Convolutional neural network-based automatic liver delineation on contrast-enhanced and non-contrast-enhanced CT images for radiotherapy planning

AimThis study evaluated a convolutional neural network (CNN) for automatically delineating the liver on contrast-enhanced or non-contrast-enhanced CT, making comparisons with a commercial automated technique (MIM Maestro®).BackgroundIntensity-modulated radiation therapy requires careful labor-intensive planning involving delineation of the target and organs on CT or MR images to ensure delivery of the effective dose to the target while avoiding organs at risk.Materials and MethodsContrast-enhanced planning CT images from 101 pancreatic cancer cases and accompanying mask images showing manually-delineated liver contours were used to train the CNN to segment the liver. The trained CNN then performed liver segmentation on a further 20 contrast-enhanced and 15 non-contrastenhanced CT image sets, producing three-dimensional mask images of the liver.ResultsFor both contrast-enhanced and non-contrast-enhanced images, the mean Dice similarity coefficients between CNN segmentations and ground-truth manual segmentations were significantly higher than those between ground-truth and MIM Maestro software (p < 0.001). Although mean CT values of the liver were higher on contrast-enhanced than on non-contrast-enhanced CT, there were no significant differences in the Hausdorff distances of the CNN segmentations, indicating that the CNN could successfully segment the liver on both image types, despite being trained only on contrast-enhanced images.ConclusionsOur results suggest that a CNN can perform highly accurate automated delineation of the liver on CT images, irrespective of whether the CT images are contrast-enhanced or not.     

A patient-specific approach for quantitative and automatic analysis of computed tomography images in lung disease: Application to COVID-19 patients

PurposeQuantitative metrics in lung computed tomography (CT) images have been widely used, often without a clear connection with physiology. This work proposes a patient-independent model for the estimation of well-aerated volume of lungs in CT images (WAVE).MethodsA Gaussian fit, with mean (Mu.f) and width (Sigma.f) values, was applied to the lower CT histogram data points of the lung to provide the estimation of the well-aerated lung volume (WAVE.f). Independence from CT reconstruction parameters and respiratory cycle was analysed using healthy lung CT images and 4DCT acquisitions. The Gaussian metrics and first order radiomic features calculated for a third cohort of COVID-19 patients were compared with those relative to healthy lungs. Each lung was further segmented in 24 subregions and a new biomarker derived from Gaussian fit parameter Mu.f was proposed to represent the local density changes.ResultsWAVE.f resulted independent from the respiratory motion in 80% of the cases. Differences of 1%, 2% and up to 14% resulted comparing a moderate iterative strength and FBP algorithm, 1 and 3 mm of slice thickness and different reconstruction kernel. Healthy subjects were significantly different from COVID-19 patients for all the metrics calculated. Graphical representation of the local biomarker provides spatial and quantitative information in a single 2D picture.ConclusionsUnlike other metrics based on fixed histogram thresholds, this model is able to consider the inter- and intra-subject variability. In addition, it defines a local biomarker to quantify the severity of the disease, independently of the observer.     

Segmentation of parotid glands from registered CT and MR images

PurposeTo develop an automatic multimodal method for segmentation of parotid glands (PGs) from pre-registered computed tomography (CT) and magnetic resonance (MR) images and compare its results to the results of an existing state-of-the-art algorithm that segments PGs from CT images only.MethodsMagnetic resonance images of head and neck were registered to the accompanying CT images using two different state-of-the-art registration procedures. The reference domains of registered image pairs were divided on the complementary PG regions and backgrounds according to the manual delineation of PGs on CT images, provided by a physician. Patches of intensity values from both image modalities, centered around randomly sampled voxels from the reference domain, served as positive or negative samples in the training of the convolutional neural network (CNN) classifier. The trained CNN accepted a previously unseen (registered) image pair and classified its voxels according to the resemblance of its patches to the patches used for training. The final segmentation was refined using a graph-cut algorithm, followed by the dilate-erode operations.ResultsUsing the same image dataset, segmentation of PGs was performed using the proposed multimodal algorithm and an existing monomodal algorithm, which segments PGs from CT images only. The mean value of the achieved Dice overlapping coefficient for the proposed algorithm was 78.8%, while the corresponding mean value for the monomodal algorithm was 76.5%.ConclusionsAutomatic PG segmentation on the planning CT image can be augmented with the MR image modality, leading to an improved RT planning of head and neck cancer.     

