Histopathologic Changes after Yttrium-90 Radioembolization of Colorectal Liver Metastases: A Pilot Feasibility Study.

PURPOSE: To evaluate the histopathologic changes and potential correlations of tumor absorbed dose (TAD) after yttrium-90 transarterial radioembolization (TARE) for colorectal liver metastases (CLMs). MATERIALS AND METHODS: This prospective pilot study assessed 12 patients with 13 CLMs through positron emission tomography (PET)/computed tomography (CT)-guided biopsies before, immediately after TARE (T0), and 3 weeks after TARE (T3). Subsequent sampling from the same location was enabled by fiducial placement. Biopsy samples were evaluated with hematoxylin and eosin, TUNEL, Ki67, OxPhos, caspase-3 (CC3), and pH2AX antibodies. Proliferation changes (Ki67) and double-strand DNA breaks (DSBs) were evaluated quantitatively. TAD was calculated on post-TARE PET/CT scan of the biopsy needle location at T0 and T3. RESULTS: Median TAD at 3 weeks after TARE was 162 Gy (interquartile range (IQR), 92-211 Gy). DSBs decreased significantly from T0 (median, 77%; IQR, 75%-100%) to T3 (median, 14%; IQR, 0%-54%; P = .03). A decrease in Ki67 was also documented (median, 73%; IQR, 70%-80% at T0 vs median, 41%; IQR, 0%-66% at T3; P = .05). There was a strong positive correlation between TAD and DSBs at T0 (r[9] = 0.68) and a strong negative correlation at T3 (r[10] = -0.855; P = .042 and P = .002, respectively). There was a strong negative correlation between TAD and Ki67 at both T0 (r[9] = -0.733; P = .025) and T3 (r[10] = -0.681; P = .03). Tumors that exhibited caspase-3 activation (8/13, 62%) at either T0 or T3 time point were more likely to develop progression (7/8 [88%] vs 1/5 [20%]; P = .015). CONCLUSIONS: Post-TARE biopsy can be used to assess TAD and histopathologic changes. Significant decreases in DSBs and proliferation index were noted after TARE. Post-TARE CC3 activation deserves further exploration.

An automated technique for global noise level measurement in CT image with a conjunction of image gradient.

Automated assessment of noise level in clinical computed tomography (CT) images is a crucial technique for evaluating and ensuring the quality of these images. There are various factors that can impact CT image noise, such as statistical noise, electronic noise, structure noise, texture noise, artifact noise, etc. In this study, a method was developed to measure the global noise index (GNI) in clinical CT scans due to the fluctuation of x-ray quanta. Initially, a noise map is generated by sliding a 10 × 10 pixel for calculating Hounsfield unit (HU) standard deviation and the noise map is further combined with the gradient magnitude map. By employing Boolean operation, pixels with high gradients are excluded from the noise histogram generated with the noise map. By comparing the shape of the noise histogram from this method with Christianson's tissue-type global noise measurement algorithm, it was observed that the noise histogram computed in anthropomorphic phantoms had a similar shape with a close GNI value. In patient CT images, excluding the HU deviation due the structure change demonstrated to have consistent GNI values across the entire CT scan range with high heterogeneous tissue compared to the GNI values using Christianson's tissue-type method. The proposed GNI was evaluated in phantom scans and was found to be capable of comparing scan protocols between different scanners. The variation of GNI when using different reconstruction kernels in clinical CT images demonstrated a similar relationship between noise level and kernel sharpness as observed in uniform phantom: sharper kernel resulted in noisier images. This indicated that GNI was a suitable index for estimating the noise level in clinical CT images with either a smooth or grainy appearance. The study's results suggested that the algorithm can be effectively utilized to screen the noise level for a better CT image quality control.

First-in-human imaging with [89Zr]Zr-DFO-SC16.56 anti-DLL3 antibody in patients with high-grade neuroendocrine tumors of the lung and prostate.

