Skip to content

Minicourse EDEP-SP09

From acquisition to image: a step-by-step tutorial in high-density diffuse optical tomography


with Sam Powell, Nida Duobaitė, and colleagues

About this course

This minicourse will demonstrate the process of collecting and analysing high-density diffuse optical tomography data, from acquisition to image reconstruction, and give you the skills and knowledge to replicate this process independently.

Our tutorial will begin with an overview of LUMO: a modular, whole-head, wearable, high-density diffuse optical tomography system. We will demonstrate the simplicity of configuring the system and acquiring a dataset, before explaining each of the steps required to process the many thousands of channels of data that result: moving from standard signal processing techniques familiar to fNIRS practitioners, through the use of high-density superficial regression to minimise physiological noise, to techniques for image reconstruction and statistical mapping.

You will be invited to follow each step of the process using your own computer (provided that you have a recent installation of MATLAB or Python on your laptop). By the end of this workshop, you will have built a fully working imaging pipeline which you can apply to your own experimental data. This workshop will particularly benefit those looking to gain an understanding of the advantages of high-density diffuse optical tomography over conventional fNIRS, and those who wish to develop an understanding of image reconstruction techniques.

Learning Outcomes

  • Gain a broad appreciation of each of the stages of a high-density diffuse optical tomography experiment.
  • Develop an understanding of the advantages of high-density data in the context of improved rejection of physiological noise
  • Develop an understanding of the central concepts of moving from the channel space to the image space, including appropriate statistical analysis.

Course Plan

Level: Introductory

Pre-Requisites: None (but attendees may wish to bring a laptop with a recent installation of MATLAB or Python in order to follow the tutorial)
Course Duration: 3 hours

Delivery Plan

10 min: Introduction
15 min: System configuration and data acquisition
15 min: Discussion, installation and configuration of analysis software
15 min: Theoretical overview of channel space analysis
25 min: Data management, pre-processing, and regression
15 min: Theoretical overview of image reconstruction
25 min: Image reconstruction
15 min: Statistical analysis
30 min: Q&A

Why enrol on this course?

The optimal use of high-density diffuse optical tomography data requires the application of a broad range of techniques from the fields of signal processing, statistical analysis, and inverse problems. In practice, excellent results can be achieved with an appropriate grounding in a small subset of each of these fields. It is our goal to guide attendees through the complexity to concentrate on the pertinent concepts and methods and, in doing so, make the significant advantages offered by HD-DOT techniques more widely accessible.