Volumetric Comparison Research Articles

Quantitative Volumetric Comparison of Direct Oral Anticoagulant and Vitamin K Antagonist Treatment for Pulmonary Thrombus Reduction During the Acute Phase in Symptomatic Patients

Recent clinical trials' findings have revealed the therapeutic noninferiority of direct oral anticoagulant (DOAC) to standard therapy with vitamin K antagonist (VKA) in patients with pulmonary thromboembolism (PTE). However, few studies have quantitatively analyzed thrombus reduction in the pulmonary artery. This observational study included 38 symptomatic PTE patients with stable hemodynamics. All patients received an intravenous heparin bolus followed by continual heparin injections immediately after the PTE diagnosis. The heparin was discontinued after edoxaban therapy began in the DOAC group (n=22) or after the therapeutic range for the prothrombin time-international normalized ratio was achieved in the VKA group (n=16). The thrombus volumes in the pulmonary arteries were quantitatively analyzed using contrast-enhanced computed tomography scans, and they were compared at baseline and at 2 weeks after admission. The pulmonary thrombus volumes declined in the VKA and DOAC groups from 7.9 to 4.2 cm3 (P=0.048) and from 7.1 to 3.7 cm3 (P < 0.01), respectively, and the thrombus reduction rates did not differ significantly between the groups (-34% vs. -64%, respectively; P=0.38). The fibrinogenolysis parameter changes during the14 days after admission were similar in both groups. Compared with the VKAgroup, the average hospital stay was 9days shorter in the DOAC group. There were no in-hospital deaths, and 1 case experienced major bleeding in the VKA group. In relation to pulmonary artery thrombus volume reduction, DOAC monotherapy for PTE may be comparable with standard therapy involving VKAs.

Longitudinal MEMRI analysis of brain phenotypes in a mouse model of Niemann-Pick Type C disease

Niemann-Pick Type C (NPC) is a rare genetic disorder characterized by progressive cell death in various tissues, particularly in the cerebellar Purkinje cells, with no known cure. Mouse models for human NPC have been generated and characterized histologically, behaviorally, and using longitudinal magnetic resonance imaging (MRI). Previous imaging studies revealed significant brain volume differences between mutant and wild-type animals, but stopped short of making volumetric comparisons of the cerebellar sub-regions. In this study, we present longitudinal manganese-enhanced MRI (MEMRI) data from cohorts of wild-type, heterozygote carrier, and homozygote mutant NPC mice, as well as deformation-based morphometry (DBM) driven brain volume comparisons across genotypes, including the cerebellar cortex, white matter, and nuclei. We also present the first comparisons of MEMRI signal intensities, reflecting brain and cerebellum sub-regional Mn2+-uptake over time and across genotypes.

Lipid-weighted intraoperative photoacoustic tomography of breast tumors: Volumetric comparison to preoperative MRI

With a lifetime risk of 1 in 8, breast cancer continues to be a major concern for women and their physicians. The optimal treatment of the disease depends on the stage of the cancer at diagnosis, which is typically assessed using medical imaging. However, currently employed imaging systems for breast tumor measurement rarely agree perfectly.Our group developed an Intraoperative Photoacoustic Screening (iPAS) soft tissue scanner featuring high bulk tissue sensitivity, a clinically compatible scan-time of 6 min, imaging depths greater than 2 cm and the capability to visualize whole breast tumors based on their lipid, rather than hemoglobin, profile. Here, we report on the first clinical experience with breast cancer patients by comparing tumor-measurement using iPAS, preoperative dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) and gold-standard pathology. Tumor size was measured volumetrically for iPAS and DCE-MRI, and separately using maximum diameters for pathology, DCE-MRI and iPAS. Comparisons were performed using Pearson’s correlation coefficients, and the non-parametric Wilcoxon signed-rank test.Twelve consecutive patients were included in the study, contingent on pathologically documented invasive carcinoma. iPAS volumetric tumor size was positively correlated to DCE-MRI (Pearson’s r = 0.78, p = 0.003) and not significantly different (Wilcoxon, p = 0.97). In comparison to pathology, tumor diameters given by iPAS were positively correlated (Pearson’s r = 0.87, p = 0.0002) and significantly different (Wilcoxon, p = 0.0015).The results indicated that volumetric-measurement of invasive breast tumors with iPAS is similar to that of DCE-MRI. On the other hand, tumor diameter measurements were less reliable. Beyond enhancing surgical specimen examination, an extension of this technology to diagnostic imaging promises a new perspective on tumor assessment, potentially improving our current understanding and treatment of breast cancer.

