CITATION.cff

cff-version: 1.2.0
message: >
  If you use the EHR Risk Framework software, please cite it using the
  metadata below (including the Zenodo DOI).
  Feedback: support@larucare.com
type: software
title: "EHR Risk Framework — Leakage-Aware, Calibrated, Explainable Open Software"
authors:
  - family-names: Hossain
    given-names: Md Rana
    email: support@larucare.com
    orcid: "https://orcid.org/0009-0005-5996-719X"
    affiliation: "Maharishi International University"
url: "https://ehr.larucare.com/"
repository-code: "https://github.com/ranasl62/ehr-chronic-disease-risk-prediction"
license: MIT
version: "1.0.0"
date-released: "2026-07-19"
abstract: >
  Open research software for leakage-aware longitudinal EHR-like binary risk
  prediction (chronic-disease and other horizon tasks). Provides index/horizon
  feature construction, isotonic calibration with Brier/ECE, SHAP and exploratory
  fairness jobs, FastAPI + Angular workbench, task YAML, training manifests, and
  reproducible synthetic verification artifacts. Research and education only —
  not a medical device; not for patient care.
keywords:
  - electronic health records
  - EHR risk prediction
  - clinical machine learning
  - chronic disease risk prediction
  - temporal leakage
  - leakage-safe AI
  - calibration
  - explainable AI
  - SHAP
  - fairness
  - reproducibility
  - research software
  - health informatics
identifiers:
  - type: doi
    value: "10.5281/zenodo.21448693"
# preferred-citation can point at a journal article DOI once the paper is published.