Background: Delta-like ligand 3 (DLL3) is aberrantly expressed on the cell surface in many neuroendocrine cancers including small cell lung cancer (SCLC) and neuroendocrine prostate cancer (NEPC). Several therapeutic agents targeting DLL3 are in active clinical development. Molecular imaging of DLL3 would enable non-invasive diagnostic assessment to inform the use of DLL3-targeting therapeutics or to assess disease treatment response. Methods: We conducted a first-in-human immuno-positron emission tomography (immunoPET) imaging study of [89Zr]Zr-DFO-SC16.56, composed of the anti-DLL3 antibody SC16.56 conjugated to desferrioxamine (DFO) and the positron-emitting radionuclide zirconium-89, in 18 patients with neuroendocrine cancers. An initial cohort of three patients received 1-2 mCi of [89Zr]Zr-DFO-SC16.56 at a total mass dose of 2·5 mg and underwent serial PET and computed tomography (CT) imaging over the course of one week. Radiotracer clearance, tumor uptake, and radiation dosimetry were estimated. An expansion cohort of 15 additional patients were imaged using the initial activity and mass dose. Retrospectively collected tumor biopsies were assessed for DLL3 by immunohistochemistry (IHC) (n = 16). Findings: Imaging of the initial 3 SCLC patients demonstrated strong tumor-specific uptake of [89Zr]Zr-DFO-SC16.56, with similar tumor: background ratios at days 3, 4, and 7 post-injection. Serum clearance was bi-phasic with an estimated terminal clearance half-time of 119 h. The sites of highest background tracer uptake were blood pool and liver. The normal tissue receiving the highest radiation dose was liver; 1·8 mGy/MBq, and the effective dose was 0.49 mSv/MBq. Tumoral uptake varied both between and within patients, and across anatomic sites, with a wide range in SUVmax (from 3·3 to 66·7). Tumor uptake by [89Zr]Zr-DFO-SC16.56 was associated with protein expression in all cases. Two non-avid DLL3 NEPC cases by PET scanning demonstrated the lowest DLL3 expression by tumor immunohistochemistry. Only one patient had a grade 1 allergic reaction, while no grade ≥2 adverse events noted. Interpretation: DLL3 PET imaging of patients with neuroendocrine cancers is safe and feasible. These results demonstrate the potential utility of [89Zr]Zr-DFO-SC16.56 for non-invasive in vivo detection of DLL3-expressing malignancies. Funding: Supported by NIH R01CA213448 (JTP), R35 CA263816 (CMR), U24 CA213274 (CMR), R35 CA232130 (JSL), and a Prostate Cancer Foundation TACTICAL Award (JSL), Scannell foundation. The Radiochemistry and Molecular Imaging Probes Core Facility is supported by NIH P30 CA08748.

Three-Dimensional Margin as a Predictor of Local Tumor Progression after Microwave Ablation: Intraprocedural versus 4-8-Week Postablation Assessment.

PURPOSE: To evaluate the prognostic accuracy of intraprocedural and 4-8-week (current standard) post-microwave ablation zone (AZ) and margin assessments for prediction of local tumor progression (LTP) using 3-dimensional (3D) software. MATERIALS AND METHODS: Data regarding 100 colorectal liver metastases (CLMs) in 75 patients were collected from 2 prospective fluorodeoxyglucose positron emission tomography (PET)/computed tomography (CT)-guided microwave ablation (MWA) trials. The target CLMs and theoretical 5- and 10-mm margins were segmented and registered intraprocedurally and at 4-8 weeks after MWA contrast-enhanced CT (or magnetic resonance [MR] imaging) using the same methodology and 3D software. Tumor and 5- and 10-mm minimal margin (MM) volumes not covered by the AZ were defined as volumes of insufficient coverage (VICs). The intraprocedural and 4-8-week post-MWA VICs were compared as predictors of LTP using receiver operating characteristic curve analysis. RESULTS: The median follow-up time was 19.6 months (interquartile range, 7.97-36.5 months). VICs for 5- and 10-mm MMs were predictive of LTP at both time assessments. The highest accuracy for the prediction of LTP was documented with the intra-ablation 5-mm VIC (area under the curve [AUC], 0.78; 95% confidence interval, 0.66-0.89). LTP for a VIC of 6-10-mm margin category was 11.4% compared with 4.3% for >10-mm margin category (P < .001). CONCLUSIONS: A 3D 5-mm MM is a critical endpoint of thermal ablation, whereas optimal local tumor control is noted with a 10-mm MM. Higher AUCs for prediction of LTP were achieved for intraprocedural evaluation than for the 4-8-week postablation 3D evaluation of the AZ.

