Deformable lung 4DCT image registration via landmark-driven cycle network.

BACKGROUND: An automated, accurate, and efficient lung four-dimensional computed tomography (4DCT) image registration method is clinically important to quantify respiratory motion for optimal motion management. PURPOSE: The purpose of this work is to develop a weakly supervised deep learning method for 4DCT lung deformable image registration (DIR). METHODS: The landmark-driven cycle network is proposed as a deep learning platform that performs DIR of individual phase datasets in a simulation 4DCT. This proposed network comprises a generator and a discriminator. The generator accepts moving and target CTs as input and outputs the deformation vector fields (DVFs) to match the two CTs. It is optimized during both forward and backward paths to enhance the bi-directionality of DVF generation. Further, the landmarks are used to weakly supervise the generator network. Landmark-driven loss is used to guide the generator's training. The discriminator then judges the realism of the deformed CT to provide extra DVF regularization. RESULTS: We performed four-fold cross-validation on 10 4DCT datasets from the public DIR-Lab dataset and a hold-out test on our clinic dataset, which included 50 4DCT datasets. The DIR-Lab dataset was used to evaluate the performance of the proposed method against other methods in the literature by calculating the DIR-Lab Target Registration Error (TRE). The proposed method outperformed other deep learning-based methods on the DIR-Lab datasets in terms of TRE. Bi-directional and landmark-driven loss were shown to be effective for obtaining high registration accuracy. The mean and standard deviation of TRE for the DIR-Lab datasets was 1.20 ± 0.72 mm and the mean absolute error (MAE) and structural similarity index (SSIM) for our datasets were 32.1 ± 11.6 HU and 0.979 ± 0.011, respectively. CONCLUSION: The landmark-driven cycle network has been validated and tested for automatic deformable image registration of patients' lung 4DCTs with results comparable to or better than competing methods.

A Couch Mounted Smartphone-based Motion Monitoring System for Radiation Therapy.

PURPOSE: Surface-guided radiation-therapy (SGRT) systems are being adopted into clinical practice for patient setup and motion monitoring. However, commercial systems remain cost prohibitive to resource-limited clinics around the world. Our aim is to develop and validate a smartphone-based application using LiDAR cameras (such as on recent Apple iOS devices) for facilitating SGRT in low-resource centers. The proposed SGRT application was tested at multiple institutions and validated using phantoms and volunteers against various commercial systems to demonstrate feasibility. METHODS AND MATERIALS: An iOS application was developed in Xcode and written in Swift using the Augmented-Reality (AR) Kit and implemented on an Apple iPhone 13 Pro with a built-in LiDAR camera. The application contains multiple features: 1) visualization of both the camera and depth video feeds (at a ∼60Hz sample-frequency), 2) region-of-interest (ROI) selection over the patient's anatomy where motion is measured, 3) chart displaying the average motion over time in the ROI, and 4) saving/exporting the motion traces and surface map over the ROI for further analysis. The iOS application was tested to evaluate depth measurement accuracy for: 1) different angled surfaces, 2) different field-of-views over different distances, and 3) similarity to a commercially available SGRT systems (Vision RT AlignRT and Varian IDENTIFY) with motion phantoms and healthy volunteers across 3 institutions. Measurements were analyzed using linear-regressions and Bland-Altman analysis. RESULTS: Compared with the clinical system measurements (reference), the iOS application showed excellent agreement for depth (r = 1.000, P < .0001; bias = -0.07±0.24 cm) and angle (r = 1.000, P < .0001; bias = 0.02±0.69°) measurements. For free-breathing traces, the iOS application was significantly correlated to phantom motion (institute 1: r = 0.99, P < .0001; bias =-0.003±0.03 cm; institute 2: r = 0.98, P < .0001; bias = -0.001±0.10 cm; institute 3: r = 0.97, P < .0001; bias = 0.04±0.06 cm) and healthy volunteer motion (institute 1: r = 0.98, P < .0001; bias = -0.008±0.06 cm; institute 2: r = 0.99, P < .0001; bias = -0.007±0.12 cm; institute 3: r = 0.99, P < .0001; bias = -0.001±0.04 cm). CONCLUSIONS: The proposed approach using a smartphone-based application provides a low-cost platform that could improve access to surface-guided radiation therapy accounting for motion.

