CT Image Sets Research Articles

Skeletal Muscle Loss Is an Imaging Biomarker of Outcome after Definitive Chemoradiotherapy for Locally Advanced Cervical Cancer.

Purpose: This study investigates the association between body composition change during concurrent chemoradiotherapy (CCRT) and outcome in patients with locally advanced cervical cancer (LACC).Experimental Design: Pre- and posttreatment CT images of 245 patients with LACC who were treated between 2004 and 2015 were analyzed. Skeletal muscle index (SMI) and density (SMD), subcutaneous adipose tissue index (SATI), and visceral adipose tissue index (VATI) were measured from two sets of CT images at the level of the L3 vertebra. Sarcopenia and a low SMD were defined using published cut-off points. Predictors of overall survival (OS) and cancer-specific survival (CSS) were analyzed using Cox regression models.Results: The median follow-up was 62.7 (range, 7.3-152.3) months. Among the 245 patients, 127 (51.8%) had pretreatment sarcopenia, and 154 (62.9%) had a low SMD. SMI did not decrease significantly during CCRT, 0.6%/150 days [95% confidence interval (CI), -1.8-0.6; P = 0.35]. However, SMI loss during CCRT of >10.0%/150 days was independently associated with poorer OS (HR, 6.02; 95% CI, 3.04-11.93; P < 0.001) and CSS (HR, 3.49; 95% CI, 1.44-8.42; P = 0.006) when adjusted for FIGO stage, pathology, and treatment. Pretreatment sarcopenia and change of SMD, SATI, and VATI during CCRT were not associated with survival.Conclusions: Skeletal muscle measurements could be imaging biomarkers to predict outcomes for patients with LACC in clinical practice. Further studies are needed to determine whether multimodal interventions can preserve skeletal muscle mass and thereby improve survival. Clin Cancer Res; 24(20); 5028-36. ©2018 AACR.

Thoracic Organ Doses and Cancer Risk from Low Pitch Helical 4-Dimensional Computed Tomography Scans

Purpose To investigate the dose depositions to organs at risk (OARs) and associated cancer risk in cancer patients scanned with 4-dimensional computed tomography (4DCT) as compared with conventional 3DCT. Methods and Materials The radiotherapy treatment planning CT image and structure sets of 102 patients were converted to CT phantoms. The effective diameters of those patients were computed. Thoracic scan protocols in 4DCT and 3DCT were simulated and verified with a validated Monte Carlo code. The doses to OARs (heart, lungs, esophagus, trachea, spinal cord, and skin) were calculated and their correlations with patient effective diameter were investigated. The associated cancer risk was calculated using the published models in BEIR VII reports. Results The average of mean dose to thoracic organs was in the range of 7.82-11.84 cGy per 4DCT scan and 0.64-0.85 cGy per 3DCT scan. The average dose delivered per 4DCT scan was 12.8-fold higher than that of 3DCT scan. The organ dose was linearly decreased as the function of patients' effective diameter. The ranges of intercept and slope of the linear function were 17.17-30.95 and -0.0278--0.0576 among patients' 4DCT scans, and 1.63-2.43 and -0.003--0.0045 among patients' 3DCT scans. Relative risk of cancer increased (with a ratio of 15.68:1) resulting from 4DCT scans as compared to 3DCT scans. Conclusions As compared to 3DCT, 4DCT scans deliver more organ doses, especially for pediatric patients. Substantial increase in lung cancer risk is associated with higher radiation dose from 4DCT and smaller patients' size as well as younger age.

Automatic Mapping of CT Scan Locations on Computational Human Phantoms for Organ Dose Estimation.

To develop an algorithm to automatically map CT scan locations of patients onto computational human phantoms to provide with patient-specific organ doses. We developed an algorithm that compares a two-dimensional skeletal mask generated from patient CTs with that of a whole body computational human phantom. The algorithm selected the scan locations showing the highest Dice Similarity Coefficient (DSC) calculated between the skeletal masks of a patient and a phantom. To test the performance of the algorithm, we randomly selected five sets of neck, chest, and abdominal CT images from the National Institutes of Health Clinical Center. We first automatically mapped scan locations of the CT images on a computational human phantom using our algorithm. We had several radiologists to manually map the same CT images on the phantom and compared the results with the automated mapping. Finally, organ doses for automated and manual mapping locations were calculated by an in-house CT dose calculator and compared to each other. The visual comparison showed excellent agreement between manual and automatic mapping locations for neck, chest, and abdomen-pelvis CTs. The difference in mapping locations averaged over the start and end in the five patients was less than 1cm for all neck, chest, and AP scans: 0.9, 0.7, and 0.9cm for neck, chest, and AP scans, respectively. Five cases out of ten in the neck scans show zero difference between the average manual and automatic mappings. Average of absolute dose differences between manual and automatic mappings was 2.3, 2.7, and 4.0% for neck, chest, and AP scans, respectively. The automatic mapping algorithm provided accurate scan locations and organ doses compared to manual mapping. The algorithm will be useful in cases requiring patient-specific organ dose for a large number of patients such as patient dose monitoring, clinical trials, and epidemiologic studies.