Technical Note: Impact of impurities on Yttrium-90 glass microsphere activity quantitation.

BACKGROUND: Glass 90 Y microspheres are produced with known radionuclide impurities. These impurities are not independently monitored. Clinical instruments, including ionization chamber dose calibrators and positron emmission tomography (PET) cameras, can be much more sensitive in detecting signals from these impurities than to signals from 90 Y itself. PURPOSE: The "typical" levels of 90 Y impurities have been studied to assess their impact on dosimetry during internal implantation, and for the management of waste. However, unaccounted-for decay spectra of impurities can also have an impact on dose calibrator and PET readings. Thus, even what might be considered negligibly small impurity fractions, can in principle cause substantial overestimates of the amount of 90 Y activity present in a sample. To our knowledge, quantitative effects of radionuclide impurities in glass microspheres on activity measurements have not been documented in the field. As activity quantitation for dosimetry and its correlations with outcome becomes more prevalent, the effects of impurities on measurements may remain unaccounted for in dosimetry studies. METHODS: In this letter, we review theoretical and physical considerations that will result in asymmetric errors in quantitation from 90 Y impurities and estimate their typical and potential impact on clinical utilization. Among the common impurities 88 Y is of particular concern for its impact on 90 Y dose measurements because of its decay characteristics, along with other isotopes 91 Y and 46 Sc which can also impact measurements. RESULTS: The typical level of 88 Y impurities reported by the manufacturer should only cause small errors in dose calibrator and PET measurements made within the 12-day label-specified use-by period, up to 2.0% and 1.6%, respectively. However, the product specification max allowable impurity levels, specified by the manufacturer, leave open the potential for much greater bias from within the 12-day use-by period, potentially as high as 13.2% for dose calibrator measurements and 10.6% for PET from the 88 Y impurities. CONCLUSIONS: While typical levels of impurities appear to have acceptable impact on patient absorbed dose, it should be noted that they can have adverse effects on 90 Y radioactivity measurements. Furthermore, there is currently minimal independent verification and/or monitoring of impurity levels within the field.

Gradient-based Volumetric PET Parameters on Immediate Pre-ablation FDG-PET Predict Local Tumor Progression in Patients with Colorectal Liver Metastasis Treated by Microwave Ablation.

PURPOSE: This study aimed to evaluate the optimal method of segmentation of colorectal liver metastasis (CLM) on immediate pre-ablation PET scans and assess the prognostic value of quantitative pre-ablation PET parameters with regards to local tumor control. A secondary objective was to correlate the target tumor size estimation by PET methods with the tumor measurements on anatomical imaging. METHODOLOGY: A prospectively accrued cohort of 55 CLMs (46 patients) treated with real-time 18F-FDG-PET/CT-guided percutaneous microwave ablation was followed-up for a median of 10.8 months (interquartile: 5.5-20.2). Total lesion glycolysis (TLG) and metabolic tumor volume (MTV) values of each CLM were derived from pre-ablation 18F-FDG-PET with gradient and threshold PET segmentation methodologies. The event was defined as local tumor progression (LTP). Time-dependent receiver operating characteristic (ROC) curve analyses were used to assess area under the curves (AUCs). Intraclass correlation (ICC) and 95.0% confidence interval (CI) were performed to measure the linear relationships between the continuous variables. RESULTS: AUCs for prediction of LTP obtained from time-dependent ROC analysis for the gradient technique were higher in comparison to the threshold methodologies (AUCs for TLG and volume were: 0.790 and 0.807, respectively). ICC between PET gradient-based and anatomical measurements were higher in comparison to threshold methodologies (ICC for the longest diameter: 733 (95.0% CI 0.538-0.846), ICC for the shortest diameter: .747 (95.0% CI 0.546-0.859), p-values < 0.001). CONCLUSIONS: The gradient-based technique had a higher AUC for prediction of LTP after microwave ablation of CLM and showed the highest correlation with anatomical imaging tumor measurements.

Yttrium-90 Activity Quantification in PET/CT-Guided Biopsy Specimens from Colorectal Hepatic Metastases Immediately after Transarterial Radioembolization Using Micro-CT and Autoradiography.