Single energy CT-based mass density and relative stopping power estimation for proton therapy using deep learning method.

Background: The number of patients undergoing proton therapy has increased in recent years. Current treatment planning systems (TPS) calculate dose maps using three-dimensional (3D) maps of relative stopping power (RSP) and mass density. The patient-specific maps of RSP and mass density were obtained by translating the CT number (HU) acquired using single-energy computed tomography (SECT) with appropriate conversions and coefficients. The proton dose calculation uncertainty of this approach is 2.5%-3.5% plus 1 mm margin. SECT is the major clinical modality for proton therapy treatment planning. It would be intriguing to enhance proton dose calculation accuracy using a deep learning (DL) approach centered on SECT. Objectives: The purpose of this work is to develop a deep learning method to generate mass density and relative stopping power (RSP) maps based on clinical single-energy CT (SECT) data for proton dose calculation in proton therapy treatment. Methods: Artificial neural networks (ANN), fully convolutional neural networks (FCNN), and residual neural networks (ResNet) were used to learn the correlation between voxel-specific mass density, RSP, and SECT CT number (HU). A stoichiometric calibration method based on SECT data and an empirical model based on dual-energy CT (DECT) images were chosen as reference models to evaluate the performance of deep learning neural networks. SECT images of a CIRS 062M electron density phantom were used as the training dataset for deep learning models. CIRS anthropomorphic M701 and M702 phantoms were used to test the performance of deep learning models. Results: For M701, the mean absolute percentage errors (MAPE) of the mass density map by FCNN are 0.39%, 0.92%, 0.68%, 0.92%, and 1.57% on the brain, spinal cord, soft tissue, bone, and lung, respectively, whereas with the SECT stoichiometric method, they are 0.99%, 2.34%, 1.87%, 2.90%, and 12.96%. For RSP maps, the MAPE of FCNN on M701 are 0.85%, 2.32%, 0.75%, 1.22%, and 1.25%, whereas with the SECT reference model, they are 0.95%, 2.61%, 2.08%, 7.74%, and 8.62%. Conclusion: The results show that deep learning neural networks have the potential to generate accurate voxel-specific material property information, which can be used to improve the accuracy of proton dose calculation. Advances in knowledge: Deep learning-based frameworks are proposed to estimate material mass density and RSP from SECT with improved accuracy compared with conventional methods.

Data-Driven Volumetric Image Generation from Surface Structures using a Patient-Specific Deep Leaning Model.

The advent of computed tomography significantly improves patients' health regarding diagnosis, prognosis, and treatment planning and verification. However, tomographic imaging escalates concomitant radiation doses to patients, inducing potential secondary cancer by 4%. We demonstrate the feasibility of a data-driven approach to synthesize volumetric images using patients' surface images, which can be obtained from a zero-dose surface imaging system. This study includes 500 computed tomography (CT) image sets from 50 patients. Compared to the ground truth CT, the synthetic images result in the evaluation metric values of 26.9 ± 4.1 Hounsfield units, 39.1 ± 1.0 dB, and 0.965 ± 0.011 regarding the mean absolute error, peak signal-to-noise ratio, and structural similarity index measure. This approach provides a data integration solution that can potentially enable real-time imaging, which is free of radiation-induced risk and could be applied to image-guided medical procedures.

2D medical image synthesis using transformer-based denoising diffusion probabilistic model.

Objective. Artificial intelligence (AI) methods have gained popularity in medical imaging research. The size and scope of the training image datasets needed for successful AI model deployment does not always have the desired scale. In this paper, we introduce a medical image synthesis framework aimed at addressing the challenge of limited training datasets for AI models.Approach. The proposed 2D image synthesis framework is based on a diffusion model using a Swin-transformer-based network. This model consists of a forward Gaussian noise process and a reverse process using the transformer-based diffusion model for denoising. Training data includes four image datasets: chest x-rays, heart MRI, pelvic CT, and abdomen CT. We evaluated the authenticity, quality, and diversity of the synthetic images using visual Turing assessments conducted by three medical physicists, and four quantitative evaluations: the Inception score (IS), Fréchet Inception Distance score (FID), feature similarity and diversity score (DS, indicating diversity similarity) between the synthetic and true images. To leverage the framework value for training AI models, we conducted COVID-19 classification tasks using real images, synthetic images, and mixtures of both images.Main results. Visual Turing assessments showed an average accuracy of 0.64 (accuracy converging to50%indicates a better realistic visual appearance of the synthetic images), sensitivity of 0.79, and specificity of 0.50. Average quantitative accuracy obtained from all datasets were IS = 2.28, FID = 37.27, FDS = 0.20, and DS = 0.86. For the COVID-19 classification task, the baseline network obtained an accuracy of 0.88 using a pure real dataset, 0.89 using a pure synthetic dataset, and 0.93 using a dataset mixed of real and synthetic data.Significance. A image synthesis framework was demonstrated for medical image synthesis, which can generate high-quality medical images of different imaging modalities with the purpose of supplementing existing training sets for AI model deployment. This method has potential applications in many data-driven medical imaging research.

