Evaluation of a PET Insert for Trimodal Imaging: A Step Towards PET/MRI-Guided Focused Ultrasound

F. Lopez-Berenguer, Andrea Gonzalez‐Montoro, Marta Freire, A. González-Tamarit, Santiago Jiménez-Serrano, Luis Vidal, Miguel Gil‐Gil, Francis Loignon-Houle, I. Torres-Espallardó, Jonathan Sachs, Joseph V. Rispoli, Nicholas J. Tustison, Stuart S. Berr, Mark B. Williams, D. Andrés, Juan J. Rodríguez-García, J. L. Alonso-Ramos, Antonio J. González, Alicia Carrión, Francisco Camarena, Teresa Guallart‐Naval, José Miguel Algarín, Rubén Bosch, Eduardo Pallás, Joseba Alonso, Natasha Sheybani, Stefanyda Maslova, Thomas Sherlock, Andrew T Thede, Zehra E.f. Demir, Michael Hoch, Jiabao He, J. Benlloch, María José Rodríguez-Álvarez, Nóe JiménezIEEE Transactions on Radiation and Plasma Medical Sciences 2025 · DOI

TL;DR

This paper introduces a novel preclinical PET insert designed for simultaneous operation with MRI and Focused Ultrasound (FUS) systems, paving the way for true trimodal imaging. The system demonstrates impressive spatial resolution and compatibility, marking a significant step towards image-guided therapeutic applications.

What Problem Does This Paper Address?

The medical imaging field is increasingly exploring the synergy of multiple modalities to gain comprehensive insights into biological processes and guide therapeutic interventions. Positron Emission Tomography (PET) offers functional and metabolic information, Magnetic Resonance Imaging (MRI) provides high-resolution anatomical detail and real-time thermometry, and Focused Ultrasound (FUS) enables precise, non-invasive therapeutic delivery. Combining these three modalities (PET/MRI-FUS) holds immense potential for applications like targeted drug delivery and neuromodulation, especially for conditions requiring precise spatial targeting and real-time monitoring. However, a significant gap exists: no commercial or academic system currently allows for simultaneous PET, MRI, and FUS imaging. Existing approaches are often sequential or limited to dual-modality combinations, hindering comprehensive, real-time assessment and optimization of therapies. This paper addresses this challenge by developing and evaluating a preclinical PET insert specifically designed for seamless integration into a trimodal PET-MRI-FUS system.

Methods

The core of this work is the design and evaluation of a preclinical PET insert. The insert utilizes 16 monolithic LYSO:Ce crystals arranged in a two-ring octagonal structure, providing an inner diameter of 72 mm and an outer diameter of 114 mm, with an axial field of view (FOV) of 67 mm. A key feature is the implementation of Depth of Interaction (DOI) capabilities, which are crucial for mitigating parallax errors and achieving uniform spatial resolution across the FOV, especially in smaller preclinical systems. The system's performance was rigorously assessed using the NEMA NU 4-2008 protocol, a standard for evaluating small animal PET scanners. This involved measuring spatial resolution, sensitivity, and count rate performance. Image reconstruction was performed using the Maximum Likelihood Expectation-Maximization (MLEM) algorithm, optimized with GPU acceleration. Compatibility tests were conducted using both phantom studies and in vivo experiments, integrating the PET insert with low-field and high-field MRI systems, as well as custom and commercial FUS devices.

Datasets

The paper does not describe training datasets. Evaluation was performed using standard phantoms:

  • NEMA IQ phantom: Used for assessing image performance, including uniformity, recovery coefficients (RCs), spill-over ratio (SOR), and standard deviation (STD).
  • Micro-Derenzo phantom: Used to evaluate the system's ability to resolve small structures and assess spatial resolution.
  • Fillable mouse phantom: Used to simulate biological structures and lesions for imaging.
  • Custom phantom: A gelatine mixture with copper sulfate and [18F] FDG, used for proof-of-concept trimodal testing.
  • Naïve murine brain: Used for in vivo trimodal compatibility testing.

Key Insights and Ideas

  • DOI is Crucial for Preclinical PET: The study strongly emphasizes that Depth of Interaction (DOI) capabilities significantly improve spatial resolution, particularly in the radial direction, which is critical for small FOV preclinical PET systems where parallax errors are more pronounced.
  • Seamless Integration is Achievable: The research demonstrates that a PET insert can be designed to be compatible with both MRI and FUS systems without substantial signal degradation or artifacts, overcoming a major hurdle in trimodal imaging.
  • Monolithic LYSO Crystals Offer Advantages: The use of monolithic LYSO crystals, combined with DOI capabilities, allows for high spatial resolution and sensitivity, making it a competitive choice for preclinical PET inserts.
  • Real-time Monitoring Potential: The successful simultaneous acquisition of PET, MRI, and FUS data in both phantom and in vivo studies highlights the potential for real-time monitoring of biological processes and therapeutic effects.