PURPOSE: To evaluate the yttrium-90 (90Y) activity distribution in biopsy tissue samples of the treated liver to quantify the dose with higher spatial resolution than positron emission tomography (PET) for accurate investigation of correlations with microscopic biological effects and to evaluate the radiation safety of this procedure. MATERIALS AND METHODS: Eighty-six core biopsy specimens were obtained from 18 colorectal liver metastases (CLMs) immediately after 90Y transarterial radioembolization (TARE) with either resin or glass microspheres using real-time 90Y PET/CT guidance in 17 patients. A high-resolution micro-computed tomography (micro-CT) scanner was used to image the microspheres in part of the specimens and allow quantification of 90Y activity directly or by calibrating autoradiography (ARG) images. The mean doses to the specimens were derived from the measured specimens' activity concentrations and from the PET/CT scan at the location of the biopsy needle tip for all cases. Staff exposures were monitored. RESULTS: The mean measured 90Y activity concentration in the CLM specimens at time of infusion was 2.4 ± 4.0 MBq/mL. The biopsies revealed higher activity heterogeneity than PET. Radiation exposure to the interventional radiologists during post-TARE biopsy procedures was minimal. CONCLUSIONS: Counting the microspheres and measuring the activity in biopsy specimens obtained after TARE are safe and feasible and can be used to determine the administered activity and its distribution in the treated and biopsied liver tissue with high spatial resolution. Complementing 90Y PET/CT imaging with this approach promises to yield more accurate direct correlation of histopathological changes and absorbed dose in the examined specimens.

Real-Time Split-Dose PET/CT-Guided Ablation Improves Colorectal Liver Metastasis Detection and Ablation Zone Margin Assessments without the Need for Repeated Contrast Injection.

BACKGROUND: Real-time split-dose PET can identify the targeted colorectal liver metastasis (CLM) and eliminate the need for repeated contrast administration before and during thermal ablation (TA). This study aimed to assess the added value of pre-ablation real-time split-dose PET when combined with non-contract CT in the detection of CLM for ablation and the evaluation of the ablation zone and margins. METHODS: A total of 190 CLMs/125 participants from two IRB-approved prospective clinical trials using PET/CT-guided TA were analyzed. Based on detection on pre-TA imaging, CLMs were categorized as detectable, non-detectable, and of poor conspicuity on CT alone, and detectable, non-detectable, and low FDG-avidity on PET/CT after the initial dose. Ablation margins around the targeted CLM were evaluated using a 3D volumetric approach. RESULTS: We found that 129/190 (67.9%) CLMs were detectable on CT alone, and 61/190 CLMs (32.1%) were undetectable or of poor conspicuity, not allowing accurate depiction and targeting by CT alone. Thus, the theoretical 5- and 10-mm margins could not be defined in these tumors (32.1%) using CT alone. When TA intraprocedural PET/CT images are obtained and inspected (fused PET/CT), only 4 CLM (2.1%) remained undetectable or had a low FDG avidity. CONCLUSIONS: The addition of PET to non-contrast CT improved CLM detection for ablation targeting, margin assessments, and continuous depiction of the FDG avid CLMs during the ablation without the need for multiple intravenous contrast injections pre- and intra-procedurally.

TriDFusion (3DF) image viewer.

BACKGROUND: An open-source, extensible medical viewing platform is described, called the TriDFusion image viewer (3DF). The 3DF addresses many broad unmet needs in nuclear medicine research; it provides a viewer with several tools not available in commercial nuclear medicine workstations, yet invaluable for imaging in research studies. RESULTS: The 3DF includes an image integration platform to register images from multiple imaging modalities together with delineated volumes of interest (VOIs), structures and dose distributions. It can process images from different vendors' systems and is therefore vendor neutral. The 3DF also provides a convenient tool for performing multi-modality image analysis and fusion. The functional components currently being distributed is open-source code that includes: (1) a high quality viewer that can display axial, coronal, and sagittal tomographic images, maximum intensity projection images, structure contours, and isointensity contour lines or dose colorwash, (2) multi-image fusion allowing multiple images to be fused with VOI and dose distributions, (3) a suite of segmentation tools to edit and/or create tumor and organ VOIs, (4) dosimetry tools for several radioisotopes, (5) clinical tools for correcting acquisition errors, including patient orientation, and (6) the ability to save the resultant image and VOI as DICOM files or to export the numerical results as comma separated values files. Because the code is written in MATLAB™, it is highly readable and is easier for the coder to make changes compared to languages such as C or C++. In what follows, we describe the content of the new TriDFusion (3DF) image viewer software platform using examples of a number of clinical research workflows. Such examples vary in complexity but illustrate the main attributes of the software. CONCLUSIONS: In summary, 3DF provides a powerful, convenient, easy-to-use suite of open-source imaging research tools for the nuclear medicine community that allows physicians, medical physicists, and academic researchers to display, manipulate, and analyze images.

