OSCR

Machine-learning prediction and risk stratification of 12-month cognitive decline in Alzheimer's disease using routine clinical and MRI data.

Code ↔ Paper

18 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 18 matches
  1. [1] § Results › Model performance in cross-validation ↔ 10_make_supp_tables_S7_S8_S9.py, lines 39–123 · score 0.84 · unnecessary interventions, clinically interpretable, probability threshold, confusion matrix, NPV, PPV
  2. [2] § Methods › Internal validation and performance metrics ↔ 01_rebuild_models_and_outputs_legacy.py, lines 64–166 · score 0.81 · isotonic regression, ROC curve, Brier score, event rate, logit, fitting
  3. [3] § Methods › Data preprocessing ↔ config.py, lines 1–21 · score 0.81 · hot encoded, single imputation, analytic sample, candidate predictor, preprocessing, reproducible
  4. [4] § Methods › Data preprocessing ↔ 01_archive_legacy_model.py, lines 1–34 · score 0.80 · hot encoded, modelling pipeline, single imputation, analytic sample, preprocessing, median
  5. [5] § Methods › Model specification and comparators ↔ 10_make_supp_tables_S7_S8_S9.py, lines 39–123 · score 0.79 · max_depth, max_features, n_estimators, random forest, S9, sqrt
  6. [6] § Methods › Model specification and comparators ↔ config.py, lines 1–21 · score 0.79 · max_depth, max_features, n_estimators, candidate predictors, sqrt, hyperparameters
  7. [7] § Methods › Internal validation and performance metrics ↔ 08_sensitivity_ventricles_icv.py, lines 114–257 · score 0.74 · isotonic regression, Brier score, calibration slope, fitting, trained, sensitivity
  8. [8] § Methods › Model specification and comparators ↔ 08_sensitivity_ventricles_icv.py, lines 1–33 · score 0.72 · replacing absolute ventricular, ICV ratio, ventricular volume, S5, Sensitivity
  9. [9] § Methods › Risk stratification and clinical use case ↔ 01_rebuild_models_and_outputs_legacy.py, lines 64–166 · score 0.69 · 0.25–0.5, random forest, MMSE change, stratum, intermediate, Median
  10. [10] § Results › Model performance in cross-validation ↔ 08_sensitivity_ventricles_icv.py, lines 1–33 · score 0.68 · ventricular volume, intracranial volume, fold stratified, absolute, S5, ICV
  11. [11] § Results › Participants and baseline characteristics ↔ 06_make_supp_tables_S2_S4.py, lines 39–147 · score 0.65 · MRI derived, baseline categorical, analytic sample, S2, demographic, age
  12. [12] § Methods › Candidate predictors ↔ config.py, lines 66–85 · score 0.61 · middle temporal, entorhinal, fusiform, intracranial, hippocampal, ventricular
  13. [13] § Methods › Candidate predictors ↔ 06_make_supp_tables_S2_S4.py, lines 39–147 · score 0.59 · APOE4 allele, genetics, Demographics, sex, age, ADNIMERGE
  14. [14] § Methods › Software ↔ 01_archive_legacy_model.py, lines 282–349 · score 0.58 · full analytic, random forest, SHAP, curve, calibration, model
  15. [15] § Methods › Candidate predictors ↔ 11_make_table1_and_suppS2_from_raw_ADNIMERGE.py, lines 48–141 · score 0.58 · CDR SB, ADAS Cog, Baseline, MMSE
  16. [16] § Results › Model performance in cross-validation ↔ 01_archive_legacy_model.py, lines 104–120 · score 0.53 · confusion matrix, NPV, PPV, Youden, FN, TN
  17. [17] § Methods › Candidate predictors ↔ config.py, lines 39–64 · score 0.51 · genetics, education, allele, carrier, Demographics, APOE4
  18. [18] § Results › Participants and baseline characteristics ↔ 05_export_submission_tables.py, lines 1–22 · score 0.51 · baseline categorical characteristics, analytic sample, S2, ADNI

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 114 lines · 4.4 KB · MIT · 4 matches