MRI-based prostate and dominant lesion segmentation using cascaded scoring convolutional neural network.

PURPOSE: Dose escalation to dominant intraprostatic lesions (DILs) is a novel treatment strategy to improve the treatment outcome of prostate radiation therapy. Treatment planning requires accurate and fast delineation of the prostate and DILs. In this study, a 3D cascaded scoring convolutional neural network is proposed to automatically segment the prostate and DILs from MRI. METHODS AND MATERIALS: The proposed cascaded scoring convolutional neural network performs end-to-end segmentation by locating a region-of-interest (ROI), identifying the object within the ROI, and defining the target. A scoring strategy, which is learned to judge the segmentation quality of DIL, is integrated into cascaded convolutional neural network to solve the challenge of segmenting the irregular shapes of the DIL. To evaluate the proposed method, 77 patients who underwent MRI and PET/CT were retrospectively investigated. The prostate and DIL ground truth contours were delineated by experienced radiologists. The proposed method was evaluated with fivefold cross-validation and holdout testing. RESULTS: The average centroid distance, volume difference, and Dice similarity coefficient (DSC) value for prostate/DIL are 4.3 ± 7.5/3.73 ± 3.78 mm, 4.5 ± 7.9/0.41 ± 0.59 cc, and 89.6 ± 8.9/84.3 ± 11.9%, respectively. Comparable results were obtained in the holdout test. Similar or superior segmentation outcomes were seen when compared the results of the proposed method to those of competing segmentation approaches. CONCLUSIONS: The proposed automatic segmentation method can accurately and simultaneously segment both the prostate and DILs. The intended future use for this algorithm is focal boost prostate radiation therapy.

Magnetic resonance imaging contrast enhancement synthesis using cascade networks with local supervision.

PURPOSE: Gadolinium-based contrast agents (GBCAs) are widely administrated in MR imaging for diagnostic studies and treatment planning. Although GBCAs are generally thought to be safe, various health and environmental concerns have been raised recently about their use in MR imaging. The purpose of this work is to derive synthetic contrast enhance MR images from unenhanced counterpart images, thereby eliminating the need for GBCAs, using a cascade deep learning workflow that incorporates contour information into the network. METHODS AND MATERIALS: The proposed workflow consists of two sequential networks: (1) a retina U-Net, which is first trained to derive semantic features from the non-contrast MR images in representing the tumor regions; and (2) a synthesis module, which is trained after the retina U-Net to take the concatenation of the semantic feature maps and non-contrast MR image as input and to generate the synthetic contrast enhanced MR images. After network training, only the non-contrast enhanced MR images are required for the input in the proposed workflow. The MR images of 369 patients from the multimodal brain tumor segmentation challenge 2020 (BraTS2020) dataset were used in this study to evaluate the proposed workflow for synthesizing contrast enhanced MR images (200 patients for five-fold cross-validation and 169 patients for hold-out test). Quantitative evaluations were conducted by calculating the normalized mean absolute error (NMAE), structural similarity index measurement (SSIM), and Pearson correlation coefficient (PCC). The original contrast enhanced MR images were considered as the ground truth in this analysis. RESULTS: The proposed cascade deep learning workflow synthesized contrast enhanced MR images that are not visually differentiable from the ground truth with and without supervision of the tumor contours during the network training. Difference images and profiles of the synthetic contrast enhanced MR images revealed that intensity differences could be observed in the tumor region if the contour information was not incorporated in network training. Among the hold-out test patients, mean values and standard deviations of the NMAE, SSIM, and PCC were 0.063±0.022, 0.991±0.007 and 0.995±0.006, respectively, for the whole brain; and were 0.050±0.025, 0.993±0.008 and 0.999±0.003, respectively, for the tumor contour regions. Quantitative evaluations with five-fold cross-validation and hold-out test showed that the calculated metrics can be significantly enhanced (p-values ≤ 0.002) with the tumor contour supervision in network training. CONCLUSION: The proposed workflow was able to generate synthetic contrast enhanced MR images that closely resemble the ground truth images from non-contrast enhanced MR images when the network training included tumor contours. These results suggest that it may be possible to minimize the use of GBCAs in cranial MR imaging studies.