Intraoperative Navigation and Cone Beam Computed Tomography for Restoring Orbital Dimensions: A Single-Center Experience.

Correction of post-traumatic orbital defects remains a challenge for the maxillofacial surgeon. We examined the added value of combined intraoperative (IO) navigation and IO cone beam computed tomography (CBCT). A retrospective cohort study was performed in all consecutive patients requiring unilateral post-traumatic orbital surgery between January 2012 and December 2018. Patients were divided into 3 groups: IO navigation (NAV), IO-CBCT (CBCT), and IO navigation with IO-CBCT (NAV-CBCT). A detailed description of our workflow is provided. Volumetric comparison of the operated orbit to the contralateral orbit was made with Brainlab. Of the 81 cases, 22 patients were included (12 males/10 females) with a mean age of 51 years. Three patients were assigned to NAV, 6 to CBCT, and 13 to NAV-CBCT. The reconstructed orbital volume did not significantly differ from the contralateral orbital volume within the 3 groups. The mean difference between the contralateral and the operated orbit was 3.05 cm3, 3.72 cm3, and 1.47 cm3 for NAV, CBCT, and NAV-CBCT, respectively, where only NAV-CBCT showed a significant smaller volumetric difference in comparison to CBCT alone. Gender or age did not correlate with difference in orbital volume. Normal function and aesthetics was seen at 6 weeks postoperative in 0 of 3, 6of 6, and 6 of 13 patients of the NAV, CBCT, and NAV-CBCT, respectively. Orbital defects can be treated effectively using IO navigation. Although our data could not demonstrate a significant added value of IO-CBCT in cases where IO navigation was used based on volumetric difference alone, the combination of IO-CBCT and IO navigation seems to give the best results considering both volumetric difference and postoperative function and aesthetics. Confirmation in a prospective, randomized trial with a larger sample size is required.

Assessment of facial asymmetry in individuals having skeletal Class II malocclusion in Uttar Pradesh population: A cephalometric study

Objective: The objective was to assess facial asymmetry in individuals having skeletal Class II malocclusion in Uttar Pradesh population. Materials and Methods: A total of sixty individuals (thirty males and thirty females) between 18 and 27 years of age were selected. The pretreatment lateral cephalograms of the samples selected (n = 60) for the study were divided into two groups, Group I and Group II, after which posteroanterior cephalograms were taken for the measurement of asymmetry. All the cephalometric parameters were defined as quantitative variables. The mean and standard deviation (SD) for each measurement was calculated. The results were presented in frequencies, percentages for qualitative data, and mean ± SD for quantitative data. Paired t-test was used to test the significance (P = 0.05 or less) in the difference between the right and left sides of the face. Chi-squared test was used to check the significance of difference in proportions. All the analyses were carried out on MS-Excel and SPSS 16.0 version (Chicago, Inc., Illinois, USA). Results: In Group II, the parameters for mandibular morphology and volumetric comparison and the mean of all parameters taken were statistically insignificant, but the mean for condylion-antegonial notch was greater for the right side, and the difference between the left and right sides was statistically significant (P = 0.019). When comparing the cephalometric parameters between the left and right sides in Group I and Group II, none of the mean values were proved to be statistically significant, but the right side parameters revealed to be greater than the left side of the measured parameters. Conclusion: All participants showed mild skeletal asymmetry on posteroanterior cephalograms, which was not statistically significant. In Group II, the relative mean of condylar asymmetry was statistically significant, which proves the presence of some amount of condylar asymmetry in this group. The other cephalometric parameters measured in our study showed that the value of the same was greater for the right side than the left side.

