Population-based deep image prior for dynamic PET denoising: A data-driven approach to improve parametric quantification.

The high noise level of dynamic Positron Emission Tomography (PET) images degrades the quality of parametric images. In this study, we aim to improve the quality and quantitative accuracy of Ki images by utilizing deep learning techniques to reduce the noise in dynamic PET images. We propose a novel denoising technique, Population-based Deep Image Prior (PDIP), which integrates population-based prior information into the optimization process of Deep Image Prior (DIP). Specifically, the population-based prior image is generated from a supervised denoising model that is trained on a prompts-matched static PET dataset comprising 100 clinical studies. The 3D U-Net architecture is employed for both the supervised model and the following DIP optimization process. We evaluated the efficacy of PDIP for noise reduction in 25%-count and 100%-count dynamic PET images from 23 patients by comparing with two other baseline techniques: the Prompts-matched Supervised model (PS) and a conditional DIP (CDIP) model that employs the mean static PET image as the prior. Both the PS and CDIP models show effective noise reduction but result in smoothing and removal of small lesions. In addition, the utilization of a single static image as the prior in the CDIP model also introduces a similar tracer distribution to the denoised dynamic frames, leading to lower Ki in general as well as incorrect Ki in the descending aorta. By contrast, as the proposed PDIP model utilizes intrinsic image features from the dynamic dataset and a large clinical static dataset, it not only achieves comparable noise reduction as the supervised and CDIP models but also improves lesion Ki predictions.

Generation of Whole-Body FDG Parametric Ki Images from Static PET Images Using Deep Learning.

FDG parametric Ki images show great advantage over static SUV images, due to the higher contrast and better accuracy in tracer uptake rate estimation. In this study, we explored the feasibility of generating synthetic Ki images from static SUV ratio (SUVR) images using three configurations of U-Nets with different sets of input and output image patches, which were the U-Nets with single input and single output (SISO), multiple inputs and single output (MISO), and single input and multiple outputs (SIMO). SUVR images were generated by averaging three 5-min dynamic SUV frames starting at 60 minutes post-injection, and then normalized by the mean SUV values in the blood pool. The corresponding ground truth Ki images were derived using Patlak graphical analysis with input functions from measurement of arterial blood samples. Even though the synthetic Ki values were not quantitatively accurate compared with ground truth, the linear regression analysis of joint histograms in the voxels of body regions showed that the mean R2 values were higher between U-Net prediction and ground truth (0.596, 0.580, 0.576 in SISO, MISO and SIMO), than that between SUVR and ground truth Ki (0.571). In terms of similarity metrics, the synthetic Ki images were closer to the ground truth Ki images (mean SSIM = 0.729, 0.704, 0.704 in SISO, MISO and MISO) than the input SUVR images (mean SSIM = 0.691). Therefore, it is feasible to use deep learning networks to estimate surrogate map of parametric Ki images from static SUVR images.

Federated Transfer Learning for Low-dose PET Denoising: A Pilot Study with Simulated Heterogeneous Data.

Positron emission tomography (PET) with a reduced injection dose, i.e., low-dose PET, is an efficient way to reduce radiation dose. However, low-dose PET reconstruction suffers from a low signal-to-noise ratio (SNR), affecting diagnosis and other PET-related applications. Recently, deep learning-based PET denoising methods have demonstrated superior performance in generating high-quality reconstruction. However, these methods require a large amount of representative data for training, which can be difficult to collect and share due to medical data privacy regulations. Moreover, low-dose PET data at different institutions may use different low-dose protocols, leading to non-identical data distribution. While previous federated learning (FL) algorithms enable multi-institution collaborative training without the need of aggregating local data, it is challenging for previous methods to address the large domain shift caused by different low-dose PET settings, and the application of FL to PET is still under-explored. In this work, we propose a federated transfer learning (FTL) framework for low-dose PET denoising using heterogeneous low-dose data. Our experimental results on simulated multi-institutional data demonstrate that our method can efficiently utilize heterogeneous low-dose data without compromising data privacy for achieving superior low-dose PET denoising performance for different institutions with different low-dose settings, as compared to previous FL methods.

Evaluation of the cumulative Cherenkov converted dose on TSET patients with multiple Cherenkov cameras.

