Lumbar and Thoracic Vertebrae Segmentation in CT Scans Using a 3D Multi-Object Localization and Segmentation CNN.

Radiation treatment of cancers like prostate or cervix cancer requires considering nearby bone structures like vertebrae. In this work, we present and validate a novel automated method for the 3D segmentation of individual lumbar and thoracic vertebra in computed tomography (CT) scans. It is based on a single, low-complexity convolutional neural network (CNN) architecture which works well even if little application-specific training data are available. It is based on volume patch-based processing, enabling the handling of arbitrary scan sizes. For each patch, it performs segmentation and an estimation of up to three vertebrae center locations in one step, which enables utilizing an advanced post-processing scheme to achieve high segmentation accuracy, as required for clinical use. Overall, 1763 vertebrae were used for the performance assessment. On 26 CT scans acquired for standard radiation treatment planning, a Dice coefficient of 0.921 ± 0.047 (mean ± standard deviation) and a signed distance error of 0.271 ± 0.748 mm was achieved. On the large-sized publicly available VerSe2020 data set with 129 CT scans depicting lumbar and thoracic vertebrae, the overall Dice coefficient was 0.940 ± 0.065 and the signed distance error was 0.109 ± 0.301 mm. A comparison to other methods that have been validated on VerSe data showed that our approach achieved a better overall segmentation performance.

Head and Neck Cancer Segmentation in FDG PET Images: Performance Comparison of Convolutional Neural Networks and Vision Transformers.

Convolutional neural networks (CNNs) have a proven track record in medical image segmentation. Recently, Vision Transformers were introduced and are gaining popularity for many computer vision applications, including object detection, classification, and segmentation. Machine learning algorithms such as CNNs or Transformers are subject to an inductive bias, which can have a significant impact on the performance of machine learning models. This is especially relevant for medical image segmentation applications where limited training data are available, and a model's inductive bias should help it to generalize well. In this work, we quantitatively assess the performance of two CNN-based networks (U-Net and U-Net-CBAM) and three popular Transformer-based segmentation network architectures (UNETR, TransBTS, and VT-UNet) in the context of HNC lesion segmentation in volumetric [F-18] fluorodeoxyglucose (FDG) PET scans. For performance assessment, 272 FDG PET-CT scans of a clinical trial (ACRIN 6685) were utilized, which includes a total of 650 lesions (primary: 272 and secondary: 378). The image data used are highly diverse and representative for clinical use. For performance analysis, several error metrics were utilized. The achieved Dice coefficient ranged from 0.833 to 0.809 with the best performance being achieved by CNN-based approaches. U-Net-CBAM, which utilizes spatial and channel attention, showed several advantages for smaller lesions compared to the standard U-Net. Furthermore, our results provide some insight regarding the image features relevant for this specific segmentation application. In addition, results highlight the need to utilize primary as well as secondary lesions to derive clinically relevant segmentation performance estimates avoiding biases.

Noise-Based Image Harmonization Significantly Increases Repeatability and Reproducibility of Radiomics Features in PET Images: A Phantom Study.

For multicenter clinical studies, characterizing the robustness of image-derived radiomics features is essential. Features calculated on PET images have been shown to be very sensitive to image noise. The purpose of this work was to investigate the efficacy of a relatively simple harmonization strategy on feature robustness and agreement. A purpose-built texture pattern phantom was scanned on 10 different PET scanners in 7 institutions with various different image acquisition and reconstruction protocols. An image harmonization technique based on equalizing a contrast-to-noise ratio was employed to generate a "harmonized" alongside a "standard" dataset for a reproducibility study. In addition, a repeatability study was performed with images from a single PET scanner of variable image noise, varying the binning time of the reconstruction. Feature agreement was measured using the intraclass correlation coefficient (ICC). In the repeatability study, 81/93 features had a lower ICC on the images with the highest image noise as compared to the images with the lowest image noise. Using the harmonized dataset significantly improved the feature agreement for five of the six investigated feature classes over the standard dataset. For three feature classes, high feature agreement corresponded with higher sensitivity to the different patterns, suggesting a way to select suitable features for predictive models.