Self-supervised learning for accelerated 3D high-resolution ultrasound imaging.

PURPOSE: Ultrasound (US) imaging has been widely used in diagnosis, image-guided intervention, and therapy, where high-quality three-dimensional (3D) images are highly desired from sparsely acquired two-dimensional (2D) images. This study aims to develop a deep learning-based algorithm to reconstruct high-resolution (HR) 3D US images only reliant on the acquired sparsely distributed 2D images. METHODS: We propose a self-supervised learning framework using cycle-consistent generative adversarial network (cycleGAN), where two independent cycleGAN models are trained with paired original US images and two sets of low-resolution (LR) US images, respectively. The two sets of LR US images are obtained through down-sampling the original US images along the two axes, respectively. In US imaging, in-plane spatial resolution is generally much higher than through-plane resolution. By learning the mapping from down-sampled in-plane LR images to original HR US images, cycleGAN can generate through-plane HR images from original sparely distributed 2D images. Finally, HR 3D US images are reconstructed by combining the generated 2D images from the two cycleGAN models. RESULTS: The proposed method was assessed on two different datasets. One is automatic breast ultrasound (ABUS) images from 70 breast cancer patients, the other is collected from 45 prostate cancer patients. By applying a spatial resolution enhancement factor of 3 to the breast cases, our proposed method achieved the mean absolute error (MAE) value of 0.90 ± 0.15, the peak signal-to-noise ratio (PSNR) value of 37.88 ± 0.88 dB, and the visual information fidelity (VIF) value of 0.69 ± 0.01, which significantly outperforms bicubic interpolation. Similar performances have been achieved using the enhancement factor of 5 in these breast cases and using the enhancement factors of 5 and 10 in the prostate cases. CONCLUSIONS: We have proposed and investigated a new deep learning-based algorithm for reconstructing HR 3D US images from sparely acquired 2D images. Significant improvement on through-plane resolution has been achieved by only using the acquired 2D images without any external atlas images. Its self-supervision capability could accelerate HR US imaging.

Interplay Effect of Target Motion and Pencil-Beam Scanning in Proton Therapy for Pediatric Patients.

PURPOSE: To investigate the effect of interplay between spot-scanning proton beams and respiration-induced tumor motion on internal target volume coverage for pediatric patients. MATERIALS AND METHODS: Photon treatments for 10 children with representative tumor motions (1-13 mm superior-inferior) were replanned to simulate single-field uniform dose- optimized proton therapy. Static plans were designed by using average computed tomography (CT) data sets created from 4D CT data to obtain nominal dose distributions. The motion interplay effect was simulated by assigning each spot in the static plan delivery sequence to 1 of 10 respiratory-phase CTs, using the actual patient breathing trace and specifications of a synchrotron-based proton system. Dose distributions for individual phases were deformed onto the space of the average CT and summed to produce the accumulated dose distribution, whose dose-volume histogram was compared with the one from the static plan. RESULTS: Tumor motion had minimal impact on the internal target volume hot spot (D2), which deviated by <3% from the nominal values of the static plans. The cold spot (D98) was also minimally affected, except in 2 patients with diaphragmatic tumor motion exceeding 10 mm. The impact on tumor coverage was more pronounced with respect to the V99 rather than the V95. Decreases of 10% to 49% in the V99 occurred in multiple patients for whom the beam paths traversed the lung-diaphragm interface and were, therefore, more sensitive to respiration-induced changes in the water equivalent path length. Fractionation alone apparently did not mitigate the interplay effect beyond 6 fractions. CONCLUSION: The interplay effect is not a concern when delivering scanning proton beams to younger pediatric patients with tumors located in the retroperitoneal space and tumor motion of <5 mm. Children and adolescents with diaphragmatic tumor motion exceeding 10 mm require special attention, because significant declines in target coverage and dose homogeneity were seen in simulated treatments of such patients.

