Evaluation of manual and automated approaches for segmentation and extraction of quantitative indices from [18F]FDG PET-CT images.

Utilisation of whole organ volumes to extract anatomical and functional information from computed tomography (CT) and positron emission tomography (PET) images may provide key information for the treatment and follow-up of cancer patients. However, manual organ segmentation, is laborious and time-consuming. In this study, a CT-based deep learning method and a multi-atlas method were evaluated for segmenting the liver and spleen on CT images to extract quantitative tracer information from Fluorine-18 fluorodeoxyglucose ([18F]FDG) PET images of 50 patients with advanced Hodgkin lymphoma (HL). Manual segmentation was used as the reference method. The two automatic methods were also compared with a manually defined volume of interest (VOI) within the organ, a technique commonly performed in clinical settings. Both automatic methods provided accurate CT segmentations, with the deep learning method outperforming the multi-atlas with a DICE coefficient of 0.93 ± 0.03 (mean ± standard deviation) in liver and 0.87 ± 0.17 in spleen compared to 0.87 ± 0.05 (liver) and 0.78 ± 0.11 (spleen) for the multi-atlas. Similarly, a mean relative error of -3.2% for the liver and -3.4% for the spleen across patients was found for the mean standardized uptake value (SUVmean) using the deep learning regions while the corresponding errors for the multi-atlas method were -4.7% and -9.2%, respectively. For the maximum SUV (SUVmax), both methods resulted in higher than 20% overestimation due to the extension of organ boundaries to include neighbouring, high-uptake regions. The conservative VOI method which did not extend into neighbouring tissues, provided a more accurate SUVmaxestimate. In conclusion, the automatic, and particularly the deep learning method could be used to rapidly extract information of the SUVmeanwithin the liver and spleen. However, activity from neighbouring organs and lesions can lead to high biases in SUVmaxand current practices of manually defining a volume of interest in the organ should be considered instead.

Bone turnover change after randomized switch from tenofovir disoproxil to tenofovir alafenamide fumarate in men with HIV.

OBJECTIVE: Bone loss in people with HIV (PWH) is poorly understood. Switching tenofovir disoproxil fumarate (TDF) to tenofovir alafenamide (TAF) has yielded bone mineral density (BMD) increases. PETRAM (NCT#:03405012) investigated whether BMD and bone turnover changes correlate. DESIGN: Open-label, randomized controlled trial. SETTING: Single-site, outpatient, secondary care. PARTICIPANTS: Nonosteoporotic, virologically suppressed, cis-male PWH taking TDF/emtricitabine (FTC)/rilpivirine (RPV) for more than 24 weeks. INTERVENTION: Continuing TDF/FTC/RPV versus switching to TAF/FTC/RPV (1 : 1 randomization). MAIN OUTCOME MEASURES: :[ 18 F]NaF-PET/CT for bone turnover (standardized uptake values, SUV mean ) and dual-energy x-ray absorptiometry for lumbar spine and total hip BMD. RESULTS: Thirty-two men, median age 51 years, 76% white, median duration TDF/FTC/RPV 49 months, were randomized between 31 August 2018 and 09 March 2020. Sixteen TAF:11 TDF were analyzed. Baseline-final scan range was 23-103 (median 55) weeks. LS-SUV mean decreased for both groups (TAF -7.9% [95% confidence interval -14.4, -1.5], TDF -5.3% [-12.1,1.5], P  = 0.57). TH-SUV mean showed minimal changes (TAF +0.3% [-12.2,12.8], TDF +2.9% [-11.1,16.9], P  = 0.77). LS-BMD changes were slightly more favorable with TAF but failed to reach significance (TAF +1.7% [0.3,3.1], TDF -0.3 [-1.8,1.2], P  = 0.06). Bone turnover markers decreased more with TAF ([CTX -35.3% [-45.7, -24.9], P1NP -17.6% [-26.2, -8.5]) than TDF (-11.6% [-28.8, +5.6] and -6.9% [-19.2, +5.4] respectively); statistical significance was only observed for CTX ( P  = 0.02, P1NP, P  = 0.17). CONCLUSION: Contrary to our hypothesis, lumbar spine and total hip regional bone formation (SUV mean ) and BMD did not differ postswitch to TAF. However, improved LS-BMD and CTX echo other TAF-switch studies. The lack of difference in SUV mean may be due to inadequate power.

A review of PET attenuation correction methods for PET-MR.