Use of lung treatment plans to evaluate DIR algorithms.

The purpose of the study is to evaluate the accuracy of two deformable image registration algorithms by examining their influence on the dose summation results obtained using 4DCT (four dimensional computed tomography) dose distributions based on '4D' planned and '4D optimal' IMRT (intensity modulated radiation therapy) plans. For ten lung cancer patients, 4D step and shoot IMRT plans were produced. The breathing cycle was divided into ten parts and for each part a set of CT images was acquired. For each patient the treatment plan was copied to the CTs of each phase and subsequently recalculated. Each phase CT was then registered to the average intensity projection (AIP) CT using a deformable image registration (DIR) algorithm and the composite dose distribution was then calculated by summing up the deformed dose distributions from all the phases ('4D' treatment plan). The '4D optimal' treatment plan was created by producing an optimal plan on the CTs of each phase of the respiratory cycle and summing up the deformed dose distributions from all the phases. The results indicate that it is possible to map the dose distributions of different breathing phases in lung using DIR, and that different DIR methods and target characteristics (motion amplitude, size, location) affect the differences between original plan, '4D' and '4D optimal' dose distributions. Although the '4D optimal' plans were designed to achieve 95% target coverage, both of the used DIR methods failed to translate that coverage in some instances. The same variation between these methods was also observed in the '4D' plan comparison. This study shows that it is feasible to perform an acceptably accurate calculation of the composite deformed dose. However, it is important to account for tumor motion and body deformation especially when the tumor volume is small and/or located in the lower lobe of the lung.

OA090] A Study on the consistency of treatment planning and delivery for endometrial cancer treatments with multi-lumen cylinders for high dose rate brachytherapy

Purpose To develop a method to evaluate the cumulative dose when using Multi Lumen Cylinder (MLC) applicator in vaginal cuff brachytherapy. Methods A retrospective analysis of 25 patients treated in 2017 with the MLC applicator 3 cm diameter, were considered for this study. All patients received 5 fractions with 5 Gy each delivered twice a week and prescribed to the applicator’s surface for2/3 of the vaginal length. The CT scans of these patients, taken for treatment plan were separately imported into the treatment planning system and paired with the initial CT scan after completing the contouring. Two sets of CT images were fused at a time together with respective to the applicator, using mutual registration. Dosimetric evaluations were performed. The maximum doses received by the rectum wall, bladder wall, bowels and PTV on were analyzed and traced over the five fractions to determine the total dose distribution over the entire prescribed treatment. Results No contour of any of the OAR was exactly similar when the CT images were fused each other. Depending on the depth of the insertion the PTV varied minimal. Each plan was performed independently and cumulated 2 at a time until all 5 fractions were added to the initial fraction. There is a difference between the doses received by the OARs between treatments and between the points of maximum dose. The maximum dose varied between 15% and 25% in rectum wall and bladder neck. The minimum dose varied between 3% and 6% in rectum wall and bladder neck. The average dose varied between 10% and 22% in rectum and bladder. The cumulative treatment does do not indicate a total maximum dose exceeding the tolerances for the rectum and bladder. Conclusion The variations in volumes of the PTVs and isodoses are minor, indicating that the desired area receives the entire prescribed treatment. Rectum wall will not exceed 85% of the prescribed cumulative dose and the bladder neck will not exceed 75% of the prescribed dose.

P277] Investigation of mean heart dose reduction inter-fraction variation in left-sided breast cancer radiotherapy using active breath coordinator moderate deep inspiration breath hold

