Skip to content

README.md

Latest commit

 

History

History
136 lines (90 loc) · 8.72 KB

README.md

File metadata and controls

136 lines (90 loc) · 8.72 KB

Epilepsy Prediction Tools

Clinical prediction calculators for epilepsy management, developed at UMC Utrecht in collaboration with international partners. The tools provide individualized probability estimates to support clinical decision-making in epilepsy onset, AED withdrawal, epilepsy surgery, and juvenile myoclonic epilepsy (JME).

Live site: https://wmotte.github.io/epilepsypredictiontools/

Calculators

Epilepsy Onset

Calculator Method Reference
First Consultation (FSC) Logistic regression van Diessen et al., Pediatrics 2018

Estimates the probability of an epilepsy diagnosis after one or more paroxysmal events at first consultation, based on 11 clinical and EEG parameters. Developed and validated at UMC Utrecht and the Martini Hospital Groningen.

AED Withdrawal

Calculator Method Reference
AED Withdrawal Risk Nomogram Lamberink et al., Lancet Neurology 2017

Predicts 2-year and 5-year seizure recurrence risk and long-term seizure freedom for patients with epilepsy in remission who are considering AED withdrawal. Based on an individual participant data meta-analysis of ten populations (children and adults). Does not apply to patients who became seizure-free through epilepsy surgery.

Epilepsy Surgery

Calculator Method Reference
TTS Postoperative Withdrawal Risk Nomogram (3 models) Lamberink et al., Epilepsia 2018
IQ ≥ 85 After Surgery Lookup table Cloppenborg et al., Neurology 2022
IQ ≥ 70 After Surgery Lookup table Cloppenborg et al., Neurology 2022
DQ ≥ 50 After Surgery Lookup table Cloppenborg et al., Neurology 2022

TTS: Predicts seizure recurrence and long-term outcome after AED withdrawal in children following epilepsy surgery (TimeToStop study, 766 children, 15 centres, 8 European countries). Three models: 2-year/5-year recurrence, long-term freedom, and cure probability.

IQ/DQ: Predicts post-operative cognitive outcome two years after pediatric epilepsy surgery, using pre-surgical IQ or DQ as the primary predictor. Collaborative study between Universitätsklinik OWL and UMC Utrecht.

Juvenile Myoclonic Epilepsy (JME)

Calculator Method Reference
JME AED Withdrawal Lookup table Stevelink et al., eClinicalMedicine 2022
JME Drug Resistance Lookup table Stevelink et al., eClinicalMedicine 2022

Both calculators are based on an individual participant data meta-analysis (Stevelink et al. 2022) identifying predictors of JME treatment outcomes across multiple international cohorts.

JME Withdrawal: Predicts 2-year and 5-year seizure recurrence probability after ASM withdrawal in JME patients who are currently seizure-free. Inputs: age at withdrawal (years), seizure-free interval (years), number of ASMs.

JME Drug Resistance: Predicts the probability of drug-resistant JME (defined as failure of two adequate ASM trials) from 9 clinical variables, covering 1024 possible combinations.

References

  1. van Diessen E, Lamberink HJ, Otte WM, Doornebal N, Brouwer OF, Jansen FE, Braun KPJ. A Prediction Model to Determine Childhood Epilepsy After 1 or More Paroxysmal Events. Pediatrics 2018; 142(6):e20180931. DOI: 10.1542/peds.2018-0931

  2. Lamberink HJ, Otte WM, Geerts AT, Pavlovic M, Ramos-Lizana J, Marson AG, Overweg J, Sauma L, Specchio LM, Tennison M, Cardoso TMO, Shinnar S, Schmidt D, Geleijns K, Braun KPJ. Individualised prediction model of seizure recurrence and long-term outcomes after withdrawal of antiepileptic drugs in seizure-free patients – a systematic review and individual participant data meta-analysis. Lancet Neurology 2017; 16(7):523–531. DOI: 10.1016/S1474-4422(17)30114-X