  1. #!/usr/bin/env python3
  2. # -*- coding: utf-8 -*-
  3. """
  4. Central configuration for the Scientific Reports reproducibility package.
  6. All "rules" are taken from the submitted manuscript:
  7. - Outcome: 12-month MMSE decline >=3 points
  8. - Analytic sample: baseline AD with complete 12-month MMSE (n=306 in the paper)
  9. - Candidate predictors: demographics/genetics + cognition/severity + MRI volumes
  10. - Internal validation: 5-fold stratified CV, within-fold preprocessing, median/mode single imputation,
  11. standardization, one-hot encoding, isotonic calibration within fold.
  12. - RF tuned hyperparameters: n_estimators=200, max_features='sqrt', max_depth=None
  13. - Threshold metrics: Table 3 (0.25 / 0.40 / 0.50) + Supp Table S3 (0.20/0.30/0.40/0.50/0.60)
  14. - Risk strata: low <0.25, intermediate 0.25–0.50, high >=0.50
  15. - DCA thresholds: 0.20–0.80
  16. """
  17. from __future__ import annotations
  18. from dataclasses import dataclass
  19. from typing import Dict, List, Sequence, Tuple
  21. RANDOM_SEED: int = 20260221
  23. # Fixed thresholds and risk strata (manuscript)
  24. TABLE3_THRESHOLDS = [0.25, 0.40, 0.50]
  25. SUPP_S3_THRESHOLDS = [0.20, 0.30, 0.40, 0.50, 0.60]
  26. DCA_THRESHOLD_GRID = [round(x, 2) for x in [i/100 for i in range(20, 81)]]
  27. RISK_STRATA = [
  28. ("low", 0.0, 0.25),
  29. ("intermediate", 0.25, 0.50),
  30. ("high", 0.50, 1.0000001),
  31. ]
  33. # Columns expected in the analysis-ready CSV created by 00_build_analysis_ready_from_ADNIMERGE.py.
  34. # The script lowercases all column names; we keep the lowercase names here.
  35. RID_COL = "rid"
  36. Y_COL = "decline_3pt"
  37. MMSE_CHANGE_COL = "mmse_change"
  39. # Candidate predictor columns (lowercase), per manuscript.
  40. # Note: ADNIMERGE naming can vary across releases; we therefore allow synonyms via COLUMN_SYNONYMS.
  41. CANDIDATE_PREDICTORS: List[str] = [
  42. # Demographics/genetics
  43. "age",
  44. "ptgender", # sex
  45. "pteducat", # years education
  46. "apoe4", # allele count -> derive carrier
  47. # Baseline cognition / severity
  48. "mmse_bl",
  49. "adas13_bl",
  50. "adas11_bl",
  51. "adasq4_bl",
  52. "cdrsb_bl",
  53. "faq_bl",
  54. # MRI volumes (baseline)
  55. "hippocampus_bl",
  56. "ventricles_bl",
  57. "wholebrain_bl",
  58. "entorhinal_bl",
  59. "fusiform_bl",
  60. "midtemp_bl",
  61. "icv_bl",
  62. # Derived ratio used in Supplementary Table S4; computed in preprocessing
  63. # "hippo_icv",
  64. ]
  66. # Synonyms mapping: when a preferred column is missing, use the first existing alias.
  67. COLUMN_SYNONYMS: Dict[str, Sequence[str]] = {
  68. "ptgender": ("ptgender", "gender", "sex"),
  69. "pteducat": ("pteducat", "educ", "education", "edu", "yrs_edu", "yearseducation"),
  70. "apoe4": ("apoe4", "apoe4allele", "apoe4_count", "apoe_e4"),
  71. "mmse_bl": ("mmse_bl", "mmse", "mmscore", "mmse_total"),
  72. "adas13_bl": ("adas13_bl", "adas13", "adas_cog13", "adascog13"),
  73. "adas11_bl": ("adas11_bl", "adas11", "adas_cog11", "adascog11"),
  74. "adasq4_bl": ("adasq4_bl", "adasq4", "adas_q4"),
  75. "cdrsb_bl": ("cdrsb_bl", "cdrsb", "cdrsb_total"),
  76. "faq_bl": ("faq_bl", "faq"),
  77. # MRI: ADNI sometimes uses volumes with suffix _ucsf, _ba, etc; keep a few common aliases.
  78. "hippocampus_bl": ("hippocampus_bl", "hippocampus", "hippocampus_ucsf", "hippocampus_u"),
  79. "ventricles_bl": ("ventricles_bl", "ventricles", "ventricles_ucsf", "ventricles_u"),
  80. "wholebrain_bl": ("wholebrain_bl", "wholebrain", "wholebrain_ucsf", "wholebrain_u"),
  81. "entorhinal_bl": ("entorhinal_bl", "entorhinal", "entorhinal_ucsf", "entorhinal_u"),
  82. "fusiform_bl": ("fusiform_bl", "fusiform", "fusiform_ucsf", "fusiform_u"),
  83. "midtemp_bl": ("midtemp_bl", "midtemp", "middletemporal", "middle_temporal", "midtemp_ucsf", "midtemp_u"),
  84. "icv_bl": ("icv_bl", "icv", "intracranialvolume", "intracranial_vol", "icv_ucsf", "icv_u"),
  85. }
  87. # Model hyperparameters (as reported)
  88. RF_PARAMS = dict(
  89. n_estimators=200,
  90. max_features="sqrt",
  91. max_depth=None,
  92. random_state=RANDOM_SEED,
  93. n_jobs=-1,
  94. )
  96. LOGREG_PARAMS = dict(
  97. solver="liblinear",
  98. max_iter=1000,
  99. random_state=RANDOM_SEED,
  100. )
  102. def resolve_column(df_columns: Sequence[str], preferred: str) -> str | None:
  103. """Return the actual column name present in df_columns matching preferred or its synonyms."""
  104. cols = set(df_columns)
  105. for c in COLUMN_SYNONYMS.get(preferred, (preferred,)):
  106. if c in cols:
  107. return c
  108. return None
  110. def apoe4_carrier_from_count(apoe4_series):
  111. """Return 1 if >=1 allele, 0 if 0, NaN if missing."""
  112. import numpy as np
  113. x = apoe4_series.astype("float")
  114. return np.where(np.isnan(x), np.nan, (x >= 1).astype(int))

