Epilepsy Case Management Team

Use this as a template for your team's page.

Team name & bio

• Lachlan - developer
• Taylor Griffin - developer
• Stuart Harris - developer

The Problem

Clinicians and medical scientists that work with largely qualitative data sets have trouble visualising the data that they have. Currently residing in xlsx files, transforming the data into something that is easier to visualise and use is difficult and time consuming. Potentially important conclusions are missed when the data looks like this

(Note: Actual data may differ from image shown)

The Solution

We have moved the data into a database, with Django modelling the data and providing the ability for more data to be added relatively easily. The Django REST Framework provides an API to the data.

Ember.js and D3.js access that data via the API to generate graphs per patient showing seizures and medication.

Application/Relevance

In particualr, this project gives clinicians the ability to quickly and easily visualise a patient's longitudinal seizure, surgery and medication history. The visuals give a quick view of how medication has changed frequency or severity of seizures experienced.

(Note: Actual visualisation may differ from image shown) Datasets

Provided by Project owners

Links

• Link to github repository
• Link to clone github repository
• Link to ECM Viz Demo
• Link to Epilepsy Case Management Admin Demo login demo/demo
• Link to ECM Front End
• Link to youtube video

You need a video uploaded for us to present at closing!

Tech stack

• Ubuntu hosted by Nectar
• Python 2.7 language
• Django web framework
• Sqlite3 database
• Ember javascript SPA framework
• D3 javascript visualisation frameworks.

Tradeoffs/analysis

• Spent more time planning architecture and infrastructure
• More time whiteboarding problem
• better integration of the Django and Ember.js
• not all team members were knowledgable in all technologies - alternatively, great to have diverse skill sets
• Project Owner's initial data format made it hard for the owner's to visualise just how much could be done
• Lack of time meant we could only implement broad stroke ideas - potentialities.
• Project owner mentioned that they weren't collecting enough or the right type of data - that can be improved as end users become more comfortable with the project.

Future functionality

• exporting data to csv - per patient, per medication, per surgery type
• visualisation on medications or surgery types rather than patients
• improving the visualisation interface to include adjustable date ranges, inclusion and exclusion of data
• search - conceivably patient numbers would be very large.
• Authorization and Authentication - restrict who can login and what they see when they are logged in
• because there is an API to the data, anything can be visualised with enough finessing.
• add more details to models - case notes for patients