Quantification of uptake in pelvis F-18 FLT PET-CT images using a 3D localization and segmentation CNN.

PURPOSE: The purpose of this work was to develop and validate a deep convolutional neural network (CNN) approach for the automated pelvis segmentation in computed tomography (CT) scans to enable the quantification of active pelvic bone marrow by means of Fluorothymidine F-18 (FLT) tracer uptake measurement in positron emission tomography (PET) scans. This quantification is a critical step in calculating bone marrow dose for radiopharmaceutical therapy clinical applications as well as external beam radiation doses. METHODS: An approach for the combined localization and segmentation of the pelvis in CT volumes of varying sizes, ranging from full-body to pelvis CT scans, was developed that utilizes a novel CNN architecture in combination with a random sampling strategy. The method was validated on 34 planning CT scans and 106 full-body FLT PET-CT scans using a cross-validation strategy. Specifically, two different training and CNN application options were studied, quantitatively assessed, and statistically compared. RESULTS: The proposed method was able to successfully locate and segment the pelvis in all test cases. On all data sets, an average Dice coefficient of 0.9396 ± $\pm$ 0.0182 or better was achieved. The relative tracer uptake measurement error ranged between 0.065% and 0.204%. The proposed approach is time-efficient and shows a reduction in runtime of up to 95% compared to a standard U-Net-based approach without a localization component. CONCLUSIONS: The proposed method enables the efficient calculation of FLT uptake in the pelvis. Thus, it represents a valuable tool to facilitate bone marrow preserving adaptive radiation therapy and radiopharmaceutical dose calculation. Furthermore, the method can be adapted to process other bone structures as well as organs.

Multisite Technical and Clinical Performance Evaluation of Quantitative Imaging Biomarkers from 3D FDG PET Segmentations of Head and Neck Cancer Images.

Quantitative imaging biomarkers (QIBs) provide medical image-derived intensity, texture, shape, and size features that may help characterize cancerous tumors and predict clinical outcomes. Successful clinical translation of QIBs depends on the robustness of their measurements. Biomarkers derived from positron emission tomography images are prone to measurement errors owing to differences in image processing factors such as the tumor segmentation method used to define volumes of interest over which to calculate QIBs. We illustrate a new Bayesian statistical approach to characterize the robustness of QIBs to different processing factors. Study data consist of 22 QIBs measured on 47 head and neck tumors in 10 positron emission tomography/computed tomography scans segmented manually and with semiautomated methods used by 7 institutional members of the NCI Quantitative Imaging Network. QIB performance is estimated and compared across institutions with respect to measurement errors and power to recover statistical associations with clinical outcomes. Analysis findings summarize the performance impact of different segmentation methods used by Quantitative Imaging Network members. Robustness of some advanced biomarkers was found to be similar to conventional markers, such as maximum standardized uptake value. Such similarities support current pursuits to better characterize disease and predict outcomes by developing QIBs that use more imaging information and are robust to different processing factors. Nevertheless, to ensure reproducibility of QIB measurements and measures of association with clinical outcomes, errors owing to segmentation methods need to be reduced.

A unified framework for simultaneous assessment of accuracy, between-, and within-reader variability of image segmentations.

Medical imaging is utilized in a wide range of clinical applications. To enable a detailed quantitative analysis, medical images must often be segmented to label (delineate) structures of interest; for example, a tumor. Frequently, manual segmentation is utilized in clinical practice (e.g., in radiation oncology) to define such structures of interest. However, it can be quite time consuming and subject to substantial between-, and within-reader variability. A more reproducible, less variable, and more time efficient segmentation approach is likely to improve medical treatment. This potential has spurred the development of segmentation algorithms which harness computational power. Segmentation algorithms' widespread use is limited due to difficulty in quantifying their performance relative to manual segmentation, which itself is subject to variation. This paper presents a statistical model which simultaneously estimates segmentation method accuracy, and between- and within-reader variability. The model is simultaneously fit for multiple segmentation methods within a unified Bayesian framework. The Bayesian model is compared to other methods used in literature via a simulation study, and application to head and neck cancer PET/CT data. The modeling framework is flexible and can be employed in numerous comparison applications. Several alternate applications are discussed in the paper.