Volumetric Comparison by Magnetic Resonance Imaging at Diagnosis and Time of Brachytherapy for Recurrent Endometrial Cancer and Vaginal Cancer

Volumetric Comparison by Magnetic Resonance Imaging at Diagnosis and Time of Brachytherapy for Recurrent Endometrial Cancer and Vaginal Cancer

Response control correlates of anomalous basal ganglia morphology in boys, but not girls, with attention-deficit/hyperactivity disorder

Anomalous basal ganglia morphology may contribute to deficient motor response control in children with attention-deficit/hyperactivity disorder (ADHD). This study expands upon recent evidence of sex differences in subcortical morphology and motor response control deficits among children with ADHD to examine basal ganglia volume and shape in relation to motor response control. Participants included 8–12 year-old children with ADHD (n = 52, 21 girls) and typically developing (TD) controls (n = 45, 19 girls). High resolution T1-weighted 3D MPRAGE images covering the whole brain were acquired for all participants on a 3 T scanner. Participants performed two computer-based go/no-go tasks that differed in the extent to which working memory was necessary to guide response selection. Shape-based morphometric analyses were performed in addition to traditional volumetric comparisons and correlations with measures of motor response control were examined. Boys with ADHD consistently demonstrated increased commission error rate and response variability, regardless of task demands, suggesting broad response control deficits. In contrast, response control deficits among girls with ADHD varied depending on task demands and performance measures. Volumetric reductions and inward deformation (compression) on the dorsal surface of the globus pallidus and within subregions of the putamen receiving projections from limbic, executive and motor cortices were observed in boys, but not girls, with ADHD relative to TD children. Mediation analyses revealed that putamen and globus pallidus volumes mediated the relationship between diagnosis and commission error rate. Furthermore, reduced volumes of these structures and localized inward deformation within executive and motor circuits correlated with poorer response control, particularly under conditions of increased cognitive load. These findings suggest that anomalous basal ganglia morphology is related to impaired motor response control among boys with ADHD.

Pre-Operative Versus Post-Operative Radiosurgery of Brain Metastases-Volumetric and Dosimetric Impact of Treatment Sequence and Margin Concept.

Background: Pre-operative radiosurgery (SRS) preceding the resection of brain metastases promises to circumvent limitations of post-operative cavity SRS. It minimizes uncertainties regarding delineation and safety margins and could reduce dose exposure of the healthy brain (HB). Methods: We performed a systematic treatment plan comparison on 24 patients who received post-operative radiosurgery of the resection cavity at our institution. Comparative treatment plans were calculated for hypofractionated stereotactic radiotherapy (7 × 5 Gray (Gy)) in a hypothetical pre-operative (pre-op) and two post-operative scenarios, either with (extended field, post-op-E) or without the surgical tract (involved field, post-op-I). Detailed volumetric comparison of the resulting target volumes was performed, as well as dosimetric comparison focusing on targets and the HB. Results: The resection cavity was significantly smaller and different in morphology from the pre-operative lesion, yielding a low Dice Similarity Coefficient (DSC) of 53% (p = 0.019). Post-op-I and post-op-E targets showed high similarity (DSC = 93%), and including the surgical tract moderately enlarged resulting median target size (18.58 ccm vs. 22.89 ccm, p < 0.001). Dosimetric analysis favored the pre-operative treatment setting since it significantly decreased relevant dose exposure of the HB (Median volume receiving 28 Gy: 6.79 vs. 10.79 for pre-op vs. post-op-E, p < 0.001). Dosimetrically, pre-operative SRS is a promising alternative to post-operative cavity irradiation that could furthermore offer practical benefits regarding delineation and treatment planning. Comparative trials are required to evaluate potential clinical advantages of this approach.