Clinical evaluation of the iterative metal artifact reduction algorithm for CT simulation in radiotherapy.

PURPOSE: To clinically evaluate an iterative metal artifact reduction (IMAR) algorithm prototype in the radiation oncology clinic setting by testing for accuracy in CT number retrieval, relative dosimetric changes in regions affected by artifacts, and improvements in anatomical and shape conspicuity of corrected images. METHODS: A phantom with known material inserts was scanned in the presence/absence of metal with different configurations of placement and sizes. The relative change in CT numbers from the reference data (CT with no metal) was analyzed. The CT studies were also used for dosimetric tests where dose distributions from both photon and proton beams were calculated. Dose differences and gamma analysis were calculated to quantify the relative changes between doses calculated on the different CT studies. Data from eight patients (all different treatment sites) were also used to quantify the differences between dose distributions before and after correction with IMAR, with no reference standard. A ranking experiment was also conducted to analyze the relative confidence of physicians delineating anatomy in the near vicinity of the metal implants. RESULTS: IMAR corrected images proved to accurately retrieve CT numbers in the phantom study, independent of metal insert configuration, size of the metal, and acquisition energy. For plastic water, the mean difference between corrected images and reference images was -1.3 HU across all scenarios (N = 37) with a 90% confidence interval of [-2.4, -0.2] HU. While deviations were relatively higher in images with more metal content, IMAR was able to effectively correct the CT numbers independent of the quantity of metal. Residual errors in the CT numbers as well as some induced by the correction algorithm were found in the IMAR corrected images. However, the dose distributions calculated on IMAR corrected images were closer to the reference data in phantom studies. Relative spatial difference in the dose distributions in the regions affected by the metal artifacts was also observed in patient data. However, in absence of a reference ground truth (CT set without metal inserts), these differences should not be interpreted as improvement/deterioration of the accuracy of calculated dose. With limited data presented, it was observed that proton dosimetry was affected more than photons as expected. Physicians were significantly more confident contouring anatomy in the regions affected by artifacts. While site specific preferences were detected, all indicated that they would consistently use IMAR corrected images. CONCLUSIONS: IMAR correction algorithm could be readily implemented in an existing clinical workflow upon commercial release. While residual errors still exist in IMAR corrected images, these images present with better overall conspicuity of the patient/phantom geometry and offer more accurate CT numbers for improved local dosimetry. The variety of different scenarios included herein attest to the utility of the evaluated IMAR for a wide range of radiotherapy clinical scenarios.

Imaging dose in variable pitch body perfusion CT scans: an analysis using TG111 formalism.

PURPOSE: To investigate the variation of imaging dose with tube potential in variable pitch body CT perfusion (CTp) protocols using the TG111 dosimetric formalism. METHODS: TG111 recommendations were followed in choosing the phantom, dosimetric equipment, and methodology. Specifically, equilibrium doses (D(eq)) were measured centrally and peripherally in a long PMMA phantom. Reference planar average equilibrium doses were determined for each tube potential, for a reference set of exposure parameters (collimation, pitch, filtration) on a Siemens Definition CT scanner. These reference values were utilized to predict the imaging dose during perfusion scans using interpretations of the TG111 formalism. As a gold reference, the midscan average planar perfusion doses (D(CTp)) were obtained directly from central and peripheral D(eq) measurements for body CTp scans (144 and 271 mm) using variable pitch acquisition. Measurement-based D(CTp) values obtained using a thimble chamber were compared to the TG111-predicted values, and to CTDI(vol) reported at the console. RESULTS: Reference planar average equilibrium dose values measured for reference uniform pitch helical scans were consistently higher than console-reported or measured values for CTDI(vol). The measurement-based perfusion dose D(CTp) was predicted accurately by the reported CTDI(vol) for the 144 mm scan. The 271 mm scans delivered systematically larger dose than reported. The TG111-based dose estimates were proven to be conservative, as they were systematically higher than both the measured and the reported imaging doses. CONCLUSIONS: Upon successful implementation of TG111 formalism, standard imaging dose was measured for a body CTp protocol using the variable pitch helical acquisition. The TG111 formalism is not directly applicable to this type of acquisition. Measurement of dose for all variable pitch protocols is strongly suggested.