The fractal geometry of bronchial trees differs by strain in mice.

Fractal biological structures are pervasive throughout the plant and animal kingdoms, with the mammalian lung being a quintessential example. The lung airway and vascular trees are generated during embryogenesis from a small set of building codes similar to Turing mechanisms that create robust trees ideally suited to their functions. Whereas the blood flow pattern generated by these fractal trees has been shown to be genetically determined, the geometry of the trees has not. We explored a newly established repository providing high-resolution bronchial trees from the four most commonly studied laboratory mice (B6C3F1, BALB/c, C57BL/6 and CD-1). The data fit a fractal model well for all animals with the fractal dimensions ranging from 1.54 to 1.67, indicating that the conducting airway of mice can be considered a self-similar and space-filling structure. We determined that the fractal dimensions of these airway trees differed by strain but not sex, reinforcing the concept that airway branching patterns are encoded within the DNA. The observations also highlight that future study design and interpretations may need to consider differences in airway geometry between mouse strains.NEW & NOTEWORTHY Similar to larger mammals such as humans, the geometries of the bronchial tree in mice are fractal structures that have repeating patterns from the trachea to the terminal branches. The airway geometries of the four most commonly studied mice are different and need to be considered when comparing results that employ different mouse strains. This variability in mouse airway geometries should be incorporated into computer models exploring toxicology and aerosol deposition in mouse models.

A 3D deep convolutional neural network approach for the automated measurement of cerebellum tracer uptake in FDG PET-CT scans.

PURPOSE: The purpose of this work was to assess the potential of deep convolutional neural networks in automated measurement of cerebellum tracer uptake in F-18 fluorodeoxyglucose (FDG) positron emission tomography (PET) scans. METHODS: Three different three-dimensional (3D) convolutional neural network architectures (U-Net, V-Net, and modified U-Net) were implemented and compared regarding their performance in 3D cerebellum segmentation in FDG PET scans. For network training and testing, 134 PET scans with corresponding manual volumetric segmentations were utilized. For segmentation performance assessment, a fivefold cross-validation was used, and the Dice coefficient as well as signed and unsigned distance errors were calculated. In addition, standardized uptake value (SUV) uptake measurement performance was assessed by means of a statistical comparison to an independent reference standard. Furthermore, a comparison to a previously reported active-shape-model-based approach was performed. RESULTS: Out of the three convolutional neural networks investigated, the modified U-Net showed significantly better segmentation performance. It achieved a Dice coefficient of 0.911 ± 0.026, a signed distance error of 0.220 ± 0.103 mm, and an unsigned distance error of 1.048 ± 0.340 mm. When compared to the independent reference standard, SUV uptake measurements produced with the modified U-Net showed no significant error in slope and intercept. The estimated reduction in total SUV measurement error was 95.1%. CONCLUSIONS: The presented work demonstrates the potential of deep convolutional neural networks in automated SUV measurement of reference regions. While it focuses on the cerebellum, utilized methods can be generalized to other reference regions like the liver or aortic arch. Future work will focus on combining lesion and reference region analysis into one approach.

lapdMouse: associating lung anatomy with local particle deposition in mice.

