You are here:

Outcomes and AI theme

Outcomes and Artificial Intelligence (AI)

Machine learning model can be applied across a number of different datasets to help recognise the patterns of epilepsy and predict effective anti-seizure medication.

Outcomes and AI theme group (L - R) Front row - Dr Richard Chang, Prof Patrick Kwan, Miss Mandara Harikar, 2^nd row – Dr Shobi Sivathamboo, Miss Janet Choi, Miss Noushin Chiniforoush, 3^rd row – Mr Zachary Daus, Dr Hugh Simpson, Dr Zhibin Chen, Backrow – Dr Haris Hakeem, Dr Duong Nhu, Mr Daniel Thom (Absent – Dr Alison Anderson, Dr Emma Foster, Dr Deval Mehta, Dr Mohamad Nazem-Zadeh, Mr Talha Ilyas)

Our work

Project Leaders - Prof Patrick Kwan and Dr Zhibin Chen
The current standard of care for epilepsy is to commence treatment with a single antiseizure medication (ASM) and if it fails, try successive drug regimens until the epilepsy is regarded as ‘pharmacoresistant’. This trial-and-error approach does not consider factors beyond broad seizure types or epilepsy syndromes in drug selection. Further, it risks damage to the brain from uncontrolled seizures while patients endure trials of ineffective treatments, and 30% patients remain uncontrolled.
There is no reliable way to predict which drug will work best for a given patient, nor to predict whether their epilepsy is likely drug-resistant. As a result, non-drug options, like surgery, are generally kept as a last resort, by which time they might be less effective, and irreversible damage might have occurred. We hypothesise that it is possible to predict response to ASM treatment. Also, it is possible to identify individuals with high risk of pharmaco-resistant epilepsy using machine learning techniques.
We have recently developed a world-first machine learning model to predict the most effective initial ASM using routinely collected clinical data from patients in two countries (UK and Australia). In this project we will improve the performance of our model by integrating it with diagnostic reports and genetic information. We will also extend the model to predict response to subsequent drug regimens, and to predict drug-resistant epilepsy at the initial diagnosis of epilepsy. We will validate these models in a separate cohort of new onset epilepsy patients in clinical practice. The models will allow personalised treatment for patients with epilepsy, leading to improved outcomes.
Project Lead - Professor Patrick Kwan
The enhanced model from the project above will be tested in a national multicentre randomised controlled trial called PERSONAL (Personalised Selection of Medication for Newly Diagnosed Adult Epilepsy) to assist with the selection of treatment for patients with epilepsy.
Project Leader - Dr Zhibin Chen
Seventy million people have epilepsy with 34 – 76 per 100,000 developing the condition every year. To formulate rational treatment plans, it is important to understand the different clinical courses and patterns of response to antiepileptic drugs, ideally by following outcomes from the point of treatment initiation. This project will perform an analysis focusing on patients presenting with a first seizure and their response to the initial therapies. It will also look at their relationship with long-term treatment outcomes and development of pharmacoresistance in newly treated epilepsy patients.
Long-term prognosis of Antiepileptic Drug Therapy in People with Newly Diagnosed and Treated Epilepsy
Project Leader - Dr Zhibin Chen and Prof Patrick Kwan
Antiseizure medications are the mainstay of treatment modality for epilepsy. People with epilepsy often require lifelong antiepileptic drug treatment. A previous study (in 2000, Glasgow) demonstrated a-third of the epilepsy patients did not respond well to antiseizure medications. Despite the introduction of more than a dozen new antiseizure medications in the past two decades, no robust data exists to suggest improvement in treatment outcomes in a recent expanded Glasgow study. To valid the patient prognosis and response patterns observed in the Glasgow studies, we will assess treatment outcomes of newly treated epilepsy patients who were seen at a First Seizure Clinic and followed up prospectively in Australia. We will extract data on seizures, diagnostic and treatment information at baseline, as well as follow-up clinical documents and construct a digital database. Various methods will be used in modelling treatment outcomes including traditional statistical methods and advanced machine learning approaches.
Project Leader - Dr Zhibin Chen
Epilepsy is one of the most common serious chronic neurological disorders and is estimated to affect approximately 68 million people worldwide. Antiseizure medications are the mainstay of treatment and suppress seizure occurrence. The Epilepsy treatment gap is a recognised public health issue in resource-poor countries where up to 80% of people with epilepsy do not receive appropriate treatment. This phenomenon in resource-rich countries has not been well studied. Our recent study in Australia demonstrated 30% of adults with epilepsy did not commence treatment immediately at diagnosis. In this project, we will investigate the differences in outcomes between patients who were treated and not treated upon diagnosis through linkage to state and national databases.

Research study

Epilepsy Cohorts for Artificial Intelligence Research (EpiCFAIR) View

Epilepsy Cohorts For AI Research (EpiCFAIR)

EpiCFAIR is a global collaboration aimed at improving personalised treatment for epilepsy through artificial intelligence (AI)

EpiCFAIR is a global, multi-cohort study led by Monash University that brings together de-identified real-world clinical data from people with newly diagnosed epilepsy. By integrating these diverse data sources, the project is building a secure and comprehensive resource to develop AI tools that predict the most effective antiseizure medication (ASM) for individual patients.