Purpose Left-sided breast radiotherapy is associated with increased risk of coronary artery disease. Active Breathing Coordinator (ABC) reduces radiation dose to the heart and preserves local control in patients with left breast cancer. The aim of this study is to estimate the dose reduction to the heart and its inter-fraction variation in moderate deep inspiration breath hold (mDIBH) breast radiotherapy using ABC (Elekta). Methods Ten patients with left-sided breast cancer with tangential radiotherapy plans and ten patients with VMAT plans performed using to Monaco (Elekta) TPS, where included in the study. All the patients were trained on ABC procedure to enhance patient compliance and to determine individual (mDIBH) levels, which was set at 70–80% of maximum inspiratory capacity. During simulation two sets of CT images were acquired for each patient, one with mDIBH with ABC system and the other with free breathing. Treatment plans were performed for both scans and the mean dose to the heart was recorded. Treatment was performed using the ABC system. 200 cone-beam CT (CBCT) scans were acquired for all patients investigated throughout their course of treatment to take into account inter-fraction patient set-up inaccuracies. CBCTs were exported to Monaco TPS for registration and mean heart dose determination after the application of the original patient plan. Results For all patients evaluated, compared to free breathing, mDIBH with ABC significantly reduced mean dose to the heart by 20% or greater independently of the planning technique (tangential fields or VMAT). This reduction remains the same with minor differences (on average 1 cGy per fraction) throughout the treatment course. Conclusion Moderate deep inspiration breath hold (mDIBH) breast radiotherapy using ABC (Elekta) can significantly reduce mean dose to the heart. Variation in mean heart dose was minor and mean dose remains significantly reduced throughout the treatment course without being affected by inter-fraction variations.

Computer simulation analysis of source‐detector position for percutaneously measured O2‐gas signal in a three‐dimensional preterm infant lung

Further improvements in the clinical care of our most vulnerable patients-preterm infants-are needed. Novel diagnostic and surveillance tools facilitate such advances. The GASMAS technique has shown potential to become a tool to, noninvasively, monitor gas in the lungs of preterm infants, by placing a laser source and a detector on the chest wall skin. It is believed that this technology will become a valuable clinical diagnostic tool for monitoring the lung function of these patients. Today, the technology is, for this application, in an early stage and further investigations are needed. In the present study, a three-dimensional computer model of the thorax of an infant is constructed, from a set of CT images. Light transport simulations are performed to provide information about the position dependence of the laser- and detector probe on the thorax of the infant. The result of the simulations, based on the study method and the specified model used in this work, indicates that measurement geometries in front and on the side of the lung are favorable in order to obtain a good gas absorption signal.

Detection of suspected brain infarctions on CT can be significantly improved with temporal subtraction images.

To assess whether temporal subtraction (TS) images of brain CT improve the detection of suspected brain infarctions. Study protocols were approved by our institutional review board, and informed consent was waived because of the retrospective nature of this study. Forty-two sets of brain CT images of 41 patients, each consisting of a pair of brain CT images scanned at two time points (previous and current) between January 2011 and November 2016, were collected for an observer performance study. The 42 sets consisted of 23 cases with a total of 77 newly developed brain infarcts or hyperdense artery signs confirmed by two radiologists who referred to additional clinical information and 19 negative control cases. To create TS images, the previous images were registered to the current images by partly using a non-rigid registration algorithm and then subtracted. Fourteen radiologists independently interpreted the images to identify the lesions with and without TS images with an interval of over 4 weeks. A figure of merit (FOM) was calculated along with the jackknife alternative free-response receiver-operating characteristic analysis. Sensitivity, number of false positives per case (FPC) and reading time were analyzed by the Wilcoxon signed-rank test. The mean FOM increased from 0.528 to 0.737 with TS images (p < 0.0001). The mean sensitivity and FPC improved from 26.5% and 0.243 to 56.0% and 0.153 (p < 0.0001 and p = 0.239), respectively. The mean reading time was 173 s without TS and 170 s with TS (p = 0.925). The detectability of suspected brain infarctions was significantly improved with TS CT images. • Although it is established that MRI is superior to CT in the detection of strokes, the first choice of modality for suspected stroke patients is often CT. • An observer performance study with 14 radiologists was performed to evaluate whether temporal subtraction images derived from a non-rigid transformation algorithm can significantly improve the detectability of newly developed brain infarcts on CT. • Temporal subtraction images were shown to significantly improve the detectability of newly developed brain infarcts on CT.