Hurdles in radiation planning for glioblastoma: Can delayed-contrast enhanced computed tomography be a potential solution?

Conformal radiation is the standard of care in treatment of glioblastoma. Although co-registration of magnetic resonance imaging (MRI) with early contrast enhanced computed tomography (CECT) is recommended for target delineation by consensus guidelines, ground realities in developing countries often result in availability of less-than-ideal MR sequences for treatment planning. Purpose of this study is to analyze the impact of incorporation of delayed-CECT sequences for radiation planning in glioblastomas, as an adjunct or alternative to MRI. Case records of all patients of glioblastoma treated at our center between 2011 and 2014 were retrospectively evaluated. Gross treatment volumes were delineated on T1 contrast MRI (m-GTV), early CECT (e-GTV) and delayed CECT (d-GTV); volumetric comparisons were made using repeated measures analysis of variance and pair-wise analysis. Although 96% of registered patients underwent postoperative MRI, only 38% of them had desirable sequences suitable for co-registration. Median duration between acquisition of postoperative MRI and surgery was 45 days (range, 33-60), whereas that between MRI and treatment-planning CT was 5 days (range, 1-10). Statistically significant differences (P<0.0001) were obtained between mean volumes of e-GTV (41.20cc), d-GTV (58.09cc) and m-GTV (60.52cc). Although the mean GTV increased by 46% between early CECT and MRI, the difference was only 4% between delayed CECT and MRI. Delayed CECT is superior to early CECT for co-registration with MRI for target delineation, especially when available MR sequences are less-than-ideal for treatment planning, and can be considered as the most appropriate adjunct as well as an alternative to MRI, compared to early CECT.

9 Development of an experimental tool to evaluate normal tissue 3D doses in external-beam radiotherapy

Introduction Recently, cancer treatment improvement enabled an increase in patients’ survival rate. In more than half of the cases, radiotherapy participates in the cure for cancer. The efficacy of this technique is well known but it inevitably delivers doses to normal tissues exposing them to radio-induced complications risks [1] . Moreover, the longer patients’ life expectancy increases the risk of developing long-term complications such as second cancers. Thus, evaluating the risks following cancer treatment involving radiotherapy constitutes a major challenge. In this context, an experimental tool has been developed to evaluate normal tissue doses delivered by the different radiotherapy techniques. Those dosimetric data are prerequisites for risk evaluation. Methods The tool developed on Matlab is based on 3D dose reconstruction from 2D dose measurements performed using radiochromic films placed between phantom slices. The tool proceeds to fine spatial films/TPS dose matching near the target volume and then reconstructs the 3D dose distribution by interpolation of films dose measurements. Finally, interesting dosimetric data such as dose-volume histograms are evaluated. A first validation of the tool has been performed with the EasyCube phantom irradiated with two films spaced 2 cm apart and placed in target volume. After scanning and contouring, the phantom was irradiated according to a treatment planning with the films in place. Films were analyzed using a rigorous protocol [2] developed by our group. Results Both 2D and volumetric comparisons between measured and planned doses gave good results. Indeed, 89.2% of points pass the 3 mm/3% gamma test when comparing interpolated dose distribution and planned dose on the middle slice. Likewise, a 2.6% discrepancy on mean PTV doses has been observed between the two distributions and the mean shift between the 0.4 Gy isodoses is of 0.10 mm on the middle slice. Conclusions The original tool developed in this work enables an experimental evaluation of normal tissue doses: essential dosimetric data for risk evaluation models. The preliminary feasibility study shows promising results. Validation and optimization on a more important volume is currently ongoing. Finally, validation using 3D gel dosimetry is considered.

