What is it about?

Lack of enough data is common in many measurement techniques, such as the novel biomedical application of positron emission particle tracking (PEPT). We used a method called the Gaussian process, a.k.a. kriging, to reconstruct the blood velocity in locally narrowed coronary arteries using the data from a limited number of particles and compare the results with the velocity field calculated using computational fluid dynamics (CFD). We studied three models representing varying severity and anatomical complexity: a narrowed straight tube, an idealized coronary bifurcation with stenosis, and patient-specific coronary arteries with a stenotic left circumflex artery. We found that kriging tended to overestimate area-averaged velocity in higher occlusion cases but accurately predicted maximum cross-sectional velocity. Importantly, kriging proved practical in estimating maximum velocity after stenosis, especially when the negative near-wall velocity did not exist.

Featured Image

Why is it important?

Using positron emission particle tracking (PEPT) in biomedical settings is novel and has many unknown aspects. This research contributes valuable insights into Gaussian process application for accurate velocity reconstruction in assessing coronary artery disease.


I hope this article encourages researchers to use Gaussian processes and in-silico approaches, i.e., simulation, to calibrate novel technologies and compensate for data sparsity in many fields. The article focuses on blood velocity reconstruction and positron emission particle tracking (PEPT); however, the proposed methodology is not limited to application. Above all else, I aspire for this article to stimulate your thoughts.

Hamed Keramati
King's College London

Read the Original

This page is a summary of: Using Gaussian process for velocity reconstruction after coronary stenosis applicable in positron emission particle tracking: An in-silico study, PLoS ONE, December 2023, PLOS,
DOI: 10.1371/journal.pone.0295789.
You can read the full text:

Open access logo


The following have contributed to this page