The OpenGATE ecosystem for Monte Carlo simulation in medical physics.

This paper reviews the ecosystem of GATE, an open-source Monte Carlo toolkit for medical physics. Based on the shoulders of Geant4, the principal modules (geometry, physics, scorers) are described with brief descriptions of some key concepts (Volume, Actors, Digitizer). The main source code repositories are detailed together with the automated compilation and tests processes (Continuous Integration). We then described how the OpenGATE collaboration managed the collaborative development of about one hundred developers during almost 20 years. The impact of GATE on medical physics and cancer research is then summarized, and examples of a few key applications are given. Finally, future development perspectives are indicated.

3D margin assessment predicts local tumor progression after ablation of colorectal cancer liver metastases.

OBJECTIVE: To determine the feasibility and prognostic value of 3D measuring of the ablation margins using a dedicated image registration software. METHODS: This retrospective study included 104 colorectal liver metastases in 68 consecutive patients that underwent microwave ablation between 08/2012 and 08/2019. The minimal ablation margin (MM) was measured in 2D using anatomic landmarks on contrast enhanced CT(CECT) 4-8 weeks post-ablation, and in 3D using an image registration software and immediate post-ablation CECT. Local tumor progression (LTP) was assessed by imaging up to 24 months after ablation. A blinded interventional radiologist provided feedback on the possibility of additional ablation after examining the 3D-margin measurements. RESULTS: The 3D-margin assessment was completed in 79/104 (76%) tumors without the need for target manipulation. In 25/104 (24%) tumors, manipulation was required due to image misregistration. LTP was observed in 40/104 (38.5%) tumors: 92.5% vs 7.5% for those with margin <5mm vs ≥5mm, respectively (p = 0.0001). The 2D and 3D-assessments identified margin <5mm in 17/104 (16%), and in 74/104 (71%) ablated tumors, respectively (p < 0.01). The sensitivity and specificity of the 3D software for predicting LTP was 93% (37/40) and 42% (27/64), respectively. Additional ablation to achieve a MM of 5 mm would have been offered in 26/37 cases if the 3D-margin assessment was available intraoperatively. CONCLUSION: Image registration software can measure ablation margins and detect MM under 5 mm intraoperatively, with significantly higher sensitivity than the 2D technique using landmarks on the post-ablation CECT. The identification of a margin under 5 mm is strongly associated with LTP.

Advanced Monte Carlo simulations of emission tomography imaging systems with GATE.

Built on top of the Geant4 toolkit, GATE is collaboratively developed for more than 15 years to design Monte Carlo simulations of nuclear-based imaging systems. It is, in particular, used by researchers and industrials to design, optimize, understand and create innovative emission tomography systems. In this paper, we reviewed the recent developments that have been proposed to simulate modern detectors and provide a comprehensive report on imaging systems that have been simulated and evaluated in GATE. Additionally, some methodological developments that are not specific for imaging but that can improve detector modeling and provide computation time gains, such as Variance Reduction Techniques and Artificial Intelligence integration, are described and discussed.

Practice and prospects for PET/CT guided interventions.

During the past 10 years, performing real-time molecular imaging with positron emission tomography (PET) in combination with computed tomography (CT) during interventional procedures has undergone rapid development. Keeping in mind the interest of the nuclear medicine readers, an update is provided of the current workflows using real-time PET/CT in percutaneous biopsies and tumor ablations. The clinical utility of PET/CT guided biopsies in cancer patients with lung, liver, lymphoma, and bone tumors are reviewed. Several technological developments, including the introduction of new PET tracers and robotic arms as well as opportunities provided through acquiring radioactive biopsy specimens are briefly reviewed.