Effects of four‐dimensional simulation of boost radiation after breast conservation surgery: A prospective dosimetric and volumetric analysis

AbstractObjectiveBreast conservation therapy is the standard of care for early breast cancer. After whole breast irradiation, lumpectomy cavity boost increases local control and survival. However, accurate identification of the tumor bed and internal displacement caused by respiratory movements poses major challenges in achieving the optimal therapeutic index. Although surgical clips are considered accurate surrogates for localization of the tumor bed, no consensus has been reached regarding the use of motion management during treatment delivery. The purpose of the present study was to evaluate the dosimetric and volumetric effects of respiratory motion on the delivery of boost radiation using 4‐D computed tomography simulation.MethodsA total of 25 patients undergoing tumor bed boost radiation were recruited prospectively. Dosimetric and volumetric comparisons were made between conformal plans formed with 3‐D and 4‐D datasets obtained from each patient.ResultsNo significant differences were observed in target volumes, degrees of inclusion, conformity indices, or homogeneity indices between the plans formed with the two datasets (P &lt; 0.05). The magnitude of internal movement of the tumor bed was insignificant compared with the margin for setup errors.ConclusionDisplacement of the tumor bed caused by respiratory motion was minimal with insignificant dosimetric or volumetric effects, and does not warrant the routine use of 4‐D simulation or the generation of internal target volumes.

OA075] Prediction of recurrences using diffusion driven growth modelling for glioblastoma

Purpose Clinical target volume (CTV) for radiotherapy of glioblastoma is typically defined as an isotropic 2 cm expansion to the area of residual enhancement on T1 weighted MR imaging plus the surgical bed. Tumor growth models allow for incorporating the biological information about preferential spread along the white matter fibers in target definition. The aim of this study was to assess the value of an image driven CTV in the prediction of non-central recurrences. Methods Out of 56 glioblastoma patients treated with radiotherapy (dose 60 Gy in 30 fractions), five had recurrences extending outside the 95% isodose. Each patient’s brain was segmented into white matter, grey matter and cerebrospinal fluid (CSF). Anatomic boundaries were delineated manually. A reaction–diffusion growth model was used assuming uniform proliferation throughout the brain and a difference in white and grey matter diffusion of a factor of 10 with no migration across the barriers or in CSF. The volume of the image driven CTV (iCTV) was fixed to that of the standard CTV. The volumetric comparison was performed using the shortest distance (SD) between all surface points in the recurrence volume and respective CTV and iCTV. Results Three of five recurrences were distant, seperated at minimum of 3 cm from the resection cavity or the residual enhancement, in the temporal, parietal and frontal lobe. The median CTV and recurrence volume was 236 cm 3 (range: 175–240 cm 3 ) and 89 cm 3 (range: 1–112 cm 3 , respectively. The median SD from the recurrence volume decreased from 7.0 mm (range: 4.7–36.0 mm) for CTV to 5.5 mm (range: 2.9–28.1 mm) for iCTV indicating that iCTV was closer to the recurrence volume. However, distant recurrences were not included in either CTV or iCTV. Conclusions The image driven CTV appears promising in the prediction of marginal recurrences outside 95% isodose compared to standard CTV. However this population was limited in size and validation is warranted in a larger cohort.

Volumetric Comparison by Magnetic Resonance Imaging at Diagnosis and Time of Brachytherapy for Recurrent Endometrial Cancer and Vaginal Cancer

Volumetric Comparison by Magnetic Resonance Imaging at Diagnosis and Time of Brachytherapy for Recurrent Endometrial Cancer and Vaginal Cancer