An alternative approach to histopathological validation of PET imaging for radiation therapy image-guidance: a proof of concept.

PURPOSE: In radiotherapy, PET images can be used to guide the delivery of selectively escalated doses to biologically relevant tumour subvolumes. Validation of PET for such applications requires demonstration of spatial coincidence between PET tracer uptake pattern and the histopathologically confirmed target. This study introduces a novel approach to histopathological validation of PET image segmentation for radiotherapy guidance. METHODS AND MATERIALS: Sequential tissue sections from surgically excised whole-tumour specimens were used to acquire full 3D-sets of both histopathological images (microscopy) and PET tracer distribution images (autoradiography). After these datasets were accurately registered, a full 3D autoradiographic distribution of PET tracer was reconstructed and used to obtain synthetic PET images (sPET) by simulating the image deterioration induced by processes involved in PET image formation. To illustrate the method, sPET images were used in this study to investigate spatial coincidence between high FDG uptake areas and the distribution of viable tissue in two small animal tumour models. RESULTS: The reconstructed 3D autoradiographic distribution of the PET tracer was spatially coherent, as indicated by the high average value of the normalised pixel-by-pixel correlation of intensities between successive slices (0.84 ± 0.05 and 0.94 ± 0.02). The loss of detail in the sPET images versus the 3D autoradiography was significant as indicated by Dice coefficient values corresponding to the two tumours (0 and 0.1 at 70% threshold). The maximum overlap between the FDG segmented volumes and the extent of the viable tissue as indicated by Dice coefficient values, was 0.8 for one tumour (for the image thresholded at 22% of max intensity) and 0.88 for the other (threshold of 14% of max intensity). CONCLUSION: It was demonstrated that the use of synthetic PET images for histopathological validation allows for bypassing a technically challenging and error-prone step of registering non-invasive PET images with histopathology.

High-resolution radioluminescence microscopy of 18F-FDG uptake by reconstructing the ß-ionization track.

UNLABELLED: Radioluminescence microscopy is a new method for imaging radionuclide uptake by single live cells with a fluorescence microscope. Here, we report a particle-counting scheme that improves spatial resolution by overcoming the ß-range limit. METHODS: Short frames (10 µs-1 s) were acquired using a high-gain camera coupled to a microscope to capture individual ionization tracks. Optical reconstruction of the ß-ionization track (ORBIT) was performed to localize individual ß decays, which were aggregated into a composite image. The new approach was evaluated by imaging the uptake of (18)F-FDG in nonconfluent breast cancer cells. RESULTS: After image reconstruction, ORBIT resulted in better definition of individual cells. This effect was particularly noticeable in small clusters (2-4 cells), which occur naturally even for nonconfluent cell cultures. The annihilation and Bremsstrahlung photon background signal was markedly lower. Single-cell measurements of (18)F-FDG uptake that were computed from ORBIT images more closely matched the uptake of the fluorescent glucose analog (Pearson correlation coefficient, 0.54 vs. 0.44, respectively). CONCLUSION: ORBIT can image the uptake of a radiotracer in living cells with spatial resolution better than the ß range. In principle, ORBIT may also allow for greater quantitative accuracy because the decay rate is measured more directly, with no dependency on the ß-particle energy.

Single-fraction simulation of relative cell survival in response to uniform versus hypoxia-targeted dose escalation.

The purpose of this study was to investigate the increase in cell kill that can be achieved by tumor irradiation with heterogeneous dose distributions targeting hypoxic regions that can be visualized with non-invasive imaging. Starting with a heterogeneous distribution of microvessels, a microscopic two-dimensional model of tumor oxygenation was developed using planar simulation of oxygen diffusion. Non-invasive imaging of hypoxia was simulated taking partial volume effect into account. A dose-modulation scheme was implemented with the goal of delivering higher doses to the hypoxic pixels, as seen in simulated hypoxia images. To determine the relative cell kill in response to hypoxia-targeting irradiation, tumor cell survival fractions were compared to those resulting from treatments delivering the same average dose to the lesion in a spatially uniform fashion. It was shown that hypoxia-targeting dose modulation may be better suited for tumors with low α/ß, low hypoxic fraction and spatially aggregated hypoxic features. Most importantly, it was determined that at low fraction doses there is no cell kill increase from targeting hypoxic regions alone versus escalating the total tumor dose. However, for higher doses per fraction (≥8 Gy/fraction), the effectiveness of hypoxia-targeting irradiation increases, resulting in the tumoricidal effect of up to 30% higher than that of uniform tumor irradiation delivering the same average tumor dose.