config.py at commit b71d69f, under MIT · at the source

Overview

Authors: Yinghui Geng1, Huijun Zhang2
  1. School of Nursing, Jinzhou Medical University, Jinzhou, Liaoning China
  2. Jinzhou Medical University, Jinzhou, Liaoning China
Institutions: Jinzhou Medical University (China)
Journal: Scientific reports, volume 16, issue 1, article 12227
Dates: received 7 November 2025; accepted 3 March 2026; published online 12 March 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1038/s41598-026-43321-1 · PMID 41820490 · PMCID PMC13076664 · OpenAlex W7135089329
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: structural MRI / diffusion (modality), human (organism), Alzheimer's / dementia (population), clinical / translational (subfield)
Methods: Machine learning, Statistics
Keywords: Alzheimer’s disease, MMSE, machine learning, random forest, structural MRI, risk stratification, Diseases, Medical research, Neurology, Neuroscience
MeSH: Alzheimer Disease*, Cognitive Dysfunction*, Machine Learning*, Magnetic Resonance Imaging*, Aged, Aged, 80 and over, Classification Algorithms, Female, Humans, Male, Mental Status and Dementia Tests, Prediction Algorithms, Predictive Learning Models, Prognosis, Random Forest, Risk Assessment, ROC Curve (* major topic)
Topic: Dementia and Cognitive Impairment Research (Psychiatry and Mental health, Medicine), according to OpenAlex
Funding: U.S. Department of Defense (U01 AG024904, W81XWH, W81XWH-12-2, W81XWH1220012, AG024904); Alzheimer's Disease Neuroimaging Initiative (W81XWH-12-2–0012, U01AG024904, AG024904, W81XWH-12-2); National Institutes of Health (W81XWH, W81XWH-12-2-0012, U01-AG024904, AG024904)
Citations: cited by 1 paper (Europe PMC); 27 references in the paper

Abstract

The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.

Repository

Its files are read in the Code ↔ Paper reader above, with 18 matches between paragraphs and lines of code.

gengyinghui/ADNI-12M-Cognitive-Decline-ML

License: MIT
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Commit: b71d69f103d372358403ced982ab62ba51fe5492, 22 February 2026
Languages: Python (15), R (1)
Size: 20 files, 16 scripts
Software Heritage: not archived
Found in: “Code availability”
Holds: README, license file, environment (requirements.txt)
Not found: CITATION.cff, tests, continuous integration, documentation
Tools: NumPy (13 files), pandas (13 files), scikit-learn (7 files), SciPy (4 files), Matplotlib (2 files), ggplot2 (1 file), SHAP (1 file), tidyverse (1 file)
Availability: 1 check, the latest on 30 September 2026: the link answers
  • 30 September 2026: the link answers
18 files
At the source: github.com/gengyinghui/ADNI-12M-Cognitive-Decline-ML

Code availability statement

The paper has a code availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s41598-026-43321-1.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 16 scripts, each with its path and the digest of its content;
  • 18 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Code and data availability statement

The paper has a code and data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s41598-026-43321-1.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 10 keywords, 17 MeSH terms, 3 funders, 26 references.