Monte Carlo verification of radiotherapy treatments with CloudMC

BackgroundA new implementation has been made on CloudMC, a cloud-based platform presented in a previous work, in order to provide services for radiotherapy treatment verification by means of Monte Carlo in a fast, easy and economical way. A description of the architecture of the application and the new developments implemented is presented together with the results of the tests carried out to validate its performance.MethodsCloudMC has been developed over Microsoft Azure cloud. It is based on a map/reduce implementation for Monte Carlo calculations distribution over a dynamic cluster of virtual machines in order to reduce calculation time. CloudMC has been updated with new methods to read and process the information related to radiotherapy treatment verification: CT image set, treatment plan, structures and dose distribution files in DICOM format. Some tests have been designed in order to determine, for the different tasks, the most suitable type of virtual machines from those available in Azure. Finally, the performance of Monte Carlo verification in CloudMC is studied through three real cases that involve different treatment techniques, linac models and Monte Carlo codes.ResultsConsidering computational and economic factors, D1_v2 and G1 virtual machines were selected as the default type for the Worker Roles and the Reducer Role respectively. Calculation times up to 33 min and costs of 16 € were achieved for the verification cases presented when a statistical uncertainty below 2% (2σ) was required. The costs were reduced to 3–6 € when uncertainty requirements are relaxed to 4%.ConclusionsAdvantages like high computational power, scalability, easy access and pay-per-usage model, make Monte Carlo cloud-based solutions, like the one presented in this work, an important step forward to solve the long-lived problem of truly introducing the Monte Carlo algorithms in the daily routine of the radiotherapy planning process.

Improved lung nodule diagnosis accuracy using lung CT images with uncertain class

Background and objective: Among all malignant tumors, lung cancer ranks in the top in mortality rate. Pulmonary nodule is the early manifestation of lung cancer, and plays an important role in its discovery, diagnosis and treatment. The technology of medical imaging has encountered a rapid development in recent years, thus the amount of pulmonary nodules can be discovered are on the raise, which means even tiny or minor changes in lung can be recorded by the CT images. This paper proposes a pulmonary nodule computer aided diagnosis (CAD) based on semi-supervised extreme learning machine(SS-ELM).Methods: First, the feature model based on the pulmonary nodules regions of lung CT images is established. After that, the same feature data sets have been put into ELM, support vector machine (SVM) methods, probabilistic neural network (PNN) and multilayer perceptron (MLP) so as to compare the performance of the methods. ELM turned out to have better performance in training time and testing accuracy compared with SVM, PNN and MLP. Then, we propose a pulmonary nodules computer aided diagnosis algorithm based on semi-supervised ELM (SS-ELM), which enables both certain class feature sets with labels and unlabeled feature sets to be input for training and computer aided diagnosing. This algorithm has provided a solution for the using of uncertain class data and improve the testing accuracy of benign and malignant diagnosis.Results: 1018 sets of thoracic CT images from the Lung Database Consortium and Image Database Resource Initiative (LIDC-IDRI) have been used in experiment in order to test the effectiveness of the algorithm. Compared with ELM, the pulmonary nodules CAD based on SS-ELM has better testing accuracy performance.Conclusions: We have proposed a pulmonary nodule CAD system based on SS-ELM, which achieving better generalization performance at faster learning speed and higher testing accuracy than ELM, SVM, PNN and MLP. The SS-ELM based pulmonary nodules CAD has been proposed to solve the problem of uncertain class data using.

Positioning accuracy and daily dose assessment for prostate cancer treatment using in-room CT image guidance at a proton therapy facility.

To evaluate the effectiveness of CT image-guided proton radiotherapy for prostate cancer by analyzing the positioning uncertainty and assessing daily dose change due to anatomical variations. Patients with prostate cancer were treated by opposed lateral proton beams based on a passive scattering method using an in-room CT image-guided system. The system employs a single couch for both CT scanning and beam delivery. The patient was positioned by matching the boundary between the prostate and the rectum's anterior region identified in the CT images to the corresponding boundary in the simulator images after bone matching. We acquired orthogonal kV x-ray images after couch movement and confirmed the body position by referring to the bony structure prior to treatment. In offline analyses, we contoured the targeted anatomical structures on 375 sets of daily in-room CT images for 10 patients. The uncertainty of the image-matching procedure was evaluated using the prostate contours and actual couch corrections. We also performed dose calculations using the same set of CT images, and evaluated daily change of dose-volume histograms (DVHs) to compare the effectiveness of the treatment using prostate matching to the bone-matching procedure. The isocenter shifts by prostate matching after bone matching were 0.5±1.8 and -0.8±2.6mm along the superior-inferior (SI) and anterior-posterior (AP) directions, respectively. The body movement errors (σ) after couch movement were 0.7, 0.5, and 0.3mm along the lateral, SI and AP direction, respectively, for 30 patients. The estimated errors (σ) in the prostate matching were 1.0 and 1.3mm, and, in conjunction with the movement errors, the total positioning uncertainty was estimated to be 1.0 and 1.4mm along the SI and AP directions, respectively. Daily DVH analyses showed that in the prostate matching, 98.7% and 86.1% of the total 375 irradiations maintained a dose condition of V95% >95% for the prostate and a dose constraint of V77% <18% for the rectum, whereas 90.4% and 66.1% of the total irradiations did so when bone matching was used. The dose constraint of the rectum and dose coverage of the prostate were better maintained by prostate matching than bone matching (P<0.001). The daily variation in the dose to the seminal vesicles (SVs) was large, and only 40% of the total irradiations maintained the initial planned values of V95% for high-risk treatment. Nevertheless, the deviations from the original value were -4±7% and -5±11% in the prostate and bone matching, respectively, and a better dose coverage of the SV was achieved by the prostate matching. The correction of repositioning along the AP and SI direction from conventional bone matching in CT image-guided proton therapy was found to be effective to maintain the dose constraint of the rectum and the dose coverage of the prostate. This work indicated that prostate cancer treatment by prostate matching using CT image guidance may be effective to reduce the rectal complications and achieve better tumor control of the prostate. However, an adaptive approach is desirable to maintain better dose coverage of the SVs.

