Medical software PIQE

CTvisualizationdiagnostic

Overview
PIQE is a Deep Learning Reconstruction (DLR) algorithm designed to maximize CT spatial resolution and deliver Super Resolution images for cardiac, lung and body examinations. It improves fine anatomical detail and low‑contrast detectability while operating on routine scan data.

Image quality
PIQE produces images with sharper anatomical detail and clearer borders between structures to aid diagnostic confidence. The 1024 matrix supports up to 4× enlargement compared with a 512 matrix, preserving spatial detail at higher zoom levels.

PIQE in practice
PIQE offers higher spatial resolution and reduced image noise versus conventional reconstruction methods. In cardiac CT it enables clearer visualization of small atherosclerotic plaques and calcifications, and reduces blooming artifacts from stents to improve stent‑strut and lumen delineation.

Benefits

• Sharper anatomical detail for improved delineation of small structures
• No additional radiation dose required to achieve enhanced detail
• 4× zoom capability via 1024 matrix (enlargement from 512 to 1024)

Trained for precision
PIQE is trained to reconstruct UHR‑CT 0.25 mm images acquired with Aquilion Precision from down‑sampled 0.5 mm data. Its 3D neural network architecture maximizes spatial resolution on routine scans to raise CT image quality standards.

Background
Canon has continuously advanced CT image quality technologies since 2011. Deep learning was first applied to reconstruction with AiCE in 2018; PIQE, introduced in 2021 (initially with 512 matrix and now improved to 1024), represents the next generation of DLR technology.

Clinical evidence and use cases
PIQE is indicated for cardiac, lung and body imaging and is associated with improved low‑contrast detectability, reduced blooming artifacts (e.g., around stents) and clearer visualization of small plaques and calcifications.

Selected testimonials
"What struck me with PIQE is that it’s really a step forward compared to other reconstruction techniques. The difference is in sharper images, increased spatial resolution, also with less image noise." — Prof. Mickaël Ohana

"PIQE offers higher spatial resolution and lower image noise compared to conventional reconstruction methods. PIQE is considered a groundbreaking technology that should be widely implemented across various areas of the body." — Dr. Fuminari Tatsugami

Technical specifications

• Type: Deep Learning Reconstruction (DLR) algorithm
• Manufacturer platform: Trained for Aquilion Precision UHR‑CT
• Reconstruction target resolution: UHR‑CT 0.25 mm from 0.5 mm down‑sampled data
• Matrix options: 1024 matrix (improved from initial 512)
• Zoom: Allows up to 4× enlargement of a 512 image using 1024 matrix
• Clinical focus: Cardiac, lung and body examinations
• Primary benefits: Increased spatial resolution, excellent low‑contrast detectability, reduced blooming artifacts
• Introduced: 2021 (initial release with 512, later improved to 1024)