Key Results

  • Spatial Resolution: Achieved a uniform spatial resolution of approximately 0.9 mm across the entire FOV with DOI correction. Without DOI, the resolution degraded to ~1.36 mm in the radial direction at the center FOV (cFOV).
  • Sensitivity: A peak sensitivity of 3.8% was achieved for an energy window of 255–766 keV.
  • Count Rate Performance: The system reached a peak noise equivalent count rate (NECR) of 80 kcps at an activity of 19 MBq.
  • Compatibility: The PET insert demonstrated successful simultaneous operation with both low-field and high-field MRI systems, as well as with commercial FUS devices, in both phantom and in vivo experiments. This included visualizing the diffusion of a radiotracer in response to FUS-induced heating and demonstrating co-localization of contrast agents in vivo.

Limitations and Future Work

  • NECR Peak: The PET insert exhibits a lower NECR peak compared to some other preclinical systems. The authors attribute this to the thinner scintillators (8 mm) used for MRI/FUS integration and the shorter axial length of the prototype.
  • Reconstruction Software: While MLEM was used, further improvements in reconstruction software are planned, including the implementation of dead-time and other quantitative corrections to ensure accurate activity estimation, especially for dynamic or high-activity studies.
  • Optimization: Future studies will focus on optimizing the PET system performance by adding quantification features, enhancing image corrections, and developing a pseudo-automatic PET-MRI co-registration algorithm. This will enable more detailed exploration of biological processes and advanced in vivo experiments.
  • Human Translation: The authors suggest that the success of these compatibility tests lays the groundwork for extending this technology to human-sized systems for clinical trials.

Summary updated

Abstract

Combining Positron Emission Tomography (PET) with Magnetic Resonance Imaging (MRI) and Focused Ultrasound (FUS) has emerged as a promising hybrid technique in the medical imaging field. Despite the potential benefits of using simultaneous PET and MRI acquisitions to monitor therapeutic effects and ensure precision and safety during FUS applications, no commercial or academic PET/MRI-guided FUS system (referred in the following as trimodal PET-MRI-FUS) exists. This work presents the design and evaluation of a preclinical PET insert for simultaneous operation with MRI and FUS systems. The proposed PET insert is based on monolithic LYSO crystals arranged in two octagonal rings that define inner and outer diameters of 72 mm and 114 mm, respectively, with an axial length of 67 mm. The system performance was evaluated according to the NEMA NU 4 2008 protocol. It achieves a uniform spatial resolution of ~0.9 mm along the whole PET Field of View (FOV) when enabling Depth of Interaction (DOI) capabilities. Moreover, we found a sensitivity peak of 3.8% for an energy window of 255-766 keV and a peak noise equivalent count rate peak of 80 kcps at an activity of 19 MBq. The compatibility of the PET insert with both low-and high-field MRI systems, as well as with custom and commercially available FUS devices, was studied using phantoms and in vivo experiments, respectively. The results of the trimodal tests have proven that the PET insert works simultaneously with both MRI and FUS systems.
BibTeX
@article{lopezberenguer2025evaluation,
  author = {F. Lopez-Berenguer and Andrea Gonzalez‐Montoro and Marta Freire and A. González-Tamarit and Santiago Jiménez-Serrano and Luis Vidal and Miguel Gil‐Gil and Francis Loignon-Houle and I. Torres-Espallardó and Jonathan Sachs and Joseph V. Rispoli and Nicholas J. Tustison and Stuart S. Berr and Mark B. Williams and D. Andrés and Juan J. Rodríguez-García and J. L. Alonso-Ramos and Antonio J. González and Alicia Carrión and Francisco Camarena and Teresa Guallart‐Naval and José Miguel Algarín and Rubén Bosch and Eduardo Pallás and Joseba Alonso and Natasha Sheybani and Stefanyda Maslova and Thomas Sherlock and Andrew T Thede and Zehra E.F. Demir and Michael Hoch and Jiabao He and J. Benlloch and María José Rodríguez-Álvarez and Nóe Jiménez},
  title = {{Evaluation of a PET Insert for Trimodal Imaging: A Step Towards PET/MRI-Guided Focused Ultrasound}},
  journal = {IEEE Transactions on Radiation and Plasma Medical Sciences},
  year = {2025},
  doi = {10.1109/trpms.2025.3615467},
}
Figure