Convolution Neural Networks and Support Vector Machines for Automatic Segmentation of Intracoronary Optical Coherence Tomography

Cardiovascular diseases are closely associated with deteriorating atherosclerotic plaques. Optical coherence tomography (OCT) is a recently developed intravascular imaging technique with high resolution approximately 10 microns and could provide accurate quantification of coronary plaque morphology. However, tissue segmentation of OCT images in clinic is still mainly performed manually by physicians which is time consuming and subjective. To overcome these limitations, two automatic segmentation methods for intracoronary OCT image based on support vector machine (SVM) and convolutional neural network (CNN) were performed to identify the plaque region and characterize plaque components. In vivo IVUS and OCT coronary plaque data from 5 patients were acquired at Emory University with patient’s consent obtained. Seventy-seven matched IVUS and OCT slices with good image quality and lipid cores were selected for this study. Manual OCT segmentation was performed by experts using virtual histology IVUS as guidance, and used as gold standard in the automatic segmentations. The overall classification accuracy based on CNN method achieved 95.8%, and the accuracy based on SVM was 71.9%. The CNN-based segmentation method can better characterize plaque compositions on OCT images and greatly reduce the time spent by doctors in segmenting and identifying plaques.     

Effects of lung tissue characterization in radiotherapy of breast cancer under deep inspiration breath hold when using Monte Carlo dosimetry

PurposeTo investigate the sensitivity of Monte Carlo (MC) calculated lung dose distributions to lung tissue characterization in external beam radiotherapy of breast cancer under Deep Inspiration Breath Hold (DIBH).MethodsEGSnrc based MC software was employed. Mean lung densities for one hundred patients were analysed. CT number frequency and clinical dose distributions were calculated for 15 patients with mean lung density below 0.14 g/cm³. Lung volume with a pre-defined CT numbers was also considered. Lung tissue was characterized by applying different CT calibrations in the low-density region and air-lung tissue thresholds. Dose impact was estimated by Dose Volume Histogram (DVH) parameters.ResultsMean lung densities below 0.14 g/cm³ were found in 10% of the patients. CT numbers below −960 HU dominated the CT frequency distributions with a high rate of CT numbers at −990 HU. Mass density conversion approach influenced the DVH shape. V_4Gy and V_8Gy varied by 7% and 5% for the selected patients and by 9% and 3.5% for the pre-defined lung volume. V_16Gy and V_20Gy, were within 2.5%. Regions above 20 Gy were affected. Variations in air- lung tissue differentiation resulted in DVH parameters within 1%. Threshold at −990 HU was confirmed by the CT number frequency distributions.ConclusionsLung dose distributions were more sensitive to variations in the CT calibration curve below lung (inhale) density than to air-lung tissue differentiation. Low dose regions were mostly affected. The dosimetry effects were found to be potentially important to 10% of the patients treated under DIBH.     

Enhancement of Hounsfield unit distribution in cone-beam CT images for adaptive radiation therapy: Evaluation of a hybrid correction approach

PurposeThis study aims to evaluate the accuracy of a hybrid approach combining the histogram matching (HM) and the multilevel threshold (MLT) to correct the Hounsfield Unit (HU) distribution in cone-beam CT (CBCT) images.Methods and MaterialsCBCT images acquired for ten prostate cancer patients were processed by matching their histograms to those of deformed planning CT (pCT) images obtained after applying a deformable registration (DR) process. Then, HU values corresponding to five tissue types in the pCT were assigned to the obtained CBCT images (CBCT_HM-MLT). Finally, the CBCT_HM-MLT images were compared to the deformed pCT visually and using different statistical metrics.ResultsThe visual assessment and the profiles comparison showed that the high discrepancies in the CBCT images were significantly reduced when using the proposed approach. Furthermore, the correlation values indicated that the CBCT_HM-MLT were in good agreement with the deformed pCT with correlation values ranging from 0.9893 to 0.9962. In addition, the root mean squared error (RMSE) over the entire volume was reduced from 64.15 ± 9.50 to 51.20 ± 6.76 HU. Similarly, the mean absolute error in specific tissue classes was significantly reduced especially in the soft tissue-air interfaces. These results confirmed that applying MLT after HM worked better than using only HM for which the correlation values were ranging from 0.9878 to 0.9955 and the RMSE was 55.95 ± 10.43 HU.ConclusionEvaluation of the proposed approach showed that the HM + MLT correction can improve the HU distribution in the CBCT images and generate corrected images in good agreement with the pCT.     