To facilitate computational toxicology, we developed an approach for generating high-resolution lung-anatomy and particle-deposition mouse models. Major processing steps of our method include mouse preparation, serial block-face cryomicrotome imaging, and highly automated image analysis for generating three-dimensional (3D) mesh-based models and volume-based models of lung anatomy (airways, lobes, sublobes, and near-acini structures) that are linked to local particle-deposition measurements. Analysis resulted in 34 mouse models covering 4 different mouse strains (B6C3F1: 8, BALB/C: 11, C57Bl/6: 8, and CD-1: 7) as well as both sexes (16 male and 18 female) and different particle sizes [2 µm (n = 15), 1 µm (n = 16), and 0.5 µm (n = 3)]. On average, resulting mouse airway models had 1,616.9 ± 298.1 segments, a centerline length of 597.6 ± 59.8 mm, and 1,968.9 ± 296.3 outlet regions. In addition to 3D geometric lung models, matching detailed relative particle-deposition measurements are provided. All data sets are available online in the lapdMouse archive for download. The presented approach enables linking relative particle deposition to anatomical structures like airways. This will in turn improve the understanding of site-specific airflows and how they affect drug, environmental, or biological aerosol deposition.NEW & NOTEWORTHY Computer simulations of particle deposition in mouse lungs play an important role in computational toxicology. Until now, a limiting factor was the lack of high-resolution mouse lung models and measured local particle-deposition information, which are required for developing accurate modeling approaches (e.g., computational fluid dynamics). With the developed imaging and analysis approach, we address this issue and provide all of the raw and processed data in a publicly accessible repository.

Machine learning with the TCGA-HNSC dataset: improving usability by addressing inconsistency, sparsity, and high-dimensionality.

BACKGROUND: In the era of precision oncology and publicly available datasets, the amount of information available for each patient case has dramatically increased. From clinical variables and PET-CT radiomics measures to DNA-variant and RNA expression profiles, such a wide variety of data presents a multitude of challenges. Large clinical datasets are subject to sparsely and/or inconsistently populated fields. Corresponding sequencing profiles can suffer from the problem of high-dimensionality, where making useful inferences can be difficult without correspondingly large numbers of instances. In this paper we report a novel deployment of machine learning techniques to handle data sparsity and high dimensionality, while evaluating potential biomarkers in the form of unsupervised transformations of RNA data. We apply preprocessing, MICE imputation, and sparse principal component analysis (SPCA) to improve the usability of more than 500 patient cases from the TCGA-HNSC dataset for enhancing future oncological decision support for Head and Neck Squamous Cell Carcinoma (HNSCC). RESULTS: Imputation was shown to improve prognostic ability of sparse clinical treatment variables. SPCA transformation of RNA expression variables reduced runtime for RNA-based models, though changes to classifier performance were not significant. Gene ontology enrichment analysis of gene sets associated with individual sparse principal components (SPCs) are also reported, showing that both high- and low-importance SPCs were associated with cell death pathways, though the high-importance gene sets were found to be associated with a wider variety of cancer-related biological processes. CONCLUSIONS: MICE imputation allowed us to impute missing values for clinically informative features, improving their overall importance for predicting two-year recurrence-free survival by incorporating variance from other clinical variables. Dimensionality reduction of RNA expression profiles via SPCA reduced both computation cost and model training/evaluation time without affecting classifier performance, allowing researchers to obtain experimental results much more quickly. SPCA simultaneously provided a convenient avenue for consideration of biological context via gene ontology enrichment analysis.

FDG PET based prediction of response in head and neck cancer treatment: Assessment of new quantitative imaging features.

INTRODUCTION: 18 F-fluorodeoxyglucose (FDG) positron emission tomography (PET) is now a standard diagnostic imaging test performed in patients with head and neck cancer for staging, re-staging, radiotherapy planning, and outcome assessment. Currently, quantitative analysis of FDG PET scans is limited to simple metrics like maximum standardized uptake value, metabolic tumor volume, or total lesion glycolysis, which have limited predictive value. The goal of this work was to assess the predictive potential of new (i.e., nonstandard) quantitative imaging features on head and neck cancer outcome. METHODS: This retrospective study analyzed fifty-eight pre- and post-treatment FDG PET scans of patients with head and neck squamous cell cancer to calculate five standard and seventeen new features at baseline and post-treatment. Cox survival regression was used to assess the predictive potential of each quantitative imaging feature on disease-free survival. RESULTS: Analysis showed that the post-treatment change of the average tracer uptake in the rim background region immediately adjacent to the tumor normalized by uptake in the liver represents a novel PET feature that is associated with disease-free survival (HR 1.95; 95% CI 1.27, 2.99) and has good discriminative performance (c index 0.791). CONCLUSION: The reported findings define a promising new direction for quantitative imaging biomarker research in head and neck squamous cell cancer and highlight the potential role of new radiomics features in oncology decision making as part of precision medicine.