Despite being thirteen years since the installation of the first PET-MR system, the scanners constitute a very small proportion of the total hybrid PET systems installed. This is in stark contrast to the rapid expansion of the PET-CT scanner, which quickly established its importance in patient diagnosis within a similar timeframe. One of the main hurdles is the development of an accurate, reproducible and easy-to-use method for attenuation correction. Quantitative discrepancies in PET images between the manufacturer-provided MR methods and the more established CT- or transmission-based attenuation correction methods have led the scientific community in a continuous effort to develop a robust and accurate alternative. These can be divided into four broad categories: (i) MR-based, (ii) emission-based, (iii) atlas-based and the (iv) machine learning-based attenuation correction, which is rapidly gaining momentum. The first is based on segmenting the MR images in various tissues and allocating a predefined attenuation coefficient for each tissue. Emission-based attenuation correction methods aim in utilising the PET emission data by simultaneously reconstructing the radioactivity distribution and the attenuation image. Atlas-based attenuation correction methods aim to predict a CT or transmission image given an MR image of a new patient, by using databases containing CT or transmission images from the general population. Finally, in machine learning methods, a model that could predict the required image given the acquired MR or non-attenuation-corrected PET image is developed by exploiting the underlying features of the images. Deep learning methods are the dominant approach in this category. Compared to the more traditional machine learning, which uses structured data for building a model, deep learning makes direct use of the acquired images to identify underlying features. This up-to-date review goes through the literature of attenuation correction approaches in PET-MR after categorising them. The various approaches in each category are described and discussed. After exploring each category separately, a general overview is given of the current status and potential future approaches along with a comparison of the four outlined categories.

Artificial Intelligence for the Characterization of Pulmonary Nodules, Lung Tumors and Mediastinal Nodes on PET/CT.

Lung cancer is the leading cause of cancer related death around the world although early diagnosis remains vital to enabling access to curative treatment options. This article briefly describes the current role of imaging, in particular 2-deoxy-2-[18F]fluoro-D-glucose (FDG) PET/CT, in lung cancer and specifically the role of artificial intelligence with CT followed by a detailed review of the published studies applying artificial intelligence (ie, machine learning and deep learning), on FDG PET or combined PET/CT images with the purpose of early detection and diagnosis of pulmonary nodules, and characterization of lung tumors and mediastinal lymph nodes. A comprehensive search was performed on Pubmed, Embase, and clinical trial databases. The studies were analyzed with a modified version of the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) and Prediction model Risk Of Bias Assessment Tool (PROBAST) statement. The search resulted in 361 studies; of these 29 were included; all retrospective; none were clinical trials. Twenty-two records evaluated standard machine learning (ML) methods on imaging features (ie, support vector machine), and 7 studies evaluated new ML methods (ie, deep learning) applied directly on PET or PET/CT images. The studies mainly reported positive results regarding the use of ML methods for diagnosing pulmonary nodules, characterizing lung tumors and mediastinal lymph nodes. However, 22 of the 29 studies were lacking a relevant comparator and/or lacking independent testing of the model. Application of ML methods with feature and image input from PET/CT for diagnosing and characterizing lung cancer is a relatively young area of research with great promise. Nevertheless, current published studies are often under-powered and lacking a clinically relevant comparator and/or independent testing.

Standardisation of conventional and advanced iterative reconstruction methods for Gallium-68 multi-centre PET-CT trials.