Investigating the impact of the CT Hounsfield unit range on radiomic feature stability using dual energy CT data

PurposeRadiomic texture calculation requires discretizing image intensities within the region-of-interest. FBN (fixed-bin-number), FBS (fixed-bin-size) and FBN and FBS with intensity equalization (FBNequal, FBSequal) are four discretization approaches. A crucial choice is the voxel intensity (Hounsfield units, or HU) binning range. We assessed the effect of this choice on radiomic features.MethodsThe dataset comprised 95 patients with head-and-neck squamous-cell-carcinoma. Dual energy CT data was reconstructed at 21 electron energies (40, 45,… 140 keV). Each of 94 texture features were calculated with 64 extraction parameters. All features were calculated five times: original choice, left shift (-10/-20 HU), right shift (+10/+20 HU). For each feature, Spearman correlation between nominal and four variants were calculated to determine feature stability. This was done for six texture feature types (GLCM, GLRLM, GLSZM, GLDZM, NGTDM, and NGLDM) separately. This analysis was repeated for the four binning algorithms. Effect of feature instability on predictive ability was studied for lymphadenopathy as endpoint.ResultsFBN and FBNequal algorithms showed good stability (correlation values consistently $>$0.9). For FBS and FBSequal algorithms, while median values exceeded 0.9, the 95% lower bound decreased as a function of energy, with poor performance over the entire spectrum. FBNequal was the most stable algorithm, and FBS the least.ConclusionsWe believe this is the first multi-energy systematic study of the impact of CT HU range used during intensity discretization for radiomic feature extraction. Future analyses should account for this source of uncertainty when evaluating the robustness of their radiomic signature.     

Impact de la méthode de segmentation dans la caractérisation des adénocarcinomes pulmonaires en TEP-TDM au 18FDG

ObjectiveThe treatment of lung adenocarcinomas is conditioned by the presence of certain genetic abnormalities. Certain quantitative parameters obtained from FDG PET-CT, at the voxel scale, provide tumour shape and texture characteristics and might predict their mutational status. Our objective was to determine the impact of the segmentation method in the characterization of lung adenocarcinomas in FDG PET-CT.MethodsForty-nine patients with pulmonary adenocarcinomas were retrospectively included, with their initial FDG PET-CT image. The studied tumours were big, heterogeneous and difficult to segment automatically. The automatic FLAB algorithm was used with and without manual adjustment. The parameters were extracted and compared to the ALK, PDL1, and KRAS status, in order to compare the performances of the two segmentation methods. Their performance was determined by the ROC curve method.ResultsSeveral parameters were significant to predict genetic status (AUC > 0.65). The best performing parameters were different according to the genes studied and according to the resampling methods used. The results were less dependent on resampling in automatic segmentation without manual adjustment. The best performing parameters were volume dependent parameters for segmentation with manual adjustment, and texture parameters for automatic segmentation without adjustment.ConclusionThe study of texture parameters is more efficient in automatic segmentation that is not manually adjusted, and it is advantageous to use a manual adjustment when studying volume-dependent parameters in the case of very heterogeneous tumors.     

Assessment of the variation in CT scanner performance (image quality and Hounsfield units) with scan parameters,for image optimisation in radiotherapy treatment planning

