Within my project with the System Modelling and Quantitive Biomedicine group with the University of Birmingham, I worked with a team of researchers who are looking into developing mathematical models of seizure forecasting in epilepsy. This project particularly attracted me as I was diagnosed with epilepsy myself 10 years ago and one of the main concerns that arose through my diagnosis, as well as many others living with epilepsy, was the unpredictable nature of seizures. Seizure propensity tends to be influenced by “triggers”, for example sleep deprivation, stress and alcohol intake. The team I worked with are analysing these such triggers and quantifying these using data from wearable devices. This will then look at giving a patient a seizure forecast for their day, giving them an idea of any seizures they might experience depending on the trigger factors they have been exposed too.

I began to look at the procedures the team were using to predict the seizures. They were effectively taking a collecting of factors or ‘triggers’ of data and comparing these against a persons usual activity to provide them with a forecast of their seizures. This is something that the patient would check every day and allow them to plan their day, something we all do when looking at the weather forecast, something that was initial thought when looking at the process. Weather forecasts are made in a similar way, by collecting data about the current state of the atmosphere. Before any forecast can be made, a meteorologist must first understand what the current weather conditions are and what is producing them, a lot like how the researchers are looking at what triggers are causing the patients seizures. This is done by examining a large quantity of observation data including surface observations, satellite imagery, radar data, radiosonde data, upper-air data, wind profilers, aircraft observations, river gauges, and simply looking outside, comparable to what my team are doing in looking at the collection of seizure triggers, such as stress hormones and sleep activity.

This similarity between weather forecasting and the seizure forecasting, led me to create a series 16 weather forecasts that would speak about a persons seizure activity rather than the weather but be presented in the same way as we usually see a weather forecast on the television. The forecasts are filmed as if they they were the same as any weather report we see on the television everyday, but behind them is a distorted image that shows activity of the seizures on the map of a brain.

This nature of a seizure forecast becoming something we see and watch in our everyday lives, not only shows people that seizure activity is a part of a patients everyday life but also bringing seizure forecasts in to their everyday activity. The work stars students Ethan and Connor from the Birmingham City University Media Production course and was filmed at Birmingham City University. 

All of the forecasts can be found on my Youtube channel.