neurofilament light chain (NFL) and fibroblast growth factor 19 (FGF-19) as the most predictive protein biomarkers for detecting DSPN in individuals with recent-onset diabetes. The proteomics model achieved an AUC of 0.66 (0.59, 0.74), while the demographic/clinical model had an AUC of 0.68 (0.62, 0.76). However, combined features boosted the model performance to an AUC of 0.75 (0.68, 0.82). Conclusion A model combining two objective blood-based biomarkers (NFL and FGF-19) and four standard demographic and clinical parameters (age, sex, waist circumference and diabetes type) has an acceptable performance for detecting early DSPN. This model could complement clinical and neurophysiological testing. Publication History Article published online: 02 May 2023 © 2023. Thieme. All rights reserved. Georg Thieme Verlag Rüdigerstraße 14, 70469 Stuttgart, Germany
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".