Tumour microenvironment heterogeneity affects the perceived spatial concordance between the intratumoural patterns of cell proliferation and 18F-fluorothymidine uptake.

BACKGROUND AND PURPOSE: PET imaging with (18)F-fluorothymidine ((18)F-FLT) can potentially be used to identify tumour subvolumes for selective dose escalation in radiation therapy. The purpose of this study is to analyse the co-localization of intratumoural patterns of cell proliferation with (18)F-FLT tracer uptake. MATERIALS AND METHODS: Mice bearing FaDu or SQ20B xenograft tumours were injected with (18)F-FLT, and bromodeoxyuridine (proliferation marker). Ex vivo images of the spatial pattern of intratumoural (18)F-FLT uptake and that of bromodeoxyuridine DNA incorporation were obtained from thin tumour tissue sections. These images were segmented by thresholding and Relative Operating Characteristic (ROC) curves and Dice similarity indices were evaluated. RESULTS: The thresholds at which maximum overlap occurred between FLT-segmented areas and areas of active cell proliferation were significantly different for the two xenograft tumour models, whereas the median Dice values were not. However, ROC analysis indicated that segmented FLT images were more specific at detecting the proliferation pattern in FaDu tumours than in SQ20B tumours. CONCLUSION: Highly dispersed patterns of cell proliferation observed in certain tumours can affect the perceived spatial concordance between the spatial pattern of (18)F-FLT uptake and that of cell proliferation even when high-resolution ex vivo autoradiography imaging is used for (18)F-FLT imaging.

Comprehensive approach to coregistration of autoradiography and microscopy images acquired from a set of sequential tissue sections.

UNLABELLED: Histopathologic validation of a PET tracer requires assessment of colocalization of the tracer with its intended biologic target. Using thin tissue section autoradiography, it is possible to visualize the spatial distribution of the PET tracer uptake and compare it with the distribution of the intended biologic target (as visualized with immunohistochemistry). The purpose of this study was to develop and evaluate an objective methodology for deformable coregistration of autoradiography and microscopy images acquired from a set of sequential tissue sections. METHODS: Tumor-bearing animals were injected with 3'-deoxy-3'-(18)F-fluorothymidine ((18)F-FLT), (14)C-FDG, and other markers of tumor microenvironment including Hoechst 33342 (blood-flow surrogate). After sacrifice, tumors were excised, frozen, and sectioned. Multiple stacks of sequential 8 µm sections were collected from each tumor. From each stack, the middle (reference) sections were used to obtain images of (18)F-FLT and (14)C-FDG uptake distributions using dual-tracer autoradiography. Sections adjacent to the reference were used to acquire all histopathologic data (e.g., images of cell proliferation, hematoxylin and eosin). Hoechst images were acquired from all sections. To correct for deformations and misalignments induced by tissue processing and image acquisition, the Hoechst image of each nonreference section was deformably registered to the reference Hoechst image. This transformation was then applied to all images acquired from the same tissue section. In this way, all microscopy images were registered to the reference Hoechst image. The Hoechst-to-autoradiography image registration was done using rigid point-set registration based on external markers visible in both images. RESULTS: The mean error of Hoechst to (18)F-FLT autoradiography registration (both images acquired from the same section) was 30.8 ± 20.1 µm. The error of Hoechst-based deformable registration of histopathologic images (acquired from sequential tissue sections) was 23.1 ± 17.9 µm. Total error of registration of autoradiography images to the histopathologic images acquired from adjacent sections was evaluated at 44.9 µm. This coregistration precision supersedes current rigid registration methods with reported errors of 100-200 µm. CONCLUSION: Deformable registration of autoradiography and histopathology images acquired from sequential sections is feasible and accurate when performed using corresponding Hoechst images.