Cite

This paper

Geng, Y., & Zhang, H. (2026). Machine-learning prediction and risk stratification of 12-month cognitive decline in Alzheimer's disease using routine clinical and MRI data. Scientific reports, 16(1), 12227. https://doi.org/10.1038/s41598-026-43321-1

BibTeX

@article{geng2026machine,
author = {Geng, Yinghui and Zhang, Huijun},
title = {{Machine-learning prediction and risk stratification of 12-month cognitive decline in Alzheimer's disease using routine clinical and MRI data}},
journal = {Scientific reports},
year = {2026},
month = mar,
volume = {16},
number = {1},
pages = {12227},
publisher = {Nature Publishing Group},
issn = {2045-2322},
doi = {10.1038/s41598-026-43321-1},
url = {https://doi.org/10.1038/s41598-026-43321-1},
pmid = {41820490},
pmcid = {PMC13076664}
}

RIS

TY - JOUR
AU - Geng, Yinghui
AU - Zhang, Huijun
TI - Machine-learning prediction and risk stratification of 12-month cognitive decline in Alzheimer's disease using routine clinical and MRI data
T2 - Scientific reports
J2 - Sci Rep
PY - 2026
DA - 2026/03/12
VL - 16
IS - 1
SP - 12227
SN - 2045-2322
PB - Nature Publishing Group
DO - 10.1038/s41598-026-43321-1
UR - https://doi.org/10.1038/s41598-026-43321-1
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41598-026-43321-1",
"type": "article-journal",
"title": "Machine-learning prediction and risk stratification of 12-month cognitive decline in Alzheimer's disease using routine clinical and MRI data",
"container-title": "Scientific reports",
"author": [
{
"family": "Geng",
"given": "Yinghui"
},
{
"family": "Zhang",
"given": "Huijun"
}
],
"container-title-short": "Sci Rep",
"volume": "16",
"issue": "1",
"page": "12227",
"DOI": "10.1038/s41598-026-43321-1",
"PMID": "41820490",
"PMCID": "PMC13076664",
"ISSN": "2045-2322",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41598-026-43321-1",
"language": "en",
"issued": {
"date-parts": [
[
2026,
3,
12
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1038/s41514-026-00390-w [code]
Mild cognitive impairment cases affect the predictive power of Alzheimer's disease diagnostic models using routine clinical variables.
Journal: npj aging
In common: SHAP, ggplot2, tidyverse, 5 other tools, Alzheimer's / dementia, clinical / translational
[2] doi:10.1073/pnas.2516601123 [code]
Unveiling the glymphatic system's role in brain aging: A comprehensive biomarker and modifiable intervention target.
Journal: Proceedings of the National Academy of Sciences of the United States of America
In common: SHAP, ggplot2, tidyverse, 5 other tools, structural MRI / diffusion, clinical / translational
[3] doi:10.1038/s41467-026-76837-1 [code]
Drug screen and machine learning predict neuroprotective agents in a preclinical human model of childhood dementia.
Journal: Nature communications
In common: SHAP, ggplot2, tidyverse, 5 other tools, Alzheimer's / dementia
[4] doi:10.1016/j.nicl.2026.104012 [code]
Structural-functional multilayer brain network properties and outcome of combined repetitive transcranial magnetic stimulation and psychotherapy for obsessive-compulsive disorder.
Journal: NeuroImage. Clinical
In common: SHAP, ggplot2, tidyverse, 5 other tools, structural MRI / diffusion
[5] doi:10.1016/j.isci.2026.115329 [code]
Brain metastases converge on shared geometric architecture and transcriptomic landscape yet remain distinct from gliomas.
Journal: iScience
In common: SHAP, ggplot2, tidyverse, 5 other tools, clinical / translational
[6] doi:10.1002/dad2.70430 [code]
Benchmarking privacy and utility in synthetic tabular cohorts for Alzheimer's disease research.
Journal: Alzheimer's & dementia (Amsterdam, Netherlands)
In common: ggplot2, tidyverse, scikit-learn, 4 other tools, Alzheimer's / dementia, 1 reference
[7] doi:10.1038/s42003-026-10957-8 [code]
Brain defence by the extracellular matrix protein Cochlin.
Journal: Communications biology
In common: SHAP, ggplot2, tidyverse, 5 other tools
[8] doi:10.1093/nargab/lqag050 [code]
TSProm: deep learning framework to predict tissue-specific regulatory logic.
Journal: NAR genomics and bioinformatics
In common: SHAP, ggplot2, tidyverse, 5 other tools
[9] doi:10.1002/advs.202600020 [code]
Early Retinal UCHL1 Dysregulation Coupled With Synaptic Loss Reflects Alzheimer's Disease Severity.
Journal: Advanced science (Weinheim, Baden-Wurttemberg, Germany)
In common: SHAP, scikit-learn, pandas, 3 other tools, Alzheimer's / dementia, 1 reference
[10] doi:10.1038/s44318-026-00818-9 [code]
FAM134B-mediated ER-phagy degrades APP and suppresses Alzheimer's disease pathology.
Journal: The EMBO journal
In common: ggplot2, tidyverse, scikit-learn, 4 other tools, Alzheimer's / dementia, 1 reference

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.