A prospective gating method to acquire a diverse set of free-breathing CT images for model-based 4DCT

Breathing motion modeling requires observation of tissues at sufficiently distinct respiratory states for proper 4D characterization. This work proposes a method to improve sampling of the breathing cycle with limited imaging dose.We designed and tested a prospective free-breathing acquisition protocol with a simulation using datasets from five patients imaged with a model-based 4DCT technique. Each dataset contained 25 free-breathing fast helical CT scans with simultaneous breathing surrogate measurements. Tissue displacements were measured using deformable image registration. A correspondence model related tissue displacement to the surrogate. Model residual was computed by comparing predicted displacements to image registration results. To determine a stopping criteria for the prospective protocol, i.e. when the breathing cycle had been sufficiently sampled, subsets of N scans where 5 ⩽ N ⩽ 9 were used to fit reduced models for each patient. A previously published metric was employed to describe the phase coverage, or ‘spread’, of the respiratory trajectories of each subset. Minimum phase coverage necessary to achieve mean model residual within 0.5 mm of the full 25-scan model was determined and used as the stopping criteria. Using the patient breathing traces, a prospective acquisition protocol was simulated.In all patients, phase coverage greater than the threshold necessary for model accuracy within 0.5 mm of the 25 scan model was achieved in six or fewer scans. The prospectively selected respiratory trajectories ranked in the (97.5 ± 4.2)th percentile among subsets of the originally sampled scans on average. Simulation results suggest that the proposed prospective method provides an effective means to sample the breathing cycle with limited free-breathing scans. One application of the method is to reduce the imaging dose of a previously published model-based 4DCT protocol to 25% of its original value while achieving mean model residual within 0.5 mm.

Calculation of dose volume parameters and indices in plan evaluation of HDR interstitial brachytherapy by MUPIT in carcinoma cervix.

Evaluation of a HDR- interstitial brachytherapy plan is a challenging job. Owing to the complexities and diversity of the normalization and optimization techniques involved, a simple objective assessment of these plans is required. This can improve the radiation dose coverage of the tumour with decreased organ toxicity. To study and document the various dose volume indices and parameters required to evaluate a HDR interstitial brachytherapy plan by Volume normalization and graphical optimization using MUPIT (Martinez Universal Perineal Interstitial Template) in patients of carcinoma cervix. Single arm, retrospective study. 35 patients of carcinoma cervix who received EBRT and HDR brachytherapy using MUPIT, were selected. The dose prescribed was 4 Gray/Fraction in four fractions (16Gy/4) treated twice daily, at least 6 hours apart. CTV and OARs were delineated on the axial CT image set. Volume normalization and graphical optimization was done for planning. Coverage Index (CI), Dose homogeneity index (DHI), Overdose index (OI), Dose non-uniformity ratio (DNR), Conformity Index (COIN) and dose volume parameters i.e. D2cc, D1cc, D0.1cc of rectum and bladder were evaluated. SPSS version 16 was used. CI was 0.95 ± 1.84 which means 95% of the target received 100% of the prescribed dose. The mean COIN was 0.841 ± 0.06 and DHI was 0.502 ± 0.11. D2cc rectum and bladder was 3.40 ± 0.56 and 2.95 ± 0.62 respectively which was within the tolerance limit of this organs. There should be an optimum balance between these indices for improving the quality of the implant and to yield maximum clinical benefit out of it, keeping the dose to the OARs in limit. Dose optimization should be carefully monitered and an institutional protocol should be devised for the acceptability criteria of these plans.