Cherenkov images can be used for the quality assurance of dose homogeneity in total skin electron therapy (TSET). For the dose mapping purpose, this study reconstructed the patient model from 3D scans using registration algorithms and computer animation techniques. The Cherenkov light emission of the patient's surface was extracted from multi-view Cherenkov images, converted into dose distribution, and projected onto the patient's 3D model, allowing for dose cumulation and evaluation. The projected result from multiple Cherenkov cameras provides additional information about Cherenkov emission on the sides of the patients, which improves the agreement between the Cherenkov converted dose and the OSLD measurements.

DuDoSS: Deep-learning-based dual-domain sinogram synthesis from sparsely sampled projections of cardiac SPECT.

PURPOSE: Myocardial perfusion imaging (MPI) using single-photon emission-computed tomography (SPECT) is widely applied for the diagnosis of cardiovascular diseases. In clinical practice, the long scanning procedures and acquisition time might induce patient anxiety and discomfort, motion artifacts, and misalignments between SPECT and computed tomography (CT). Reducing the number of projection angles provides a solution that results in a shorter scanning time. However, fewer projection angles might cause lower reconstruction accuracy, higher noise level, and reconstruction artifacts due to reduced angular sampling. We developed a deep-learning-based approach for high-quality SPECT image reconstruction using sparsely sampled projections. METHODS: We proposed a novel deep-learning-based dual-domain sinogram synthesis (DuDoSS) method to recover full-view projections from sparsely sampled projections of cardiac SPECT. DuDoSS utilized the SPECT images predicted in the image domain as guidance to generate synthetic full-view projections in the sinogram domain. The synthetic projections were then reconstructed into non-attenuation-corrected and attenuation-corrected (AC) SPECT images for voxel-wise and segment-wise quantitative evaluations in terms of normalized mean square error (NMSE) and absolute percent error (APE). Previous deep-learning-based approaches, including direct sinogram generation (Direct Sino2Sino) and direct image prediction (Direct Img2Img), were tested in this study for comparison. The dataset used in this study included a total of 500 anonymized clinical stress-state MPI studies acquired on a GE NM/CT 850 scanner with 60 projection angles following the injection of 99m Tc-tetrofosmin. RESULTS: Our proposed DuDoSS generated more consistent synthetic projections and SPECT images with the ground truth than other approaches. The average voxel-wise NMSE between the synthetic projections by DuDoSS and the ground-truth full-view projections was 2.08% ± 0.81%, as compared to 2.21% ± 0.86% (p < 0.001) by Direct Sino2Sino. The averaged voxel-wise NMSE between the AC SPECT images by DuDoSS and the ground-truth AC SPECT images was 1.63% ± 0.72%, as compared to 1.84% ± 0.79% (p < 0.001) by Direct Sino2Sino and 1.90% ± 0.66% (p < 0.001) by Direct Img2Img. The averaged segment-wise APE between the AC SPECT images by DuDoSS and the ground-truth AC SPECT images was 3.87% ± 3.23%, as compared to 3.95% ± 3.21% (p = 0.023) by Direct Img2Img and 4.46% ± 3.58% (p < 0.001) by Direct Sino2Sino. CONCLUSIONS: Our proposed DuDoSS is feasible to generate accurate synthetic full-view projections from sparsely sampled projections for cardiac SPECT. The synthetic projections and reconstructed SPECT images generated from DuDoSS are more consistent with the ground-truth full-view projections and SPECT images than other approaches. DuDoSS can potentially enable fast data acquisition of cardiac SPECT.

Computational dose visualization & comparison in total skin electron treatment suggests superior coverage by the rotational versus the Stanford technique.

