Alternative headphones for patient noise protection and communication in PET-MR studies of the brain

Lutz Tellmann, Hans Herzog, Frank Boers, Christoph Lerche, N. Jon ShahEJNMMI 2018 · DOI

TL;DR

This paper introduces a novel, low-cost earphone system for PET-MR imaging that effectively protects patients from loud scanner noise and facilitates communication without introducing the image artifacts caused by traditional headphones. This solution is crucial for improving the accuracy and reliability of PET-MR brain studies.

What Problem Does This Paper Address?

Combined PET-MR scanners offer powerful diagnostic capabilities by integrating functional (PET) and anatomical/physiological (MR) imaging. However, a significant challenge arises from the intense noise generated by the MR scanner's gradient coils, necessitating noise protection for patients. Traditional solutions involve non-magnetic headphones. The problem is that these standard headphones, particularly when used in conjunction with modern, tightly packed multichannel head coils, can cause significant attenuation artifacts in PET images. These artifacts arise because the headphone material absorbs PET photons, leading to inaccurate quantitative measurements and potentially misleading visual interpretations, especially in brain imaging where precise quantification is critical. Furthermore, the limited space within advanced MR coils makes fitting conventional headphones difficult.

Methods

The researchers developed an alternative headphone system by modifying commercially available, disposable foam earplugs. They connected these earplugs to the pneumatic tubes of the original manufacturer-supplied headphones, creating a custom earphone system.

To evaluate the impact of these earphones, they employed a two-pronged approach:

  1. Transmission Imaging: They used a dedicated PET system with a 68Ge transmission source and a head phantom. Transmission scans were performed with the standard headphones, the custom earphones, and without any headgear. This allowed them to measure the linear attenuation coefficient of each device.
  2. Emission Imaging: The head phantom was filled with an 18F solution, and emission scans were acquired in the PET-MR scanner with and without the headphones/earphones. This enabled them to assess the influence of the devices on PET image quality and quantification.

Both transmission and emission data were reconstructed using standard PET reconstruction algorithms (OSEM2D and OP-OSEM3D). Regional image analysis with Regions of Interest (ROIs) was performed to quantify attenuation coefficients and activity concentrations.

Datasets

The study utilized a lida brain phantom for both transmission and emission imaging experiments. This phantom was filled with 18F solution for emission studies. No patient datasets were used for the primary evaluation of the headphone/earphone systems.

Key Insights and Ideas

  • Material Matters: The choice of material for patient accessories in PET-MR is critical. Even seemingly minor items like headphones can introduce significant quantitative errors if they attenuate PET photons.
  • Artifacts are Quantifiable: The study demonstrated that standard headphones cause a measurable underestimation of activity in PET images, directly linked to their attenuation properties.
  • DIY Solutions Can Be Effective: A simple modification of readily available earplugs can create a highly effective alternative that avoids the pitfalls of standard equipment.
  • Patient Comfort is Key: While the technical performance is paramount, patient acceptance is also a crucial factor for clinical adoption. The developed earphones were well-received by patients.

Key Results

  • Attenuation: The standard headphones had a linear attenuation coefficient of 0.11 cm⁻¹, while the custom earphones were not identifiable in the transmission images, indicating negligible attenuation.
  • PET Image Artifacts: Emission images showed an activity underestimation of up to 10% near the standard headphones. In contrast, the custom earphones showed no discernible influence on the emission images, with a deviation of only 0.8% compared to the reference.
  • MR Image Quality: Neither the standard headphones nor the custom earphones caused any visible artifacts in the MR images.
  • Patient Acceptance: A questionnaire revealed high patient acceptance for the earphones, with 89% rating the overall communication quality positively. Noise reduction (76%) and comfort (68%) were also rated positively.

Limitations and Future Work

Limitations:

  • Preparation Time: The authors note that the preparation time for inserting the custom earphones was slightly increased compared to standard headphones.
  • Stereo and Volume Control: The air-driven system used for the custom earphones lacks stereo discrimination and individual volume control, making it unsuitable for audiometric studies requiring precise audio delivery.

Future Work:

  • The authors suggest that for audiometric studies (e.g., in fMRI), commercial, non-pneumatic earphone systems with better acoustic delivery and noise attenuation should be considered.
  • They also mention that an equivalent commercial solution (MagnaCoils) exists, but their in-house built earphones are cost-effective and perform comparably, suggesting no immediate need to replace their current solution.

Summary updated

Abstract

The presented earphones, which can be connected to the existing patient communication system, are a preferable alternative to the conventional headphones, as, in contrast to the use of headphones, qualitative and quantitative errors in the PET images can be avoided. Patient acceptance of the earphones was high, despite the increase in preparation time before the PET-MR study.
BibTeX
@inproceedings{tellmann2018alternative,
  author = {Lutz Tellmann and Hans Herzog and Frank Boers and Christoph Lerche and N. Jon Shah},
  title = {{Alternative headphones for patient noise protection and communication in PET-MR studies of the brain}},
  booktitle = {EJNMMI},
  year = {2018},
  doi = {10.1186/s13550-018-0457-6},
}
Figure