PurposeTo define a method and investigate how the adjustment of scan parameters affected the image quality and Hounsfield units (HUs) on a CT scanner used for radiotherapy treatment planning. A lack of similar investigations in the literature may be a contributing factor in the apparent reluctance to optimise radiotherapy CT protocols.MethodA Catphan phantom was used to assess how image quality on a Toshiba Aquilion LB scanner changed with scan parameters. Acquisition and reconstruction field-of-view (FOV), collimation, image slice thickness, effective mAs per rotation and reconstruction algorithm were varied. Changes were assessed for HUs of different materials, high contrast spatial resolution (HCSR), contrast-noise ratio (CNR), HU uniformity, scan direction low contrast and CT dose-index.ResultsCNR and HCSR varied most with reconstruction algorithm, reconstruction FOV and effective mAs. Collimation, but not image slice width, had a significant effect on CT dose-index with narrower collimation giving higher doses. Dose increased with effective mAs. Highest HU differences were seen when changing reconstruction algorithm: 56 HU for densities close to water and 117 HU for bone-like materials. Acquisition FOV affected the HUs but reconstruction FOV and effective mAs did not.ConclusionsAll the scan parameters investigated affected the image quality metrics. Reconstruction algorithm, reconstruction FOV, collimation and effective mAs were most important. Reconstruction algorithm and acquisition FOV had significant effect on HU. The methodology is applicable to radiotherapy CT scanners when investigating image quality optimisation, prior to assessing the impact of scan protocol changes on clinical CT images and treatment plans.     

Segmentation automatique du volume tumoral des cancers pulmonaires non à petites cellules sur la 18F-FDG TEP-TDM en vue d’un traitement par radiothérapie

PurposeRadiotherapy belongs to the treatment of certain stages of non-small cell lung cancer. It requires a preliminary delimitation of the tumour by the radiotherapist with a contouring of the tumour on the CT but this method does not take into account the biological characteristics of the target and remains dependent on the operator. To optimize the irradiation volume, several teams created methods of automatic segmentation on the PET-CT, still under evaluation.Material and methodsWe undertook a retrospective study on 17 patients to evaluate the difference between obtained volumes by CT delimitation with those deduced by three algorithms of automatic segmentation based on PET-CT (Black et al., Nestle et al. and Tylski et al.). A confrontation with histological tumoral volume has been carried on in case of surgery.ResultsOn average, the three methods under evaluated tumoral volume compared to the CT of 8% for Nestle et al., 22% for Black et al. and 30% for Tylski et al., but these results hide disparities, which tended to decrease with a grouping of tumours by size. The most important differences were due to heterogeneous uptake of FDG (necrosis, spicules, atelectasis).ConclusionThe method of Black et al. was the most discrepant one. For tumours less than 45 cm³, Nestle's et al. algorithm tends to overestimate the CT volume and thus makes it possible to integrate safety margins into final volume (microscopic extension). The method of Tylski et al. presents an interesting approach (correction of partial volume effect) but still requires developments because it under evaluates too much the target volume.     

Evaluation of a cycle-generative adversarial network-based cone-beam CT to synthetic CT conversion algorithm for adaptive radiation therapy

PurposeImage-guided radiation therapy could benefit from implementing adaptive radiation therapy (ART) techniques. A cycle-generative adversarial network (cycle-GAN)-based cone-beam computed tomography (CBCT)-to-synthetic CT (sCT) conversion algorithm was evaluated regarding image quality, image segmentation and dosimetric accuracy for head and neck (H&N), thoracic and pelvic body regions.MethodsUsing a cycle-GAN, three body site-specific models were priorly trained with independent paired CT and CBCT datasets of a kV imaging system (XVI, Elekta). sCT were generated based on first-fraction CBCT for 15 patients of each body region. Mean errors (ME) and mean absolute errors (MAE) were analyzed for the sCT. On the sCT, manually delineated structures were compared to deformed structures from the planning CT (pCT) and evaluated with standard segmentation metrics. Treatment plans were recalculated on sCT. A comparison of clinically relevant dose-volume parameters (D₉₈, D₅₀ and D₂ of the target volume) and 3D-gamma (3%/3mm) analysis were performed.ResultsThe mean ME and MAE were 1.4, 29.6, 5.4 Hounsfield units (HU) and 77.2, 94.2, 41.8 HU for H&N, thoracic and pelvic region, respectively. Dice similarity coefficients varied between 66.7 ± 8.3% (seminal vesicles) and 94.9 ± 2.0% (lungs). Maximum mean surface distances were 6.3 mm (heart), followed by 3.5 mm (brainstem). The mean dosimetric differences of the target volumes did not exceed 1.7%. Mean 3D gamma pass rates greater than 97.8% were achieved in all cases.ConclusionsThe presented method generates sCT images with a quality close to pCT and yielded clinically acceptable dosimetric deviations. Thus, an important prerequisite towards clinical implementation of CBCT-based ART is fulfilled.     