43. CBCT dose calculation: accuracy assessment of four different methods

Introduction Cone Beam CT (CBCT) images aim at target volume repositioning before treatment fraction. These images may be used for dose calculation in order to dosimetrically quantify anatomical variations that occurs during treatment. The main limits are image quality, Hounsfield numbers (HU) consistency and CBCT reconstruction size. This study aims to assess accuracy of several CBCT dose calculation methods. Methods This study deals with four CBCT (XVI, Elekta) and CT image sets acquired on the same day in context of weekly dose planning of one H&N patient treated with VMAT technique. CT-CBCT sets were used to quantify CBCT dose calculation accuracy by investigating four methods: (i) use of a density to HU relation from phantom CBCT image (phantom HU-D relation), (ii) use of a density to HU relation from patient CBCT image (patient HU-D relation), (iii) density override by three classes threshold (air, soft tissues, bone) and (iv) elastic deformation (Admire, Elekta) between planning CT and CBCT image, by creating a pseudo-CT. After rigid registration, plan characteristics are projected on CBCT images (or pseudo-CT) with Pinnacle v.9.10 (Philips). Dose distribution comparison between CT-based (reference) and CBCT-based calculation was performed by a gamma analysis in 2D and 3D ( γ 2D and γ 3D, criteria: 2%/2 mm, 30% threshold). Then, dose differences were reported in terms of 95% target coverage, parotid glands mean dose (Dmean) and PRV spinal cord maximum dose (D2%). Results On average, γ 3D (percentage of points with gamma The gamma analysis showed location of discrepancy areas: near body contour, regarding shoulders and in presence of heterogeneities ( Fig. 1 ). These differences can be explained by the poor image quality and the HU inconsistency of CBCT images in some areas. The method with pseudo-CT seems to be the best to pass over these issues because densities come from planning CT. Conclusions The four evaluated methods show that dose differences can reach 5% for parotid glands and are lower than 2% for spinal cord and target volumes. The 3D gamma analysis results (2%/2 mm) are higher than 95% for three among four methods. However, additional uncertainties have to be taken into consideration in case of CBCT image delineation and/or calculation on truncated images. This preliminary study shows CBCT dose calculation feasibility to quantify impact of patient anatomical variations during treatment and trigger a new plan based on CT if necessary. The future work is real delivered dose monitoring, based on CBCT image, in perspective of dose guided adaptive radiotherapy.

Brain medical image diagnosis based on corners with importance-values

BackgroundBrain disorders are one of the top causes of human death. Generally, neurologists analyze brain medical images for diagnosis. In the image analysis field, corners are one of the most important features, which makes corner detection and matching studies essential. However, existing corner detection studies do not consider the domain information of brain. This leads to many useless corners and the loss of significant information. Regarding corner matching, the uncertainty and structure of brain are not employed in existing methods. Moreover, most corner matching studies are used for 3D image registration. They are inapplicable for 2D brain image diagnosis because of the different mechanisms. To address these problems, we propose a novel corner-based brain medical image classification method. Specifically, we automatically extract multilayer texture images (MTIs) which embody diagnostic information from neurologists. Moreover, we present a corner matching method utilizing the uncertainty and structure of brain medical images and a bipartite graph model. Finally, we propose a similarity calculation method for diagnosis.ResultsBrain CT and MRI image sets are utilized to evaluate the proposed method. First, classifiers are trained in N-fold cross-validation analysis to produce the best θ and K. Then independent brain image sets are tested to evaluate the classifiers. Moreover, the classifiers are also compared with advanced brain image classification studies. For the brain CT image set, the proposed classifier outperforms the comparison methods by at least 8% on accuracy and 2.4% on F1-score. Regarding the brain MRI image set, the proposed classifier is superior to the comparison methods by more than 7.3% on accuracy and 4.9% on F1-score. Results also demonstrate that the proposed method is robust to different intensity ranges of brain medical image.ConclusionsIn this study, we develop a robust corner-based brain medical image classifier. Specifically, we propose a corner detection method utilizing the diagnostic information from neurologists and a corner matching method based on the uncertainty and structure of brain medical images. Additionally, we present a similarity calculation method for brain image classification. Experimental results on two brain image sets show the proposed corner-based brain medical image classifier outperforms the state-of-the-art studies.