CD38-targeted Immuno-PET of Multiple Myeloma: From Xenograft Models to First-in-Human Imaging.

Background Current measurements of multiple myeloma disease burden are suboptimal. Daratumumab is a monoclonal antibody that targets CD38, an antigen expressed on nearly all myeloma cells. Purpose To demonstrate preclinical and first-in-human application of an antibody composed of the native daratumumab labeled with the positron-emitting radionuclide zirconium 89 (89Zr) through the chelator deferoxamine (DFO), or 89Zr-DFO-daratumumab, for immunologic PET imaging of multiple myeloma. Materials and Methods 89Zr-DFO-daratumumab was synthesized by conjugating 89Zr to daratumumab with DFO. A murine xenograft model using CD38-positive OPM2 multiple myeloma cells was used to evaluate CD38-specificity of 89Zr-DFO-daratumumab. Following successful preclinical imaging, a prospective phase I study of 10 patients with multiple myeloma was performed. Study participants received 74 MBq (2 mCi) of intravenous 89Zr-DFO-daratumumab. Each participant underwent four PET/CT scans over the next 8 days, as well as blood chemistry and whole-body counts, to determine safety, tracer biodistribution, pharmacokinetics, and radiation dosimetry. Because 89Zr has a half-life of 78 hours, only a single administration of tracer was needed to obtain all four PET/CT scans. Results 89Zr-DFO-daratumumab was synthesized with radiochemical purity greater than 99%. In the murine model, substantial bone marrow uptake was seen in OPM2 mice but not in healthy mice, consistent with CD38-targeted imaging of OPM2 multiple myeloma cells. In humans, 89Zr-DFO-daratumumab was safe and demonstrated acceptable dosimetry. 89Zr-DFO-daratumumab uptake was visualized at PET in sites of osseous myeloma. Conclusion These data demonstrate successful CD38-targeted immunologic PET imaging of multiple myeloma in a murine model and in humans. © RSNA, 2020 Online supplemental material is available for this article.

Technical Note: Scintillation well counters and particle counting digital autoradiography devices can be used to detect activities associated with genomic profiling adequacy of biopsy specimens obtained after a low activity 18 F-FDG injection.

PURPOSE: Genomic profiling of biopsied tissue is the basis for precision cancer therapy. However, biopsied materials may not contain sufficient amounts of tumor deoxyribonucleonic acid needed for the analysis. We propose a method to determine the adequacy of specimens for performing genomic profiling by quantifying their metabolic activity. METHODS: We estimated the average density of tumor cells in biopsy specimens needed to successfully perform genomic analysis following the Memorial Sloan Kettering Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) protocol from the minimum amount of deoxyribonucleonic acid needed and the volume of tissue typically used for analysis. The average 18 F-FDG uptake per cell was assessed by incubating HT-29 adenocarcinoma tumor cells in 18 F-FDG containing solution and then measuring their activity with a scintillation well counter. Consequently, we evaluated the response of two devices around the minimum expected activities which would indicate genomic profiling adequacy of biopsy specimens obtained under 18 F-FDG PET/CT guidance. Surrogate samples obtained using 18G core needle biopsies of gels containing either 18 F-FDG-loaded cells in the expected concentrations or the corresponding activity were measured using autoradiography and a scintillation well counter. Autoradiography was performed using a CCD-based device with real-time image display as well as with digital autoradiography imaging plates following a 30-min off-line protocol for specimen activity determination against previously established calibration. RESULTS: Cell incubation experiments and estimates obtained from quantitative autoradiography of biopsy specimens (QABS) indicate that specimens acquired under 18 F-FDG PET/CT guidance that contained the minimum amount of cells needed for genomic profiling would have an average activity concentration in the range of about 3 to about 9 kBq/mL. When exposed to specimens with similar activity concentration, both a CCD-based autoradiography device and a scintillation well counter produced signals with sufficient signal-to-background ratio for specimen genomic adequacy identification in less than 10 min, which is short enough to allow procedure guidance. CONCLUSION: Scintillation well counter measurements and CCD-based autoradiography have adequate sensitivity to detect the tumor burden needed for genomic profiling during 18 F-FDG PET/CT-guided 18G core needle biopsies of liver adenocarcinoma metastases.