FLT PET Radiomics for Response Prediction to Chemoradiation Therapy in Head and Neck Squamous Cell Cancer.

Radiomics is an image analysis approach for extracting large amounts of quantitative information from medical images using a variety of computational methods. Our goal was to evaluate the utility of radiomic feature analysis from 18F-fluorothymidine positron emission tomography (FLT PET) obtained at baseline in prediction of treatment response in patients with head and neck cancer. Thirty patients with advanced-stage oropharyngeal or laryngeal cancer, treated with definitive chemoradiation therapy, underwent FLT PET imaging before treatment. In total, 377 radiomic features of FLT uptake and feature variants were extracted from volumes of interest; these features variants were defined by either the primary tumor or the total lesion burden, which consisted of the primary tumor and all FLT-avid nodes. Feature variants included normalized measurements of uptake, which were calculated by dividing lesion uptake values by the mean uptake value in the bone marrow. Feature reduction was performed using clustering to remove redundancy, leaving 172 representative features. Effects of these features on progression-free survival were modeled with Cox regression and P-values corrected for multiple comparisons. In total, 9 features were considered significant. Our results suggest that smaller, more homogenous lesions at baseline were associated with better prognosis. In addition, features extracted from total lesion burden had a higher concordance index than primary tumor features for 8 of the 9 significant features. Furthermore, total lesion burden features showed lower interobserver variability.

A Bayesian framework for performance assessment and comparison of imaging biomarker quantification methods.

Quantitative biomarkers derived from medical images are being used increasingly to help diagnose disease, guide treatment, and predict clinical outcomes. Measurement of quantitative imaging biomarkers is subject to bias and variability from multiple sources, including the scanner technologies that produce images, the approaches for identifying regions of interest in images, and the algorithms that calculate biomarkers from regions. Moreover, these sources may differ within and between the quantification methods employed by institutions, thus making it difficult to develop and implement multi-institutional standards. We present a Bayesian framework for assessing bias and variability in imaging biomarkers derived from different quantification methods, comparing agreement to a reference standard, studying prognostic performance, and estimating sample size for future clinical studies. The statistical methods are illustrated with data obtained from a positron emission tomography challenge conducted by members of the NCI's Quantitative Imaging Network program, in which tumor volumes were measured manually and with seven different semi-automated segmentation algorithms. Estimates and comparisons of bias and variability in the resulting measurements are provided along with an R software package for the technical performance analysis and an online web application for sample size and power analysis.

Pulmonary lobe separation in expiration chest CT scans based on subject-specific priors derived from inspiration scans.

Segmentation of pulmonary lobes in inspiration and expiration chest CT scan pairs is an important prerequisite for lobe-based quantitative disease assessment. Conventional methods process each CT scan independently, resulting typically in lower segmentation performance at expiration compared to inspiration. To address this issue, we present an approach, which utilizes CT scans at both respiratory states. It consists of two main parts: a base method that processes a single CT scan and an extended method that utilizes the segmentation result obtained on the inspiration scan as a subject-specific prior for segmentation of the expiration scan. We evaluated the methods on a diverse set of 40 CT scan pairs. In addition, we compare the performance of our method to a registration-based approach. On inspiration scans, the base method achieved an average distance error of 0.59, 0.64, and 0.91 mm for the left oblique, right oblique, and right horizontal fissures, respectively, when compared with expert-based reference tracings. On expiration scans, the base method's errors were 1.54, 3.24, and 3.34 mm, respectively. In comparison, utilizing proposed subject-specific priors for segmentation of expiration scans allowed decreasing average distance errors to 0.82, 0.79, and 1.04 mm, which represents a significant improvement ([Formula: see text]) compared with all other methods investigated.

Chest wall strapping increases expiratory airflow and detectable airway segments in computer tomographic scans of normal and obstructed lungs.