Quantification of emphysema using low-dose chest CT: Effect of iterative reconstruction

Background: Iterative reconstruction (IR) is widely used in low-dose CT examinations for reducing image noise and saving radiation dose. Lung density profile has been reported to be significantly influenced by IR. The aim of this study was to evaluate the changes of emphysema quantification according to variation of low-density thresholds, percentile of lung density, and volume of low-attenuation clusters in low-dose chest CT using IR comparing with those using filtered back projection (FBP). Methods: 50 patients (39 men and 11 women; mean age 56.7 years) who underwent low dose (20 mA, dose modulation) chest CT scans for lung cancer screening were included in the study. CT images were reconstructed using both FBP and IR (iDose; Philips Healthcare, Netherlands). Lung volume, mean lung density, low density volumes with the threshold of -950 HU (LDV-950) and -910 HU (LDV-910), 15th percentile of lung density, and volume of low attenuation clusters (cluster sizes 2–8,8–65, 66–186, >186 mm2) were measured in each set of low-dose CT images. The measurement results of two reconstruction methods were compared using concordance correlation coefficient (ρc) and a Bland-Altman plot. Findings: Although total lung volume and mean lung density showed almost perfect agreement (ρc=0.99), all other measurements showed poor agreement between IR and FBP (0.090–0.855). LDV-910 showed relatively better agreement (ρc=0.842) compared with LDV-950 (ρc=0.648) and 15th percentile (ρc=0.527). Mean differences of LDV-950, LDV-910, and 15th percentile measured in two algorithms were 3.9%, 3.3%, and 28.7 HU. Mean differences of volume of low attenuation clusters were 2.1%, 3.0%, 0.6%, and 1.9%, respectively. Interpretation: Measurement of emphysema quantification with low-dose CT using IR showed significantly different results than with those using FBP. Low-density volume using higher threshold showed better agreement. Funding: None. Conflicts of interest: We declare that we have no conflicts of interest.

4D CBCT image guidance in liver SBRT

To analyze 4D CBCT image guided accuracy and dose volume histograms in SBRT for liver tumors. The treatment with SBRT in hepatic cancer using 4D image guidance is analyzed in this work. Patients are inmovilized employing an arm support, knee fix and a dampening system. Respiration correlated 4D CT and diagnostic CT performed using IV contrast media are used to delineate gross tumor volume (GTV). GTVs in eight 4DCT sets and two contrast sets are combined to create an internal target volume (ITV). Comparing center of mass of GTV in every respiratory phase a dampening device is selected and a set of CT images is chosen as reference for planning and simulation. In this reference CT organs at risk are contoured and it is selected to match in the daily image guided radiotherapy (IGRT). IGRT using anatomy fiducials respiration correlated 4DCBCT with automatic registration of the planning reference CT is used to setup patient in every treatment session. Results were compared for three registration techniques: 3D CBCT using mean ITV and soft tissue registration, 4dCBCT clipbox and bone anatomy registration (4D bCBCT), and 4D soft tissue registration using a 0.5 cm mask from ITV (4D mCBCT). Significant differences in ITV localization were observed between 4D mCBCT and the other modalities, up to 16.0 mm ± 6.0 mm (3D vector). The differences between 4D bCBCT and 4D mCBCT are closer than 4DbCBCT and 3D CBCT. These differences are larger when respiratory amplitude increase. Mean respiratory amplitude was 21.1 mm ± 7.0 mm when measured in simulation CT, in agreement with 20.0 ± 7.8 mm when measured in 4D CBCT. Visual comparison of reference respiratory curve and treatment daily respiratory curve makes possible to asses reproducibility of patients and tumor position reproducibility. Interobserver variability was reduced using 4D mCBCT and soft tissue automatic registration. With this protocol we can assert that interpatient variability in dose volume histograms is due mainly to treatment planning technique and tumor volume definition, and not directly correlated with motion amplitude. Respiration correlated 4D CBCT with soft tissue registration improves accuracy of IGRT because of the more precise target localization in the presence of respiratory motion, allowing the comparison of reference respiratory curve with treatment respiratory curve. Moreover, this method enables the realization of 4D CBCT without external surrogates with good correlation with reference 4D CT using external sensors.

Can Robust Optimization for Range Uncertainty in Proton Therapy Act as a Surrogate for Biological Optimization?