INTRODUCTION/BACKGROUND: The goal of Total Skin Electron Therapy (TSET) is to achieve a uniform surface dose, although assessment of this is never really done and typically limited points are sampled. A computational treatment simulation approach was developed to estimate dose distributions over the body surface, to compare uniformity of (i) the 6 pose Stanford technique and (ii) the rotational technique. METHODS: The relative angular dose distributions from electron beam irradiation was calculated by Monte Carlo simulation for cylinders with a range of diameters, approximating body part curvatures. These were used to project dose onto a 3D body model of the TSET patient's skin surfaces. Computer animation methods were used to accumulate the dose values, for display and analysis of the homogeneity of coverage. RESULTS: The rotational technique provided more uniform coverage than the Stanford technique. Anomalies of under dose were observed in lateral abdominal regions, above the shoulders and in the perineum. The Stanford technique had larger areas of low dose laterally. In the rotational technique, 90% of the patient's skin was within ±10% of the prescribed dose, while this percentage decreased to 60% or 85% for the Stanford technique, varying with patient body mass. Interestingly, the highest discrepancy was most apparent in high body mass patients, which can be attributed to the loss of tangent dose at low angles of curvature. DISCUSSION/CONCLUSION: This simulation and visualization approach is a practical means to analyze TSET dose, requiring only optical surface body topography scans. Under- and over-exposed body regions can be found, and irradiation could be customized to each patient. Dose Area Histogram (DAH) distribution analysis showed the rotational technique to have better uniformity, with most areas within 10% of the umbilicus value. Future use of this approach to analyze dose coverage is possible as a routine planning tool.

Generation of parametric Ki images for FDG PET using two 5-min scans.

PURPOSE: The net uptake rate constant (Ki ) derived from dynamic imaging is considered the gold standard quantification index for FDG PET. In this study, we investigated the feasibility and assessed the clinical usefulness of generating Ki images for FDG PET using only two 5-min scans with population-based input function (PBIF). METHODS: Using a Siemens Biograph mCT, 10 subjects with solid lung nodules underwent a single-bed dynamic FDG PET scan and 13 subjects (five healthy and eight cancer patients) underwent a whole-body dynamic FDG PET scan in continuous-bed-motion mode. For each subject, a standard Ki image was generated using the complete 0-90 min dynamic data with Patlak analysis (t* = 20 min) and individual patient's input function, while a dual-time-point Ki image was generated from two 5-min scans based on the Patlak equations at early and late scans with the PBIF. Different start times for the early (ranging from 20 to 55 min with an increment of 5 min) and late (ranging from 50 to 85 min with an increment of 5 min) scans were investigated with the interval between scans being at least 30 min (36 protocols in total). The optimal dual-time-point protocols were then identified. Regions of interest (ROI) were drawn on nodules for the lung nodule subjects, and on tumors, cerebellum, and bone marrow for the whole-body-imaging subjects. Quantification accuracy was compared using the mean value of each ROI between standard Ki (gold standard) and dual-time-point Ki , as well as between standard Ki and relative standardized uptake value (SUV) change that is currently used in clinical practice. Correlation coefficients and least squares fits were calculated for each dual-time-point protocol and for each ROI. Then, the predefined criteria for identifying a reliable dual-time-point Ki estimation for each ROI were empirically determined as: (1) the squared correlation coefficient (R2 ) between standard Ki and dual-time-point Ki is larger than 0.9; (2) the absolute difference between the slope of the equality line (1.0) and that of the fitted line when plotting standard Ki versus dual-time-point Ki is smaller than 0.1; (3) the absolute value of the intercept of the fitted line when plotting standard Ki versus dual-time-point Ki normalized by the mean of the standard Ki across all subjects for each ROI is smaller than 10%. Using Williams' one-tailed t test, the correlation coefficient (R) between standard Ki and dual-time-point Ki was further compared with that between standard Ki and relative SUV change, for each dual-time-point protocol and for each ROI. RESULTS: Reliable dual-time-point Ki images were obtained for all the subjects using our proposed method. The percentage error introduced by the PBIF on the dual-time-point Ki estimation was smaller than 1% for all 36 protocols. Using the predefined criteria, reliable dual-time-point Ki estimation could be obtained in 25 of 36 protocols for nodules and in 34 of 36 protocols for tumors. A longer time interval between scans provided a more accurate Ki estimation in general. Using the protocol of 20-25 min plus 80-85 or 85-90 min, very high correlations were obtained between standard Ki and dual-time-point Ki (R2  = 0.994, 0.980, 0.971 and 0.925 for nodule, tumor, cerebellum, and bone marrow), with all the slope values with differences ≤0.033 from 1 and all the intercept values with differences ≤0.0006 mL/min/cm3 from 0. The corresponding correlations were much lower between standard Ki and relative SUV change (R2  = 0.673, 0.684, 0.065, 0.246). Dual-time-point Ki showed a significantly higher quantification accuracy with respect to standard Ki than relative SUV change for all the 36 protocols (p < 0.05 using Williams' one-tailed t test). CONCLUSIONS: Our proposed approach can obtain reliable Ki images and accurate Ki quantification from dual-time-point scans (5-min per scan), and provide significantly higher quantification accuracy than relative SUV change that is currently used in clinical practice.