PURPOSE: To assess the applicability of the Fluorine-18 performance specifications defined by EANM Research Ltd (EARL), in Gallium-68 multi-centre PET-CT trials using conventional (ordered subset expectation maximisation, OSEM) and advanced iterative reconstructions which include the systems' point spread function (PSF) and a Bayesian penalised likelihood algorithm (BPL) commercially known as Q.CLEAR. The possibility of standardising the two advanced reconstruction methods was examined. METHODS: The NEMA image quality phantom was filled with Gallium-68 and scanned on a GE PET-CT system. PSF and BPL with varying post-reconstruction Gaussian filter width (2-6.4 mm) and penalisation factor (200-1200), respectively, were applied. The average peak-to-valley ratio from six profiles across each sphere was estimated to inspect any edge artefacts. Image noise was assessed using background variability and image roughness. Six GE and Siemens PET-CT scanners provided Gallium-68 images of the NEMA phantom using both conventional and advanced reconstructions from which the maximum, mean and peak recoveries were drawn. Fourteen patients underwent 68Ga-PSMA PET-CT imaging. BPL (200-1200) reconstructions of the data were compared against PSF smoothed with a 6.4-mm Gaussian filter. RESULTS: A Gaussian filter width of approximately 6 mm for PSF and a penalisation factor of 800 for BPL were needed to suppress the edge artefacts. In addition, those reconstructions provided the closest agreement between the two advanced iterative reconstructions and low noise levels with the background variability and the image roughness being lower than 7.5% and 11.5%, respectively. The recoveries for all methods generally performed at the lower limits of the EARL specifications, especially for the 13- and 10-mm spheres for which up to 27% (conventional) and 41% (advanced reconstructions) lower limits are suggested. The lesion standardised uptake values from the clinical data were significantly different between BPL and PSF smoothed with a Gaussian filter of 6.4 mm wide for all penalisation factors except for 800 and 1000. CONCLUSION: It is possible to standardise the advanced reconstruction methods with the reconstruction parameters being also sufficient for minimising the edge artefacts and noise in the images. For both conventional and advanced reconstructions, Gallium-68 specific recovery coefficient limits were required, especially for the smallest phantom spheres.

Robustness and Generalizability of Deep Learning Synthetic Computed Tomography for Positron Emission Tomography/Magnetic Resonance Imaging-Based Radiation Therapy Planning of Patients With Head and Neck Cancer.

PURPOSE: Radiotherapy planning based only on positron emission tomography/magnetic resonance imaging (PET/MRI) lacks computed tomography (CT) information required for dose calculations. In this study, a previously developed deep learning model for creating synthetic CT (sCT) from MRI in patients with head and neck cancer was evaluated in 2 scenarios: (1) using an independent external dataset, and (2) using a local dataset after an update of the model related to scanner software-induced changes to the input MRI. METHODS AND MATERIALS: Six patients from an external site and 17 patients from a local cohort were analyzed separately. Each patient underwent a CT and a PET/MRI with a Dixon MRI sequence over either one (external) or 2 (local) bed positions. For the external cohort, a previously developed deep learning model for deriving sCT from Dixon MRI was directly applied. For the local cohort, we adapted the model for an upgraded MRI acquisition using transfer learning and evaluated it in a leave-one-out process. The sCT mean absolute error for each patient was assessed. Radiotherapy dose plans based on sCT and CT were compared by assessing relevant absorbed dose differences in target volumes and organs at risk. RESULTS: The MAEs were 78 ± 13 HU and 76 ± 12 HU for the external and local cohort, respectively. For the external cohort, absorbed dose differences in target volumes were within ± 2.3% and within ± 1% in 95% of the cases. Differences in organs at risk were <2%. Similar results were obtained for the local cohort. CONCLUSIONS: We have demonstrated a robust performance of a deep learning model for deriving sCT from MRI when applied to an independent external dataset. We updated the model to accommodate a larger axial field of view and software-induced changes to the input MRI. In both scenarios dose calculations based on sCT were similar to those of CT suggesting a robust and reliable method.

Experimental validation of estimated spatially variant radioisotope-specific point spread functions using published positron range simulations and fluorine-18 measurements.

In this work we compare spatially variant radioisotope-specific point spread functions (PSFs) derived from published positron range data with measured data using a high resolution research tomograph (HRRT). Spatially variant PSFs were measured on a HRRT for fluorine-18, carbon-11 and gallium-68 using an array of printed point sources. For gallium-68, this required modification of the original design to handle its longer positron range. Using the fluorine-18 measurements and previously published data from Monte-Carlo simulations of positron range, estimated PSFs for carbon-11 and gallium-68 were calculated and compared with experimental data. A double 3D Gaussian function was fitted to the estimated and measured data and used to model the spatially varying PSFs over the scanner field of view (FOV). Differences between the measured and estimated PSFs were quantified using the full-width-at-half-maximum (FWHM) and full-width-at-tenth-maximum (FWTM) in the tangential, radial and axial directions. While estimated PSFs were generally in agreement with the measured PSFs over the entire FOV better agreement was observed (FWHM and FWTM differences of less than 10%) when using one of the two sets of positron range simulations, especially for gallium-68 and for the FWTM. Spatially variant radioisotope specific PSFs can be accurately estimated from fluorine-18 measurements and published positron range data. We have experimentally validated this approach for carbon-11 and gallium-68, and such an approach may be applicable to other radioisotopes such as oxygen-15 for which measurements are not practical.