The first MICCAI challenge on PET tumor segmentation.

INTRODUCTION: Automatic functional volume segmentation in PET images is a challenge that has been addressed using a large array of methods. A major limitation for the field has been the lack of a benchmark dataset that would allow direct comparison of the results in the various publications. In the present work, we describe a comparison of recent methods on a large dataset following recommendations by the American Association of Physicists in Medicine (AAPM) task group (TG) 211, which was carried out within a MICCAI (Medical Image Computing and Computer Assisted Intervention) challenge. MATERIALS AND METHODS: Organization and funding was provided by France Life Imaging (FLI). A dataset of 176 images combining simulated, phantom and clinical images was assembled. A website allowed the participants to register and download training data (n = 19). Challengers then submitted encapsulated pipelines on an online platform that autonomously ran the algorithms on the testing data (n = 157) and evaluated the results. The methods were ranked according to the arithmetic mean of sensitivity and positive predictive value. RESULTS: Sixteen teams registered but only four provided manuscripts and pipeline(s) for a total of 10 methods. In addition, results using two thresholds and the Fuzzy Locally Adaptive Bayesian (FLAB) were generated. All competing methods except one performed with median accuracy above 0.8. The method with the highest score was the convolutional neural network-based segmentation, which significantly outperformed 9 out of 12 of the other methods, but not the improved K-Means, Gaussian Model Mixture and Fuzzy C-Means methods. CONCLUSION: The most rigorous comparative study of PET segmentation algorithms to date was carried out using a dataset that is the largest used in such studies so far. The hierarchy amongst the methods in terms of accuracy did not depend strongly on the subset of datasets or the metrics (or combination of metrics). All the methods submitted by the challengers except one demonstrated good performance with median accuracy scores above 0.8.

Toward a standard for the evaluation of PET-Auto-Segmentation methods following the recommendations of AAPM task group No. 211: Requirements and implementation.

PURPOSE: The aim of this paper is to define the requirements and describe the design and implementation of a standard benchmark tool for evaluation and validation of PET-auto-segmentation (PET-AS) algorithms. This work follows the recommendations of Task Group 211 (TG211) appointed by the American Association of Physicists in Medicine (AAPM). METHODS: The recommendations published in the AAPM TG211 report were used to derive a set of required features and to guide the design and structure of a benchmarking software tool. These items included the selection of appropriate representative data and reference contours obtained from established approaches and the description of available metrics. The benchmark was designed in a way that it could be extendable by inclusion of bespoke segmentation methods, while maintaining its main purpose of being a standard testing platform for newly developed PET-AS methods. An example of implementation of the proposed framework, named PETASset, was built. In this work, a selection of PET-AS methods representing common approaches to PET image segmentation was evaluated within PETASset for the purpose of testing and demonstrating the capabilities of the software as a benchmark platform. RESULTS: A selection of clinical, physical, and simulated phantom data, including "best estimates" reference contours from macroscopic specimens, simulation template, and CT scans was built into the PETASset application database. Specific metrics such as Dice Similarity Coefficient (DSC), Positive Predictive Value (PPV), and Sensitivity (S), were included to allow the user to compare the results of any given PET-AS algorithm to the reference contours. In addition, a tool to generate structured reports on the evaluation of the performance of PET-AS algorithms against the reference contours was built. The variation of the metric agreement values with the reference contours across the PET-AS methods evaluated for demonstration were between 0.51 and 0.83, 0.44 and 0.86, and 0.61 and 1.00 for DSC, PPV, and the S metric, respectively. Examples of agreement limits were provided to show how the software could be used to evaluate a new algorithm against the existing state-of-the art. CONCLUSIONS: PETASset provides a platform that allows standardizing the evaluation and comparison of different PET-AS methods on a wide range of PET datasets. The developed platform will be available to users willing to evaluate their PET-AS methods and contribute with more evaluation datasets.

Evaluation of the tumor registration error in biopsy procedures performed under real-time PET/CT guidance.

