A projection-domain deep learning approach for respiratory motion correction of myocardial perfusion imaging…

AI / informaticsMedical physics· Aug 2026

A projection-domain deep learning approach for respiratory motion correction of myocardial perfusion imaging using a multi-pinhole solid-state SPECT.

Dylan J Malenfant, Terrence D Ruddy, R Glenn Wells

Cardiovascular ImagingAI / informaticsQ1IF 3.5

AI estimated respiratory motion within about 2 mm of reference alignment in cardiac SPECT.

A projection-domain deep learning approach for respiratory motion correction of myocardial perfusion imaging using a multi-pinhole solid-state SPECT.

Dylan J Malenfant … R Glenn Wells

Medical physics · Aug 2026 · Q1 in Radiology

Background

Myocardial perfusion imagining (myocardial perfusion imagining) is a nuclear medicine technique used in the assessment of coronary artery disease.

Purpose

An AI-based method for estimating respiratory motion from gated projection data for myocardial perfusion imagining SPECT using a dedicated pinhole scanner was developed.

Methods

Respiratory motion motion parameters were calculated for rest and stress myocardial perfusion imagining scans using a minimized root mean squared (root mean squared) alignment of reconstructed respiratory gates for 90 patients supplemented with 33 NCAT simulated patients.

n = 90 patients

Results

AI alignment stayed close to reference, well under 2 mm error

Clinical noise levels

1.86mm

1/8 signal

2.07mm

More results

Motion correction with AI was shown to provide a statistically significant increase anterior wall contrast and decrease measured wall thickness relative to no motion correction, with comparable results to root mean squared motion correction.

“

Conclusion · 1 of 2

Direct AI-based motion correction parameter estimation using gated projection data was shown to be feasible in clinical settings.

Conclusion · 2 of 2

Corrections based on AI estimates showed improved contrast and reduced myocardial wall thickness when incorporated into image reconstruction.

A projection-domain deep learning approach for respiratory motion correction of myocardial perfusion imaging using a multi-pinhole solid-state SPECT.

A projection-domain deep learning approach for respiratory motion correction of myocardial perfusion imaging using a multi-pinhole solid-state SPECT.

Related papers