  3. Lamberink HJ, Boshuisen K, Otte WM, Geleijns K, Braun KPJ, on behalf of the TimeToStop Study Group. Individualized prediction of seizure relapse and outcomes following antiepileptic drug withdrawal after pediatric epilepsy surgery. Epilepsia 2018. DOI: 10.1111/epi.14020

  4. Cloppenborg T, van Schooneveld MMJ, Hagemann A, Hopf JL, Kalbhenn T, Otte WM, Polster T, Bien CG, Braun KPJ. Development and validation of prediction models for developmental and intellectual outcome following pediatric epilepsy surgery. Neurology 2022; 98:e225–e235. DOI: 10.1212/WNL.0000000000013065

  5. Stevelink R, Otte WM, et al. Individualised prediction of drug resistance and seizure recurrence after medication withdrawal in people with juvenile myoclonic epilepsy: a systematic review and individual participant data meta-analysis. eClinicalMedicine 2022; 53:101732. DOI: 10.1016/j.eclinm.2022.101732

Repository Structure

docs/                          # Static site (served via GitHub Pages)
├── index.html                 # Landing page with 8 calculator cards
├── aed-withdrawal.html        # AED Withdrawal Risk
├── tts-withdrawal.html        # TTS Postoperative Withdrawal Risk
├── first-consultation.html    # First Consultation (FSC)
├── iq85.html                  # IQ > 85 After Surgery
├── iq70.html                  # IQ > 70 After Surgery
├── dq50.html                  # DQ > 50 After Surgery
├── jme-withdrawal.html        # JME AED Withdrawal
├── jme-resistance.html        # JME Drug Resistance
├── css/
│   └── style.css              # UMC Utrecht house style
└── js/
    ├── ui.js                  # Shared UI logic (modal, result display)
    ├── aed-calc.js            # AED nomogram calculations
    ├── tts-calc.js            # TTS nomogram calculations (3 models)
    ├── fsc-calc.js            # FSC logistic regression
    ├── iq85-calc.js           # IQ > 85 lookup table (400 rows)
    ├── iq70-calc.js           # IQ > 70 lookup table (200 rows)
    ├── dq50-calc.js           # DQ > 50 lookup table (160 rows)
    ├── jme-withdrawal-calc.js # JME withdrawal lookup (20,400 rows, 2y+5y)
    └── jme-resistance-calc.js # JME drug resistance lookup (1024 rows)

EpilepsyPredictionTools/       # Original PHP/CouchCMS source (archived)
├── www/                       # Original website PHP source
├── IQ_php_calculator/         # IQ/DQ calculator PHP source and lookup CSVs
└── JME_calculator/            # JME calculator PHP source and lookup CSVs

Technical Implementation

The site is fully static — no server, database, or build step required. It runs directly from file:// or any static host.

  • No dependencies: Pure vanilla JavaScript (ES5), no frameworks or jQuery
  • Nomogram calculators (AED, TTS): Risk tables encoded as JavaScript arrays; scores are summed from per-variable point tables and used as direct array indices
  • Logistic regression (FSC): Coefficients applied directly in JS with Math.exp()
  • Lookup table calculators (IQ, DQ, JME): Source CSV data embedded in JS as either Map objects (keyed by variable combination) or compact flat integer arrays with a direct index formula, depending on table size
  • JME Withdrawal uses a flat array of 20,400 integer values (probabilities × 10,000) indexed by (asm−1)×5100 + sf×100 + (age−1), providing real individualized 2-year and 5-year recurrence probabilities from the Stevelink 2022 model

Deployment

The site is deployed to GitHub Pages from the docs/ folder:

  1. In the repository Settings → Pages, set source to Deploy from a branch, branch main, folder /docs
  2. The site will be available at https://wmotte.github.io/epilepsypredictiontools/

To run locally, open docs/index.html directly in a browser — no web server needed.

Disclaimer

These tools are intended for research and educational purposes only. They are not certified medical devices. Clinical decisions must always be made using professional judgment and in consideration of the individual patient's circumstances. The tools do not replace consultation with a qualified healthcare professional.

© UMC Utrecht