Chest wall strapping (CWS) induces breathing at low lung volumes but also increases parenchymal elastic recoil. In this study, we tested the hypothesis that CWS dilates airways via airway-parenchymal interdependence. In 11 subjects (6 healthy and 5 with mild to moderate COPD), pulmonary function tests and lung volumes were obtained in control (baseline) and the CWS state. Control and CWS-CT scans were obtained at 50% of control (baseline) total lung-capacity (TLC). CT lung volumes were analyzed by CT volumetry. If control and CWS-CT volumetry did not differ by more than 25%, airway dimensions were analyzed via automated airway segmentation. CWS-TLC was reduced on average to 71% of control-TLC in normal subjects and 79% of control-TLC in subjects with COPD. CWS increased expiratory airflow at 50% of control-TLC by 41% (3.50 ± 1.6 vs. 4.93 ± 1.9 l/s, P = 0.04) in normals and 316% in COPD(0.25 ± 0.05 vs 0.79 ± 0.39 l/s, P = 0.04). In 10 subjects (5 normals and 5 COPD), control and CWS-CT scans at 50% control-TLC did not differ more than 25% on CT volumetry and were included in the airway structure analysis. CWS increased the mean number of detectable airways with a diameter of ≤2 mm by 32.5% (65 ± 10 vs. 86 ± 124, P = 0.01) in normal subjects and by 79% (59 ± 19 vs. 104 ± 16, P = 0.01) in subjects with COPD. There was no difference in the number of detectable airways with diameters 2-4 mm and >4 mm in normal or in COPD subjects. In conclusion, CWS enhances the detection of small airways via automated CT airway segmentation and increases expiratory airflow in normal subjects as well as in subjects with mild to moderate COPD. NEW & NOTEWORTHY In normal and COPD subjects, chest wall strapping(CWS) increased the number of detectable small airways using automated CT airway segmentation. The concept of dysanapsis expresses the physiological variation in the geometry of the tracheobronchial tree and lung parenchyma based on development. We propose a dynamic concept to dysanapsis in which CWS leads to breathing at lower lung volumes with a corresponding increase in the size of small airways, a potentially novel, nonpharmacological treatment for COPD.

Automated model-based quantitative analysis of phantoms with spherical inserts in FDG PET scans.

PURPOSE: Quality control plays an increasingly important role in quantitative PET imaging and is typically performed using phantoms. The purpose of this work was to develop and validate a fully automated analysis method for two common PET/CT quality assurance phantoms: the NEMA NU-2 IQ and SNMMI/CTN oncology phantom. The algorithm was designed to only utilize the PET scan to enable the analysis of phantoms with thin-walled inserts. METHODS: We introduce a model-based method for automated analysis of phantoms with spherical inserts. Models are first constructed for each type of phantom to be analyzed. A robust insert detection algorithm uses the model to locate all inserts inside the phantom. First, candidates for inserts are detected using a scale-space detection approach. Second, candidates are given an initial label using a score-based optimization algorithm. Third, a robust model fitting step aligns the phantom model to the initial labeling and fixes incorrect labels. Finally, the detected insert locations are refined and measurements are taken for each insert and several background regions. In addition, an approach for automated selection of NEMA and CTN phantom models is presented. The method was evaluated on a diverse set of 15 NEMA and 20 CTN phantom PET/CT scans. NEMA phantoms were filled with radioactive tracer solution at 9.7:1 activity ratio over background, and CTN phantoms were filled with 4:1 and 2:1 activity ratio over background. For quantitative evaluation, an independent reference standard was generated by two experts using PET/CT scans of the phantoms. In addition, the automated approach was compared against manual analysis, which represents the current clinical standard approach, of the PET phantom scans by four experts. RESULTS: The automated analysis method successfully detected and measured all inserts in all test phantom scans. It is a deterministic algorithm (zero variability), and the insert detection RMS error (i.e., bias) was 0.97, 1.12, and 1.48 mm for phantom activity ratios 9.7:1, 4:1, and 2:1, respectively. For all phantoms and at all contrast ratios, the average RMS error was found to be significantly lower for the proposed automated method compared to the manual analysis of the phantom scans. The uptake measurements produced by the automated method showed high correlation with the independent reference standard (R2 ≥ 0.9987). In addition, the average computing time for the automated method was 30.6 s and was found to be significantly lower (P ≪ 0.001) compared to manual analysis (mean: 247.8 s). CONCLUSIONS: The proposed automated approach was found to have less error when measured against the independent reference than the manual approach. It can be easily adapted to other phantoms with spherical inserts. In addition, it eliminates inter- and intraoperator variability in PET phantom analysis and is significantly more time efficient, and therefore, represents a promising approach to facilitate and simplify PET standardization and harmonization efforts.

