Image quality and radiation dose in spinal surgery: a comparison of three imaging systems in phantom

Margherita Casiraghi; Maria Antonietta Piliero; Luca Bellesi; Mathieu Dufour; Anindita Chatterjea; Pietro Scarone; Stefano Presilla

doi:10.1117/1.JMI.10.3.035001

Image quality and radiation dose in spinal surgery: a comparison of three imaging systems in phantom

J Med Imaging (Bellingham). 2023 May;10(3):035001. doi: 10.1117/1.JMI.10.3.035001. Epub 2023 Jun 9.

Authors

Margherita Casiraghi¹, Maria Antonietta Piliero¹, Luca Bellesi¹, Mathieu Dufour¹, Anindita Chatterjea², Pietro Scarone³, Stefano Presilla¹

Affiliations

¹ Imaging Institute of Southern Switzerland, Medical Physics Division, Bellinzona, Switzerland.
² Philips Healthcare, Image Guided Therapy Systems, Best, The Netherlands.
³ Neurocenter of Southern Switzerland, Neurosurgical Service, Lugano, Switzerland.

PMID: 37304527
PMCID: PMC10251430 (available on 2024-06-09)
DOI: 10.1117/1.JMI.10.3.035001

Abstract

Purpose: Using optimal settings for x-ray scans is crucial for obtaining three-dimensional images of high quality while keeping the patient dose low. Our work compares dose and image quality (IQ) of three intraoperative imaging systems [O-arm cone-beam computed tomography (CBCT), ClarifEye C-arm CBCT, and Airo computed tomography] used for spinal surgery.

Approach: Patients of 70, 90, and 110 kg were simulated with an anthropomorphic phantom by adding tissue-equivalent material. Titanium inserts were placed in the phantom spine for reproducing metal artifacts in the images. Organ dose was measured with thermo-luminescent dosimeters for effective dose ( $E$ ) calculation. Subjective IQ was assessed by ranking the images acquired with the manufacturer-defined imaging protocols. Objective IQ was assessed with a customized Catphan phantom.

Results: The ClarifEye protocols resulted in the lowest $E$ ranging from 1.4 to 5.1 mSv according to phantom size and protocol. The highest $E$ was measured for the high-definition protocol of O-arm ( $E$ 2.2 to 9 mSv) providing the best subjective IQ for imaging of the spine without titanium inserts. For the images with metal, the best IQ was obtained with ClarifEye. Airo ( $E$ 5.5 to 8.4 mSv) was ranked with the lowest IQ for images without metal while the rank improved for images with metal. Airo images had better uniformity, noise, and contrast sensitivity compared with CBCTs but worse high-contrast resolution. The values of these parameters were comparable between the CBCT systems.

Conclusions: Both CBCT systems provided better IQ compared with Airo for navigation of lumbar spinal surgery for the original phantom. Metal artifacts particularly affect O-arm images decreasing the subjective IQ. The high spatial resolution of CBCT systems resulted in a relevant parameter for the visibility of anatomical features important for spine navigation. Low dose protocols were enough to obtain a clinically acceptable contrast-to-noise ratio in the bones.

Keywords: computed tomography; cone-beam computed tomography; effective dose; image quality; intraoperative imaging; spinal surgery.