Our goal is to improve patient outcomes worldwide by reducing the trial-and-error approach in treatment, identifying individuals at risk of drug-resistant epilepsy and supporting personalised care.

ACTRN registration number: ACTRN12625000881437
Ethics Approval: Monash University Human Research Ethics Committee (Project ID: 40732)

Facts and figures

Collaborating Centres

34
Clinical, EEG, MRI, Genetic data types
ISO 9001 Quality Framework

Expression of interest form

What we do

Data collection

De-identified clinical, genetic, EEG and MRI from routine care, contributed by participating epilepsy centres

AI modelling

Machine learning/Deep learning are used to

predict individual responses to antiseizure medication (ASM)
identify indicators of drug-resistant epilepsy

Comparative analysis

Analysis of prescribing patterns across geographic regions.

Security & Privacy

All data are fully de-identified and securely stored at Monash University under robust governance processes.

Vision

To transform the global epilepsy care through data-driven, personalised treatment.

Mission

To build a secure, collaborative network that harnesses clinical data to power AI tools predicting the most effective ASM for epilepsy patients.

Our Global Collaborators

EpiCFAIR brings together clinicians, researchers and centres worldwide with a shared commitment to improving epilepsy outcomes. Hover over the map to see our collaborators

Quality Management System (QMS)

EpiCFAIR is supported by a Quality Management System (QMS) that ensures accurate, consistent and compliant data handling across all sites, providing a secure and standardised approach to research data.

The EpiCFAIR Quality Management System has been reviewed by Monash University Office of the Pro Vice-Chancellor (Research Infrastructure) to follow the framework of ISO 9001. It is the first clinical registry recognised to follow this framework at Monash University.

EpiCFAIR has adopted an ISO 9001-aligned Quality Management System View

Introduction

The EPICFAIR Platform, developed by Monash University’s Epilepsy & Personalised Medicine Group, supports data collection and management for the EPICFAIR research project. The platform aligns with Monash University's research infrastructure standards and applies quality management principles to ensure accuracy, consistency and compliance.

The EpiCFAIR Quality Management System has been reviewed by Monash University Office of the Pro Vice-Chancellor (Research Infrastructure) to follow the framework of ISO 9001. It is the first clinical registry recognised to follow this framework at Monash University

Overview

The EPICFAIR Platform provides a centralized system for managing the electronic data lifecycle of the EPICFAIR project. It supports structured, secure and traceable workflows that span data collection, validation, storage, access and distribution. Its framework is adaptable and can be extended to a range of research and sample-management contexts.

Scope of the Platform

Component	Description
Pre-Approval and Transfer Control	Enforces approval protocols prior to data or sample transfers, ensuring compliance.
Data Collection	Facilitates entry, validation and storage of clinical data, including notes and assessments.
Quality Control	Performs internal data checks to maintain accuracy and consistency.
Storage and Access	Ensures secure data storage with role-based access controls
Data Distribution	Supports compliant reporting and controlled data sharing.

Purpose

The platform’s Quality Management System (QMS) ensures that research data is handled in a standardised, compliant and auditable manner. It establishes consistent workflows, supports readiness for internal quality reviews (iQMS), and enables continuous improvement across data management practices.

The QMS framework provides a foundation for:

Standardised data workflows across research activities
Audit-readiness under Monash University’s internal QMS (iQMS)
Continuous improvement of data handling practices

This foundation allows EPICFAIR to meet current project requirements while maintaining flexibility for future research needs.

Join the EpiCFAIR Network

We invite clinicians and researchers with longitudinal registries of newly diagnosed epilepsy to contribute de-identified datasets with ≥12 months follow-up. Contributing centres will receive

updates on model development and research outputs
eligibility for co-authorship on resulting publications
opportunities for networking, collaboration and knowledge exchange

Ready to collaborate?

Complete the Expression of Interest form and our team will be in touch.

Expression of interest form

Prefer email? EpiCFAIR@monash.edu

Publication highlights

Enhancing Drug Response Prediction in Epilepsy with Emerging Multimodal Models: Focus on Clinical, Pharmacologic, and Genomic Factors. CNS Drugs May 2026
A Anderson, RSK Chang, M Howard, Z Chen, D Nhu, Z Ge, P Kwan

Role of machine learning in the management of epilepsy: a systematic review protocol. BMJ open, January 2024 RS Chang, S Nguyen, Z Chen, E Foster, P Kwan

Experience of waiting for seizure freedom and perception of machine learning technologies to support treatment decision: A qualitative study in adults with recent onset epilepsy. Epilepsy Research, February 2023 Reeder S, Foster E, Vishwanath S, Kwan P

Markov modelling of treatment response in a 30-year cohort study of newly diagnosed epilepsy. Brain May 2022 Simpson, H. D., Foster, E., Ademi, Z., Lawn, N., Brodie, M. J., Chen, Z. & Kwan, P.

Department of Neuroscience

Medicine, Nursing and Health Sciences