Alan Aberdeen


Now

Founder — Ground Truth Labs — a spatial biomarker platform for bone marrow histology to better diagnose, monitor treatment response, and improve outcomes for cancer patients.


Recently

Designed and developed AIDA; an open-source web-based annotation tool for histology whole-slide-imaging and bioinformatics research.

AIDA

This project addresses problems with current practice: Until recently pathology image annotation has been largely manual. Even when digital recording is used, it tends to be on stand alone systems.

So we built a lightweight, easy to use browser-based web-service using open source APIs.

In current work, we explore the use of computer supported interaction to enhance annotation. Machine learning methods are used to provide context aware tools and human-in-the-loop processes that leverage the knowledge of the expert human user and the processing strength of the machine.

Design for multimodal interfaces. Thinking about speech interaction: where it has come from, what's changed and how it is implemented.

Study of computer supported interfaces, how can we best leverage machines in our interactive tools.

Design and evaluation of interactive systems. How to build in quality, control direction and measure engagement. Strong emphasis on user-centered design techniques.

Gained a deeper appreciation for Information Visualisation, particularly interactive representations of abstract data.

They lure you in with promise of building cool things... then it's maths, maths and more maths. Sometimes still kinda cool.

We cover all engineering disciplines; structural, electrical, mechanical and chemical... Eventually, I moved towards a focus on software and information engineering in a biomedical context.

Thesis: Automated Monitoring of Cell Populations

A detailed understanding of cell dynamics is essential for combating diseases and improving health. One of the main tools for analysing cell behaviour is cell-imaging over time. Tracking cells in a high-throughput environment is difficult and produces large amounts of data.

This is an open-source Python implementation of a cell tracking technique, Coupled Minimum-Cost Flow Tracking (CMCFT) (Padfield, Rittscher, & Roysam., 2011), for monitoring cell populations in-vitro.