To investigate how robust optimization (to address range uncertainties) affects the dose and linear energy transfer (LET) distribution in the target and organs at risk (OARs) compared to LET-based inverse optimization (to address relative biological effectiveness - RBE - effects) in intensity modulated proton therapy (IMPT). A GPU-based Monte Carlo code was used to calculate both physical dose and dose-averaged LET for all pencil beams for IMPT proton plans for previously treated ependymoma patients. The plans were initially optimized based on physical dose constraints. A previously developed LET-based optimization algorithm, using a prioritized optimization scheme, was then applied to modify the LET distribution while constraining the physical dose objectives within 3% of the initial plan. The objective functions for this step were based on the product of LET and physical dose (LETxD) as an approximation of the additional biological dose that is caused by high LET. For the robust optimization all pencil beam dose and dose-averaged LET calculations were repeated for patient CT image sets with modified HU values, resulting in uniform ±2.33% and ±4.77% range deviations, for a total of five different scenarios. The level of range deviation and probability of these scenarios was estimated assuming a Gaussian-shaped probability of range uncertainty, with 1.5 sigma value of 3.5%, which is the clinically used range uncertainty in our institution. The dose and LETxD distributions resulting from the initial, LET-based and robust optimizations were then compared in terms of target coverage and OAR sparing. Our results show that both LET-based and robust optimization techniques of IMPT plans resulted in significantly reduced maximum LETxD values in the OARs adjacent to the target compared to the initial plan. Robust optimization avoids dose gradients in beam direction in the dose contributions of individual beams, effectively resembling single field uniformly optimized plans. For an ependymoma patient (Table 1) this results in higher median LETxD values for both the target and brainstem, compared to the initial plan. LET-based optimization most effectively reduced the median and mean LETxD to the brainstem, however also resulted in somewhat smaller median LETxD value for the target. Both inverse LET-based and robust optimization techniques can be a realistic approach to manage biological dose hot spots in the OARs adjacent to the target. Depending on the clinical dosimetric priorities set for each patient case, each one can offer distinct benefits towards avoiding potentially increased risk of side effects from elevated RBE of proton beams near the end of range.Abstract 230; Table 1Initial planLET-reoptimized planRobustly optimized planTargetD98 (Gy)484848.2D50 (Gy)50.750.650.9Median LETxD4.64.25.5BrainstemD50 (Gy)49.749.449.5Median LETxD4.83.85.7Max LETxD (2%)9.86.78.5 Open table in a new tab

The effect of disease and respiration on airway shape in patients with moderate persistent asthma.

Computational models of gas transport and aerosol deposition frequently utilize idealized models of bronchial tree structure, where airways are considered a network of bifurcating cylinders. However, changes in the shape of the lung during respiration affect the geometry of the airways, especially in disease conditions. In this study, the internal airway geometry was examined, concentrating on comparisons between mean lung volume (MLV) and total lung capacity (TLC). A set of High Resolution CT images were acquired during breath hold on a group of moderate persistent asthmatics at MLV and TLC after challenge with a broncho-constrictor (methacholine) and the airway trees were segmented and measured. The airway hydraulic diameter (Dh) was calculated through the use of average lumen area (Ai) and average internal perimeter (Pi) at both lung volumes and was found to be systematically higher at TLC by 13.5±9% on average, with the lower lobes displaying higher percent change in comparison to the lower lobes. The average internal diameter (Din) was evaluated to be 12.4±6.8% (MLV) and 10.8±6.3% (TLC) lower than the Dh, for all the examined bronchi, a result displaying statistical significance. Finally, the airway distensibility per bronchial segment and per generation was calculated to have an average value of 0.45±0.28, exhibiting high variability both between and within lung regions and generations. Mixed constriction/dilation patterns were recorded between the lung volumes, where a number of airways either failed to dilate or even constricted when observed at TLC. We conclude that the Dh is higher than Din, a fact that may have considerable effects on bronchial resistance or airway loss at proximal regions. Differences in caliber changes between lung regions are indicative of asthma-expression variability in the lung. However, airway distensibility at generation 3 seems to predict distensibility more distally.

The accuracy of treatment planning system dose modelling in the presence of brass mesh bolus.

This work assesses the dosimetric accuracy of three commercial treatment planning system (TPS) photon dose calculation algorithms in the presence of brass mesh used as a bolus. Bolus material is used in radiotherapy to provide dose build-up where superficial tissues require irradiation. They are generally water equivalent but high density materials can also be used. Dose calculations were performed on Monaco and Masterplan TPS (Elekta AB, Sweden) using phantoms defined by the three DICOM CT image sets of water equivalent blocks (no bolus, 1 layer and 2 layers of brass mesh) exported from the CT scanner. The effect of the mesh on monitor units, build-up dose, phantom exit dose and beam penumbra were compared to measured data. Dose calculations for 6 and 15MV photon beams on plain water equivalent phantoms were seen to agree well with measurement validating the basic planning system algorithms and models. Dose in the build-up region, phantom exit dose and beam penumbra were poorly modelled in the presence of the brass mesh. The beam attenuation created by the bolus material was overestimated by all three calculation algorithms, at both photon energies, e.g. 1.6% for one layer and up to 3.1% for two layers at 6MV. The poor modelling of the physical situation in the build-up region is in part a consequence of the high HU artefact caused by the mesh in the CT image. CT imaging is not recommended with the brass mesh bolus in situ due to the poor accuracy of the subsequent TPS modelling.