Cherenkov imaging for Total Skin Electron Therapy - an evaluation of dose uniformity.

Total Skin Electron Therapy (TSET) utilizes high-energy electrons to treat cancers on the entire body surface. The otherwise invisible radiation beam can be observed via the optical Cherenkov photons emitted from interaction between the high-energy electron beam and tissue. Cherenkov emission can be used to evaluate the dose uniformity on the surface of the patient in real-time using a time-gated intensified camera system. Each patient was monitored during TSET by in-vivo detectors (IVD) as well as Scintillators. Patients undergoing TSET in various conditions (whole body and half body) were imaged and analyzed. A rigorous methodology for converting Cherenkov intensity to surface dose as products of correction factors, including camera vignette correction factor, incident radiation correction factor, and tissue optical properties correction factor. A comprehensive study has been carried out by inspecting various positions on the patients such as vertex, chest, perineum, shins, and foot relative to the umbilicus point (the prescription point).

Computer animation body surface analysis of total skin electron radiation therapy dose homogeneity via Cherenkov imaging.

Purpose: Quality assurance (QA) of dose homogeneity in total skin electron therapy (TSET) is challenging since each patient is positioned in six standing poses with two beam angles. Our study tested the feasibility of a unique approach for TSET QA through computational display of the cumulative dose, constructed and synthesized by computer animation methods. Approach: Dose distributions from Cherenkov emission images were projected onto a scanned 3D body model. Topographically mapped surfaces of the patient were recorded in each of six different delivery positions, while a Cherenkov camera acquired images. Computer animation methods allowed a fitted 3D human body model of the patient to be created with deformation of the limbs and torso to each position. A two-dimensional skin map was extracted from the 3D model of the full surface of the patient. This allowed the dose mapping to be additively accumulated independent of body position, with the total dose summed in a 2D map and reinterpreted on the 3D body display. Results: For the body model, the mean Hausdorff error distance was below 2 cm, setting the spatial accuracy limit. The dose distribution over the patient's 3D model generally matched the Cherenkov/dose images. The dose distribution mapping was estimated to be near 1.5 cm accuracy based upon a phantom study. The body model must most closely match at the edges of the mesh to ensure that high dose gradients are not projected onto the wrong location. Otherwise 2 to 3 cm level errors in positioning in the mesh do not appear to cause larger than 5% dose errors. The cumulative dose images showed regions of overlap laterally and regions of low intensity in the posterior arms. Conclusions: The proposed modeling and animation can be used to visualize and analyze the accumulated dose in TSET via display of the summed dose/Cherenkov images on a single body surface.

Cherenkov imaging for total skin electron therapy (TSET).

BACKGROUND: Total skin electron therapy (TSET) utilizes high-energy electrons to treat malignancies on the entire body surface. The otherwise invisible radiation beam can be observed via the optical Cherenkov photons emitted from interactions between the high-energy electron beam and tissue. METHODS AND MATERIALS: With a time-gated intensified camera system, the Cherenkov emission can be used to evaluate the dose uniformity on the surface of the patient in real time. Fifteen patients undergoing TSET in various conditions (whole body and half body) were imaged and analyzed. Each patient was monitored during TSET via in vivo detectors (IVD) in nine locations. For accurate Cherenkov imaging, a comparison between IVD and Cherenkov profiles was conducted using a polyvinyl chloride board to establish the perspective corrections. RESULTS AND DISCUSSION: With proper corrections developed in this study including the perspective and inverse square corrections, the Cherenkov imaging provided two-dimensional maps proportional to dose and projected on patient skin. The results of ratio between chest and umbilicus points were in good agreement with in vivo point dose measurements, with a standard deviation of 2.4% compared to OSLD measurements. CONCLUSIONS: Cherenkov imaging is a viable tool for validating patient-specific dose distributions during TSET.