Development of a radiobiological evaluation tool to assess the expected clinical impacts of contouring accuracy between manual and semi-automated segmentation algorithms.

RADEval is a tool developed to assess the expected clinical impact of contouring accuracy when comparing manual contouring and semi-automated segmentation. The RADEval tool, designed to process large scale datasets, imported a total of 2,760 segmentation datasets, along with a Simultaneous Truth and Performance Level Estimation (STAPLE) to act as ground truth tumor segmentations. Virtual dose-maps were created within RADEval and two different tumor control probability (TCP) values using a Logistic and a Poisson TCP models were calculated in RADEval using each STAPLE and each dose-map. RADEval also virtually generated a ring of normal tissue. To evaluate clinical impact, two different uncomplicated TCP (UTCP) values were calculated in RADEval by using two TCP-NTCP correlation parameters (δ = 0 and 1). NTCP values showed that semi-automatic segmentation resulted in lower NTCP with an average 1.5 - 1.6 % regardless of STAPLE design. This was true even though each normal tissue was created from each STAPLE (p <; 0.00001). TCP and UTCP presented no statistically significant differences (p ≥ 0.1884). The intra-operator standard deviations (SDs) for TCP, NTCP and UTCP were significantly lower for the semi-automatic segmentation method regardless of STAPLE design (p <; 0.0331). Both intra-and inter-operator SDs of TCP, NTCP and UTCP were significantly lower for semi-automatic segmentation for the STAPLE 1 design (p <;0.0331). RADEval was able to efficiently process 4,920 datasets of two STAPLE designs and successfully assess the expected clinical impact of contouring accuracy.

Multi-site quality and variability analysis of 3D FDG PET segmentations based on phantom and clinical image data.

PURPOSE: Radiomics utilizes a large number of image-derived features for quantifying tumor characteristics that can in turn be correlated with response and prognosis. Unfortunately, extraction and analysis of such image-based features is subject to measurement variability and bias. The challenge for radiomics is particularly acute in Positron Emission Tomography (PET) where limited resolution, a high noise component related to the limited stochastic nature of the raw data, and the wide variety of reconstruction options confound quantitative feature metrics. Extracted feature quality is also affected by tumor segmentation methods used to define regions over which to calculate features, making it challenging to produce consistent radiomics analysis results across multiple institutions that use different segmentation algorithms in their PET image analysis. Understanding each element contributing to these inconsistencies in quantitative image feature and metric generation is paramount for ultimate utilization of these methods in multi-institutional trials and clinical oncology decision making. METHODS: To assess segmentation quality and consistency at the multi-institutional level, we conducted a study of seven institutional members of the National Cancer Institute Quantitative Imaging Network. For the study, members were asked to segment a common set of phantom PET scans acquired over a range of imaging conditions as well as a second set of head and neck cancer (HNC) PET scans. Segmentations were generated at each institution using their preferred approach. In addition, participants were asked to repeat segmentations with a time interval between initial and repeat segmentation. This procedure resulted in overall 806 phantom insert and 641 lesion segmentations. Subsequently, the volume was computed from the segmentations and compared to the corresponding reference volume by means of statistical analysis. RESULTS: On the two test sets (phantom and HNC PET scans), the performance of the seven segmentation approaches was as follows. On the phantom test set, the mean relative volume errors ranged from 29.9 to 87.8% of the ground truth reference volumes, and the repeat difference for each institution ranged between -36.4 to 39.9%. On the HNC test set, the mean relative volume error ranged between -50.5 to 701.5%, and the repeat difference for each institution ranged between -37.7 to 31.5%. In addition, performance measures per phantom insert/lesion size categories are given in the paper. On phantom data, regression analysis resulted in coefficient of variation (CV) components of 42.5% for scanners, 26.8% for institutional approaches, 21.1% for repeated segmentations, 14.3% for relative contrasts, 5.3% for count statistics (acquisition times), and 0.0% for repeated scans. Analysis showed that the CV components for approaches and repeated segmentations were significantly larger on the HNC test set with increases by 112.7% and 102.4%, respectively. CONCLUSION: Analysis results underline the importance of PET scanner reconstruction harmonization and imaging protocol standardization for quantification of lesion volumes. In addition, to enable a distributed multi-site analysis of FDG PET images, harmonization of analysis approaches and operator training in combination with highly automated segmentation methods seems to be advisable. Future work will focus on quantifying the impact of segmentation variation on radiomics system performance.

