Summary

Many software tools have been developed and are essential for the clinical and preclinical analysis of single photon emission computed tomography (SPECT) and positron emission tomography (PET) imaging.
The commercially available software covers various research areas, including oncology, neurology, and cardiovascular entities. Software tools bear great potential to analyze data by automated anatomical volume of interest quantifications.
However, comparing software applications is still challenging and can result in different dosimetry analyses of clinical data derived from peptide receptor radionuclide therapy (PRRT) using ¹⁷⁷Lu-DOTATATE. For over a decade, tools for automated quantifying myocardial ischemia and wall motion defects in cardiac SPECT imaging have offered valid programs and algorithms. However, differences in certain features and artifacts might need to be manually corrected by the user despite automated processing. ⁸²Rubidium PET myocardial perfusion quantification in patients was recently used to validate the novel Carimas software. Cross-comparison studies of compartment models in hypertrophic cardiomyopathy (HCM) used the software tools Carimas, Flowquant, and PMOD to evaluate ¹³N-ammonia PET for myocardial perfusion. This study showed consistent global and regional myocardial blood flow (MBF) values. However, there was significant variability in segmental values supplied by the circumflex branch of the left coronary artery. These differences limit the interchangeability of the studied software tools. Another study quantified the MBF in HCM patients. It showed that PMOD and QPET (Cedars Sinai) could not be used interchangeably due to anatomic characteristics in HCM patients compared to non-HCM patients.
Comparison of QPET, syngo MBF, and PMOD resulted in excellent correlations in myocardial flow reserve (MFR) in ¹³N-ammonia PET and the respective vascular territories. However, reproducibility for other cardiac tracers is still challenging. In previous reports, ⁸²rubidium imaging quantification depended on software tools (e.g., PMOD, FlowQuant, and syngo MBF).
Gated data acquisition in myocardial perfusion SPECT and PET allows analysis of wall motion and calculation of left ventricular volumes, including end-diastolic (EDV), end-systolic volumes (ESV), and left ventricular ejection fraction (EF). Left ventricular cardiac volumes and EF are reliable prognostic parameters in patients. In the clinical setting, ¹⁸F-FDG represents a widely used tracer detecting hibernating myocardium, prosthetic valve endocarditis and device infections, and sarcoidosis. Previous clinical studies comparing QGS and 4D-MSPECT to magnetic resonance imaging (MRI) showed a good agreement for EDV, ESV, and EF in patients with coronary heart disease. Several validated software packages are commercially available for humans. In addition, the ¹⁸F-FDG tracer can also be utilized in small animal PET research for the detection of myocardial defects and to assess murine heart function.
To date, no studies are comparing different software tools for cardiac volumes and function in basic murine research.
In this study, the authors sought to compare the left ventricular parameters using PMOD, MIM, and QGS in cardiac small animal ¹⁸F-FDG PET/CT imaging, thereby contributing to the existing literature on its feasibility and differences and assessing interchangeability.

Results from the nanoScan PET/CT

ECG-gated ¹⁸F-FDG-PET/CT scans were performed using the Mediso nanoScan small-animal PET/CT scanner. The animals had free access to food and water until before the scan, as described previously. No prior fasting was used in the protocol due to enhanced 18F-FDG uptake upon isoflurane anesthesia and exclusion of further evaluation regarding tracer uptake such as SUVs/cardiac injected activity per gram. Bedding was changed regularly to avoid ingestion of bedding.
Anesthesia was induced (2.5%) and maintained (2.0%) with isoflurane delivered in pure oxygen at a rate of 1.5 L/min via a face mask. The core body temperature was maintained within the normal range using a heating pad and monitored by a rectal thermometer.
After placing an intravenous catheter into a tail vein, approximately 20 MBq of ¹⁸F-FDG was injected in a volume of ~ 0.1 ml. The catheter was then flushed with 0.05 ml of isotonic saline solution. Animals remained anesthetized during the entire scan and were placed in a prone position within the PET/CT scanner.
Modified neonatal needle ECG electrodes (Kendall, Cardinal Health, Dublin, Ireland) were placed into both forepaws and the left hind paw. An integrated physiological monitoring system recorded the ECG signal and vital parameters.
First, a CT scan (semicircular full trajectory, maximum field of view, 480 projections, scan mode helical, pitch 1.0, X-ray power 35 kVp × 900 µA, exposure per projection 170 ms, and 1:4 binning) was acquired for attenuation correction. The ECG-gated PET recording was initiated 30 min after the tracer injection and lasted 15 min. Recovery from anesthesia and the PET/CT scan was monitored closely by a veterinarian. Gated mouse PET studies were reconstructed using the Tera-Tomo 3-dimensional reconstruction algorithm (Nucline NanoScan, Mediso), which includes point-spread correction and the following settings: 8 iterations, 6 subsets, normal regularization, median and spike filter on, edge artifact reduction on, voxel size of 0.5 mm, and 400- to 600- keV energy window, and coincidence mode 1–3. Gating parameters were set at 16 frames. All PET data were corrected for randoms, scatter, attenuation, and decay.
Figure 1. illustrates representative PET images after the DICOM file importing into the respective software tools, showing cardiac small animal ¹⁸F-FDG PET images in QGS, PMOD, and MIM. The cardiac PET images illustrating the commercially available software QGS (left), PMOD (middle), and MIM (right) in the same healthy mouse, each in end-diastole (ED) and end-systole (ES). The upper row shows short axis view (SAX). The middle row shows the horizontal long axis (HLA), and the bottom row illustrates the  vertical long axis (VLA). Anterior (ANT), septal (SEP), lateral (LAT), inferior (INF). Color scale ranging from 0 to 100% for each software are displayed below.

• Analyzing heart volume and function in small animal research is a cornerstone in preclinical cardiac research. Regarding the previous lack of published literature, the present work compares three commercially available software tools for cardiac function in healthy mice using gated ¹⁸F-FDG PET/CT.
• The authors' results indicate that these software tools are feasible in small animal research. The calculated cardiac values are consistent with the published literature. There was a good to excellent correlation for EDV, ESV, and SV comparing the software tools.
• Bland–Altman analysis showed differences in agreement, limiting the interchangeability of PMOD, MIM, and QGS for murine cardiac volume and function assessment.

Full article on ejnmmires.springeropen.com