Cherenkov imaging for linac beam shape analysis as a remote electronic quality assessment verification tool.

PURPOSE: A remote imaging system tracking Cherenkov emission was analyzed to verify that the linear accelerator (linac) beam shape could be quantitatively measured at the irradiation surface for Quality Audit (QA). METHODS: The Cherenkov camera recorded 2D dose images delivered on a solid acrylonitrile butadiene styrene (ABS) plastic phantom surface for a range of square beam sizes, and 6 MV photons. Imaging was done at source to surface distance (SSD) of 100 cm and compared to GaF film images and linac light fields of the same beam sizes, ranging over 5 × 5 cm2 up to 20 × 20 cm2 . Line profiles of each field were compared in both X and Y jaw directions. Each measurement was repeated on two different Clinac2100 machines. An interreader comparison of the beam width interpretation was completed using procedures commonly employed for beam to light field coincidence verification. Cherenkov measurements are also done for beams of complex treatment plan and isocenter QA. RESULTS: The Cherenkov image widths matched with the measured GaF images and light field images, with accuracy in the range of ±1 mm standard deviation. The differences between the measurements were minor and within tolerance of geometrical requirement of standard linac QA procedures conducted by human setup verification, which had a similar error range. The measurement made by the remote imaging system allowed for beam shape extraction of radiation fields at the SSD location of the beam. CONCLUSIONS: The proposed Cherenkov image acquisition system provides a valid way to remotely confirm radiation field sizes and provides similar information to that obtained from the linac light field or GaF film estimates of the beam size. The major benefit of this approach is that with a fixed installation of the camera, testing could be done completely under software control with automated image analysis, potentially simplifying conventional QA procedures with appropriate calibration of boundary definitions, and the natural extension to capturing dynamic treatment beamlets at SSD could have future value, such as verification of beam plans with complex beam shapes, like IMRT or "star-shot" QA for the isocenter.

Cherenkov excited short-wavelength infrared fluorescence imaging in vivo with external beam radiation.

Cherenkov emission induced by external beam radiation therapy from a clinical linear accelerator (LINAC) can be used to excite phosphors deep in biological tissues. As with all luminescence imaging, there is a desire to minimize the spectral overlap between the excitation light and emission wavelengths, here between the Cherenkov and the phosphor. Cherenkov excited short-wavelength infrared (SWIR, 1000 to 1700 nm) fluorescence imaging has been demonstrated for the first time, using long Stokes-shift fluorophore PdSe quantum dots (QD) with nanosecond lifetime and an optimized SWIR detection. The 1 / λ2 intensity spectrum characteristic of Cherenkov emission leads to low overlap of this into the fluorescence spectrum of PdSe QDs in the SWIR range. Additionally, using a SWIR camera itself inherently ignores the stronger Cherenkov emission wavelengths dominant across the visible spectrum. The SWIR luminescence was shown to extend the depth sensitivity of Cherenkov imaging, which could be used for applications in radiotherapy sensing and imaging in human tissue with targeted molecular probes.

Observation of short wavelength infrared (SWIR) Cherenkov emission.

Cherenkov emission induced by external beam radiation from a clinical linear accelerator has been shown in preclinical molecular imaging and clinical imaging. The broad spectrum Cherenkov emission should have a short wavelength infrared (SWIR, 1000-1700 nm) component, as predicted theoretically. To the best of our knowledge, this Letter is the first experimental observation of this SWIR Cherenkov emission induced by external beam radiation. The measured spectrum of SWIR Cherenkov emission matches the theoretical prediction, with a fluence rate near one-third of the visible and near-infrared red emissions (Vis-NIR, 400-900 nm). Imaging in water-based phantoms and biological tissues indicates that there is a sufficient fluence rate for radiotherapy dosimetry applications. The spatial resolution is improved approximately 5.3 times with SWIR Cherenkov emission detection versus Vis-NIR Cherenkov emission, which provides some improvement in the potential for higher resolution Cherenkov emission dosimetry and molecular sensing during clinical radiotherapy by imaging with SWIR wavelengths.