Deep learning based automated delineation of the intraprostatic gross tumour volume in PSMA-PET for patients with primary prostate cancer.

PURPOSE: With the increased use of focal radiation dose escalation for primary prostate cancer (PCa), accurate delineation of gross tumor volume (GTV) in prostate-specific membrane antigen PET (PSMA-PET) becomes crucial. Manual approaches are time-consuming and observer dependent. The purpose of this study was to create a deep learning model for the accurate delineation of the intraprostatic GTV in PSMA-PET. METHODS: A 3D U-Net was trained on 128 different 18F-PSMA-1007 PET images from three different institutions. Testing was done on 52 patients including one independent internal cohort (Freiburg: n = 19) and three independent external cohorts (Dresden: n = 14 18F-PSMA-1007, Boston: Massachusetts General Hospital (MGH): n = 9 18F-DCFPyL-PSMA and Dana-Farber Cancer Institute (DFCI): n = 10 68Ga-PSMA-11). Expert contours were generated in consensus using a validated technique. CNN predictions were compared to expert contours using Dice similarity coefficient (DSC). Co-registered whole-mount histology was used for the internal testing cohort to assess sensitivity/specificity. RESULTS: Median DSCs were Freiburg: 0.82 (IQR: 0.73-0.88), Dresden: 0.71 (IQR: 0.53-0.75), MGH: 0.80 (IQR: 0.64-0.83) and DFCI: 0.80 (IQR: 0.67-0.84), respectively. Median sensitivity for CNN and expert contours were 0.88 (IQR: 0.68-0.97) and 0.85 (IQR: 0.75-0.88) (p = 0.40), respectively. GTV volumes did not differ significantly (p > 0.1 for all comparisons). Median specificity of 0.83 (IQR: 0.57-0.97) and 0.88 (IQR: 0.69-0.98) were observed for CNN and expert contours (p = 0.014), respectively. CNN prediction took 3.81 seconds on average per patient. CONCLUSION: The CNN was trained and tested on internal and external datasets as well as histopathology reference, achieving a fast GTV segmentation for three PSMA-PET tracers with high diagnostic accuracy comparable to manual experts.

Robustness and performance of radiomic features in diagnosing cystic renal masses.

PURPOSE: We study the inter-reader variability in manual delineation of cystic renal masses (CRMs) presented in computerized tomography (CT) images and its effect on the classification performance of a machine learning algorithm in distinguishing benign from potentially malignant CRMs. In addition, we assessed whether the inclusion of higher-order robust radiomic features improves the classification performance over the use of first-order features. METHODS: 230 CRMs were independently delineated by two radiologists. Through a combination of random fluctuations, dilation, and erosion operations over the original region of interests (ROIs), we generated four additional sets of synthetic ROIs to capture the inter-reader variability realistically, as confirmed by dice coefficient measurements and visual assessment. We then identified the robust features based on the intra-class coefficient (ICC > 0.85) across these datasets. We applied a tenfold stratified cross-validation (CV) to train and test the performance of the random forest model for the classification of CRMs into benign and potentially malignant. RESULTS: The mean area under the curve (AUC), sensitivity, specificity, positive predictive value, and negative predictive value were 0.87, 0.82, 0.90, 0.85, and 0.93, respectively. With the usage of first-order features alone, the corresponding values were nearly identical. CONCLUSION: AUC ranged for the robust and uncorrelated features from 0.83 ± 0.09 to 0.93 ± 0.04 and for the first-order features from 0.84 ± 0.09 to 0.91 ± 0.04. Our study indicates that the first-order features alone are sufficient for the classification of CRMs, and that inclusion of higher-order features does not necessarily improve performance.

Retrospective fractional dose reduction in Tc-99m cardiac perfusion SPECT/CT patients: A human and model observer study.