An approach for reducing the error rate in automated lung segmentation.

Robust lung segmentation is challenging, especially when tens of thousands of lung CT scans need to be processed, as required by large multi-center studies. The goal of this work was to develop and assess a method for the fusion of segmentation results from two different methods to generate lung segmentations that have a lower failure rate than individual input segmentations. As basis for the fusion approach, lung segmentations generated with a region growing and model-based approach were utilized. The fusion result was generated by comparing input segmentations and selectively combining them using a trained classification system. The method was evaluated on a diverse set of 204 CT scans of normal and diseased lungs. The fusion approach resulted in a Dice coefficient of 0.9855±0.0106 and showed a statistically significant improvement compared to both input segmentation methods. In addition, the failure rate at different segmentation accuracy levels was assessed. For example, when requiring that lung segmentations must have a Dice coefficient of better than 0.97, the fusion approach had a failure rate of 6.13%. In contrast, the failure rate for region growing and model-based methods was 18.14% and 15.69%, respectively. Therefore, the proposed method improves the quality of the lung segmentations, which is important for subsequent quantitative analysis of lungs. Also, to enable a comparison with other methods, results on the LOLA11 challenge test set are reported.

Semiautomated segmentation of head and neck cancers in 18F-FDG PET scans: A just-enough-interaction approach.

PURPOSE: The purpose of this work was to develop, validate, and compare a highly computer-aided method for the segmentation of hot lesions in head and neck 18F-FDG PET scans. METHODS: A semiautomated segmentation method was developed, which transforms the segmentation problem into a graph-based optimization problem. For this purpose, a graph structure around a user-provided approximate lesion centerpoint is constructed and a suitable cost function is derived based on local image statistics. To handle frequently occurring situations that are ambiguous (e.g., lesions adjacent to each other versus lesion with inhomogeneous uptake), several segmentation modes are introduced that adapt the behavior of the base algorithm accordingly. In addition, the authors present approaches for the efficient interactive local and global refinement of initial segmentations that are based on the "just-enough-interaction" principle. For method validation, 60 PET/CT scans from 59 different subjects with 230 head and neck lesions were utilized. All patients had squamous cell carcinoma of the head and neck. A detailed comparison with the current clinically relevant standard manual segmentation approach was performed based on 2760 segmentations produced by three experts. RESULTS: Segmentation accuracy measured by the Dice coefficient of the proposed semiautomated and standard manual segmentation approach was 0.766 and 0.764, respectively. This difference was not statistically significant (p = 0.2145). However, the intra- and interoperator standard deviations were significantly lower for the semiautomated method. In addition, the proposed method was found to be significantly faster and resulted in significantly higher intra- and interoperator segmentation agreement when compared to the manual segmentation approach. CONCLUSIONS: Lack of consistency in tumor definition is a critical barrier for radiation treatment targeting as well as for response assessment in clinical trials and in clinical oncology decision-making. The properties of the authors approach make it well suited for applications in image-guided radiation oncology, response assessment, or treatment outcome prediction.