Brain metastasis growth on preradiosurgical magnetic resonance imaging

PurposeA previous analysis showed that brain metastases that are treated with frameless stereotactic radiation surgery (SRS) and planned with magnetic resonance imaging (MRI) >14 days before SRS had worse local control (LC). To evaluate if worse LC may be due to unaccounted interval metastasis growth and radiosurgical marginal miss, we quantified growth before SRS on preradiosurgical imaging. Methods and materialsWe reviewed data from patients who were treated with fixed-frame SRS for brain metastases at our institution between 2010 and 2013 and had pretreatment diagnostic brain MRI and SRS-planning MRI scans available. Metastases were contoured on the pretreatment MRI scan and the day-of-treatment planning MRI scan for volumetric comparison. Growth rates were calculated. Serial volumetric contour expansions on the pretreatment MRI scans were used to determine the minimum margin necessary to encompass the entire metastasis on day of the SRS. LC was estimated by Kaplan-Meier method. ResultsAmong 411 brain metastases in 165 patients, the time between pretreatment and treatment MRI was associated with metastasis growth (P < .001) with a mean growth rate of 0.02 ml/day (95% confidence interval, 0.01-0.03) and a 1.35-fold volume increase at 14 days. Time between MRI scans was associated with the amount of margin that was needed to target the entire brain metastasis volume on the day of the SRS (P < .001), as were volume of metastasis on the pre-treatment MRI (P < .001) and melanoma histology (P < .001). LC was not associated with growth rate among patients who underwent fixed-frame SRS. ConclusionsTime between pretreatment MRI and SRS is associated with brain metastasis growth, but LC is not compromised when patients receive fixed-frame SRS with same-day MRI planning. Margins may be needed for metastases that are treated with frameless SRS to account for growth between the planning MRI and SRS delivery. SummaryIn this study, we quantify brain metastasis growth over time by taking advantage of the availability of 2 pretreatment magnetic resonance imaging scans taken at 2 time points among patients treated with frame-fixed radiation surgery. We found that metastasis growth is associated with time, initial metastasis size, melanoma histology, and concurrent chemotherapy. Performing serial margin expansions demonstrated factors that are associated with the amount of margin that is needed to target the entire metastasis on the day of radiation surgery.

Volumetric and dosimetric comparison of organs at risk between the prone and supine positions in postoperative radiotherapy for prostate cancer

BackgroundThe aim of this study was to evaluate the effects of patient positioning on the volume of organs at risk (OARs) in or near the planning target volume (PTV) and the dose distribution in adjuvant or salvage radiotherapy for prostate cancer after prostatectomy.MethodsSeventeen patients who received intensity-modulated radiation therapy (66 Gy in 33 fractions) as adjuvant or salvage therapy after prostatectomy were evaluated. All patients underwent CT scans in both the prone (on a belly board) and supine positions. The target volumes and OARs were delineated on each CT series. The planning target volume (PTV) was extended in every direction to generate the PTV + 0.5 cm, PTV + 1 cm, PTV + 2 cm, PTV + 3 cm, and PTV + 4 cm values. The volumes of the OARs overlapping with the PTV and the extended target volumes in the prone and supine position were compared using the Wilcoxon signed-rank test. Dose-volume histogram (DVH) parameters in the prone and supine position were compared using the paired t-test.ResultsThe mean overlapping volumes of the small intestine for each of the PTV values were as follows (prone position vs. supine position [mean ± SD]): PTV, 1.5 ± 5.5 cm3 vs. 7.9 ± 15.7 cm3 (P = 0.028); PTV + 0.5 cm, 2.6 ± 8.9 cm3 vs. 12.1 ± 22.6 cm3 (P = 0.028); PTV + 1 cm, 3.5 ± 11.4 cm3 vs. 17.1 ± 29.8 cm3 (P = 0.028); PTV + 2 cm, 5.6 ± 14.5 cm3 vs. 26.8 ± 46.9 cm3 (P = 0.028); and PTV + 3 cm, 9.0 ± 17.4 cm3 vs. 36.5 ± 63.2 cm3 (P = 0.019), respectively. Some of the overlapping volumes of the rectum and bladder were significantly smaller in the prone position. On the other hand, when the target volume was extended by ≥2 cm, the overlapping volumes of the femurs were significantly larger in the prone position. V15 of the rectum and mean dose and V65 of the bladder were significantly lower in the prone position.ConclusionsThis study indicated that the volumes of the small intestine, rectum, and bladder in or near the PTV decreased when the patient was placed in the prone position on a belly board in postoperative radiotherapy for prostate cancer. The dose distribution seemed superior in the prone position to the supine position.