PURPOSE: The purpose of this study is to quantify tumor displacement during real-time PET/CT guided biopsy and to investigate correlations between tumor displacement and false-negative results. METHODS: 19 patients who underwent real-time 18 F-FDG PET-guided biopsy and were found positive for malignancy were included in this study under IRB approval. PET/CT images were acquired for all patients within minutes prior to biopsy to visualize the FDG-avid region and plan the needle insertion. The biopsy needle was inserted and a post-insertion CT scan was acquired. The two CT scans acquired before and after needle insertion were registered using a deformable image registration (DIR) algorithm. The DIR deformation vector field (DVF) was used to calculate the mean displacement between the pre-insertion and post-insertion CT scans for a region around the tip of the biopsy needle. For 12 patients one biopsy core from each was tracked during histopathological testing to investigate correlations of the mean displacement between the two CT scans and false-negative or true-positive biopsy results. For 11 patients, two PET scans were acquired; one at the beginning of the procedure, pre-needle insertion, and an additional one with the needle in place. The pre-insertion PET scan was corrected for intraprocedural motion by applying the DVF. The corrected PET was compared with the post-needle insertion PET to validate the correction method. RESULTS: The mean displacement of tissue around the needle between the pre-biopsy CT and the postneedle insertion CT was 5.1 mm (min = 1.1 mm, max = 10.9 mm and SD = 3.0 mm). For mean displacements larger than 7.2 mm, the biopsy cores gave false-negative results. Correcting pre-biopsy PET using the DVF improved the PET/CT registration in 8 of 11 cases. CONCLUSIONS: The DVF obtained from DIR of the CT scans can be used for evaluation and correction of the error in needle placement with respect to the FDG-avid area. Misregistration between the pre-biopsy PET and the CT acquired with the needle in place was shown to correlate with false negative biopsy results.

Classification and evaluation strategies of auto-segmentation approaches for PET: Report of AAPM task group No. 211.

PURPOSE: The purpose of this educational report is to provide an overview of the present state-of-the-art PET auto-segmentation (PET-AS) algorithms and their respective validation, with an emphasis on providing the user with help in understanding the challenges and pitfalls associated with selecting and implementing a PET-AS algorithm for a particular application. APPROACH: A brief description of the different types of PET-AS algorithms is provided using a classification based on method complexity and type. The advantages and the limitations of the current PET-AS algorithms are highlighted based on current publications and existing comparison studies. A review of the available image datasets and contour evaluation metrics in terms of their applicability for establishing a standardized evaluation of PET-AS algorithms is provided. The performance requirements for the algorithms and their dependence on the application, the radiotracer used and the evaluation criteria are described and discussed. Finally, a procedure for algorithm acceptance and implementation, as well as the complementary role of manual and auto-segmentation are addressed. FINDINGS: A large number of PET-AS algorithms have been developed within the last 20 years. Many of the proposed algorithms are based on either fixed or adaptively selected thresholds. More recently, numerous papers have proposed the use of more advanced image analysis paradigms to perform semi-automated delineation of the PET images. However, the level of algorithm validation is variable and for most published algorithms is either insufficient or inconsistent which prevents recommending a single algorithm. This is compounded by the fact that realistic image configurations with low signal-to-noise ratios (SNR) and heterogeneous tracer distributions have rarely been used. Large variations in the evaluation methods used in the literature point to the need for a standardized evaluation protocol. CONCLUSIONS: Available comparison studies suggest that PET-AS algorithms relying on advanced image analysis paradigms provide generally more accurate segmentation than approaches based on PET activity thresholds, particularly for realistic configurations. However, this may not be the case for simple shape lesions in situations with a narrower range of parameters, where simpler methods may also perform well. Recent algorithms which employ some type of consensus or automatic selection between several PET-AS methods have potential to overcome the limitations of the individual methods when appropriately trained. In either case, accuracy evaluation is required for each different PET scanner and scanning and image reconstruction protocol. For the simpler, less robust approaches, adaptation to scanning conditions, tumor type, and tumor location by optimization of parameters is necessary. The results from the method evaluation stage can be used to estimate the contouring uncertainty. All PET-AS contours should be critically verified by a physician. A standard test, i.e., a benchmark dedicated to evaluating both existing and future PET-AS algorithms needs to be designed, to aid clinicians in evaluating and selecting PET-AS algorithms and to establish performance limits for their acceptance for clinical use. The initial steps toward designing and building such a standard are undertaken by the task group members.