BACKGROUND: In the ongoing efforts to reduce cardiac perfusion dose (injected radioactivity) for conventional SPECT/CT systems, we performed a human observer study to confirm our clinical model observer findings that iterative reconstruction employing OSEM (ordered-subset expectation-maximization) at 25% of the full dose (quarter-dose) has a similar performance for detection of hybrid cardiac perfusion defects as FBP at full dose. METHODS: One hundred and sixty-six patients, who underwent routine rest-stress Tc-99m sestamibi cardiac perfusion SPECT/CT imaging and clinically read as normally perfused, were included in the study. Ground truth was established by the normal read and the insertion of hybrid defects. In addition to the reconstruction of the 25% of full-dose data using OSEM with attenuation (AC), scatter (SC), and spatial resolution correction (RC), FBP and OSEM (with AC, SC, and RC) both at full dose (100%) were done. Both human observer and clinical model observer confidence scores were obtained to generate receiver operating characteristics (ROC) curves in a task-based image quality assessment. RESULTS: Average human observer AUC (area under the ROC curve) values of 0.725, 0.876, and 0.890 were obtained for FBP at full dose, OSEM at 25% of full dose, and OSEM at full dose, respectively. Both OSEM strategies were significantly better than FBP with P values of 0.003 and 0.01 respectively, while no significant difference was recorded between OSEM methods (P = 0.48). The clinical model observer results were 0.791, 0.822, and 0.879, respectively, for the same patient cases and processing strategies used in the human observer study. CONCLUSIONS: Cardiac perfusion SPECT/CT using OSEM reconstruction at 25% of full dose has AUCs larger than FBP and closer to those of full-dose OSEM when read by human observers, potentially replacing the higher dose studies during clinical reading.

Theranostics: The Role of Quantitative Nuclear Medicine Imaging.

Theranostics is a precision medicine discipline that integrates diagnostic nuclear medicine imaging with radionuclide therapy in a manner that provides both a tumor phenotype and personalized therapy to patients with cancer using radiopharmaceuticals aimed at the same target-specific biological pathway or receptor. The aim of quantitative nuclear medicine imaging is to plan the alpha or beta-emitting therapy based on an accurate 3-dimensional representation of the in-vivo distribution of radioactivity concentration within the tumor and normal organs/tissues in a noninvasive manner. In general, imaging may be either based on positron emission tomography (PET) or single photon emission computed tomography (SPECT) invariably in combination with X-ray CT (PET/CT; SPECT/CT) or, to a much lesser extent, MRI. PET and SPECT differ in terms of the radionuclides and physical processes that give rise to the emission of high energy photons, as well as the sets of technologies involved in their detection. Using a variety of standardized quantitative parameters, system calibration, patient preparation, imaging acquisition and reconstruction protocols, and image analysis protocols, an accurate quantification of the tracer distribution can be obtained, which helps prescribe the therapeutic dose for each patient.

The clinical utilities of multi-pinhole single photon emission computed tomography.

Single photon emission computed tomography (SPECT) is an important imaging modality for various applications in nuclear medicine. The use of multi-pinhole (MPH) collimators can provide superior resolution-sensitivity trade-off when imaging small field-of-view compared to conventional parallel-hole and fan-beam collimators. Besides the very successful application in small animal imaging, there has been a resurgence of the use of MPH collimators for clinical cardiac and brain studies, as well as other small field-of-view applications. This article reviews the basic principles of MPH collimators and introduces currently available and proposed clinical MPH SPECT systems.

Comparison of the scanning linear estimator (SLE) and ROI methods for quantitative SPECT imaging.

In quantitative emission tomography, tumor activity is typically estimated from calculations on a region of interest (ROI) identified in the reconstructed slices. In these calculations, unpredictable bias arising from the null functions of the imaging system affects ROI estimates. The magnitude of this bias depends upon the tumor size and location. In prior work it has been shown that the scanning linear estimator (SLE), which operates on the raw projection data, is an unbiased estimator of activity when the size and location of the tumor are known. In this work, we performed analytic simulation of SPECT imaging with a parallel-hole medium-energy collimator. Distance-dependent system spatial resolution and non-uniform attenuation were included in the imaging simulation. We compared the task of activity estimation by the ROI and SLE methods for a range of tumor sizes (diameter: 1-3 cm) and activities (contrast ratio: 1-10) added to uniform and non-uniform liver backgrounds. Using the correct value for the tumor shape and location is an idealized approximation to how task estimation would occur clinically. Thus we determined how perturbing this idealized prior knowledge impacted the performance of both techniques. To implement the SLE for the non-uniform background, we used a novel iterative algorithm for pre-whitening stationary noise within a compact region. Estimation task performance was compared using the ensemble mean-squared error (EMSE) as the criterion. The SLE method performed substantially better than the ROI method (i.e. EMSE(SLE) was 23-174 times lower) when the background is uniform and tumor location and size are known accurately. The variance of the SLE increased when a non-uniform liver texture was introduced but the EMSE(SLE) continued to be 5-20 times lower than the ROI method. In summary, SLE outperformed ROI under almost all conditions that we tested.