Algorithmic and modeling insights via volumetric comparison of polyhedral relaxations

This is mostly a survey on some mathematical results concerning volumes of polytopes of interest in non-convex optimization. Our motivation is in geometrically comparing relaxations in the context of mixed-integer linear and nonlinear optimization, with the goal of gaining algorithmic and modeling insights. We consider relaxations of: fixed-charge formulations, vertex packing polytopes, boolean-quadric polytopes, and relaxations of graphs of monomials on box domains. Besides surveying the area, we do give a few new results, and we provide many directions for further work.

Assessment of Accuracy in the Identification of Rock Formations from Aerial and Terrestrial Laser-Scanning Data

Rock formations are among the most spectacular landscape features both for experts and the public. However, information about these objects is often stored inaccurately in existing spatial databases, their corresponding elevations are missing, or the entire rock object is completely absent. Cartographic depiction is also reduced to a point of areal symbology of a largely generalized character. This paper discusses options in identifying and analyzing rock formations from two digital terrain models (DTMs), DMR 5G and DMR 5G+, and irregularly spaced points of airborne laser-scanning (ALS) data with different point densities. A semi-automatic method allowing rock formations to be identified from DTMs is introduced at the beginning of the paper. A method to evaluate elevation models (volume differences) is subsequently applied and a 3D model of a selected rock object is created from terrestrial laser-scanning data. Finally, positional and volumetric comparisons of that 3D object are performed in 2D, 2.5D, and 3D. The results of the pilot study confirmed that the digital terrain models studied are a reliable source in identifying and updating rock formations using the semi-automatic method introduced. The results show that DMR 5G model quality decreases with increasing fragmentation and relative rock formation height, while the proportion of gross errors increases. The complementary DMR 5G+ is better in terms of location and altitude.

EP-1483: Early anal cancer: traditional large field vs modified small field - a volumetric comparison

EP-1483: Early anal cancer: traditional large field vs modified small field - a volumetric comparison

Volumetric and Volume Fractional Comparison of the Nasal Structures of the Stork (Ciconia ciconia) and Seagull (Larus fuscus) Using Computed Tomography Images

Volumetric and Volume Fractional Comparison of the Nasal Structures of the Stork (Ciconia ciconia) and Seagull (Larus fuscus) Using Computed Tomography Images

Deep Learning Algorithm for Auto-Delineation of High-Risk Oropharyngeal Clinical Target Volumes With Built-In Dice Similarity Coefficient Parameter Optimization Function

Automating and standardizing the contouring of clinical target volumes (CTVs) can reduce interphysician variability, which is one of the largest sources of uncertainty in head and neck radiation therapy. In addition to using uniform margin expansions to auto-delineate high-risk CTVs, very little work has been performed to provide patient- and disease-specific high-risk CTVs. The aim of the present study was to develop a deep neural network for the auto-delineation of high-risk CTVs. Fifty-two oropharyngeal cancer patients were selected for the present study. All patients were treated at The University of Texas MD Anderson Cancer Center from January 2006 to August 2010 and had previously contoured gross tumor volumes and CTVs. We developed a deep learning algorithm using deep auto-encoders to identify physician contouring patterns at our institution. These models use distance map information from surrounding anatomic structures and the gross tumor volume as input parameters and conduct voxel-based classification to identify voxels that are part of the high-risk CTV. In addition, we developed a novel probability threshold selection function, based on the Dice similarity coefficient (DSC), to improve the generalization of the predicted volumes. The DSC-based function is implemented during an inner cross-validation loop, and probability thresholds are selected a priori during model parameter optimization. We performed a volumetric comparison between the predicted and manually contoured volumes to assess our model. The predicted volumes had a median DSC value of 0.81 (range 0.62-0.90), median mean surface distance of 2.8mm (range 1.6-5.5), and median 95th Hausdorff distance of 7.5mm (range 4.7-17.9) when comparing our predicted high-risk CTVs with the physician manual contours. These predicted high-risk CTVs provided close agreement to the ground-truth compared with current interobserver variability. The predicted contours could be implemented clinically, with only minor or no changes.