Hierarchical clustering applied to automatic atlas based segmentation of 25 cardiac sub-structures

PurposeSegmentation of cardiac sub-structures for dosimetric analyses is usually performed manually in time-consuming procedure. Automatic segmentation may facilitate large-scale retrospective analysis and adaptive radiotherapy. Various approaches, among them Hierarchical Clustering, were applied to improve performance of atlas-based segmentation (ABS).MethodsTraining dataset of ABS consisted of 36 manually contoured CT-scans. Twenty-five cardiac sub-structures were contoured as regions of interest (ROIs). Five auto-segmentation methods were compared: simultaneous automatic contouring of all 25 ROIs (Method-1); automatic contouring of all 25 ROIs using lungs as anatomical barriers (Method-2); automatic contouring of a single ROI for each contouring cycle (Method-3); hierarchical cluster-based automatic contouring (Method-4); simultaneous truth and performance level estimation (STAPLE). Results were evaluated on 10 patients. Dice similarity coefficient (DSC), average Hausdorff distance (AHD), volume comparison and physician score were used as validation metrics.ResultsAtlas performance improved increasing number of atlases. Among the five ABS methods, Hierarchical Clustering workflow showed a significant improvement maintaining a clinically acceptable time for contouring. Physician scoring was acceptable for 70% of the ROI automatically contoured. Inter-observer evaluation showed that contours obtained by Hierarchical Clustering method are statistically comparable with them obtained by a second, independent, expert contourer considering DSC. Considering AHD, distance from the gold standard is lower for ROIs segmented by ABS.ConclusionsHierarchical clustering resulted in best ABS results for the primarily investigated platforms and compared favorably to a second benchmark system. Auto-contouring of smaller structures, being in range of variation between manual contourers, may be ideal for large-scale retrospective dosimetric analysis.     

Comparison of different deep learning architectures for synthetic CT generation from MR images

PurposeAmong the different available methods for synthetic CT generation from MR images for the task of MR-guided radiation planning, the deep learning algorithms have and do outperform their conventional counterparts. In this study, we investigated the performance of some most popular deep learning architectures including eCNN, U-Net, GAN, V-Net, and Res-Net for the task of sCT generation. As a baseline, an atlas-based method is implemented to which the results of the deep learning-based model are compared.MethodsA dataset consisting of 20 co-registered MR-CT pairs of the male pelvis is applied to assess the different sCT production methods' performance. The mean error (ME), mean absolute error (MAE), Pearson correlation coefficient (PCC), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR) metrics were computed between the estimated sCT and the ground truth (reference) CT images.ResultsThe visual inspection revealed that the sCTs produced by eCNN, V-Net, and ResNet, unlike the other methods, were less noisy and greatly resemble the ground truth CT image. In the whole pelvis region, the eCNN yielded the lowest MAE (26.03 ± 8.85 HU) and ME (0.82 ± 7.06 HU), and the highest PCC metrics were yielded by the eCNN (0.93 ± 0.05) and ResNet (0.91 ± 0.02) methods. The ResNet model had the highest PSNR of 29.38 ± 1.75 among all models. In terms of the Dice similarity coefficient, the eCNN method revealed superior performance in major tissue identification (air, bone, and soft tissue).ConclusionsAll in all, the eCNN and ResNet deep learning methods revealed acceptable performance with clinically tolerable quantification errors.     

Multi-Atlas Based Adaptive Active Contour Model with Application to Organs at Risk Segmentation in Brain MR Images

BackgroundAccurate delineation of organs at risk (OARs) is critical in radiotherapy. Manual delineation is tedious and suffers from both interobserver and intraobserver variability. Automatic segmentation of brain MR images has a wide range of applications in brain tumor radiotherapy. In this paper, we propose a multi-atlas based adaptive active contour model for OAR automatic segmentation in brain MR images.MethodsThe proposed method consists of two parts: multi-atlas based OAR contour initiation and an adaptive edge and local region based active contour evolution. In the adaptive active contour model, we define an energy functional with an adaptive edge intensity fitting force which is responsible for evaluating contour inwards or outwards, and a local region intensity fitting force which guides the evolution of the contour.ResultsExperimental results show that the proposed method achieved more accurate segmentation results in brainstem, eyes and lens automatic segmentation with the Dice Similar Coefficient (DSC) value of 87.19%, 91.96%, 77.11% respectively. Besides, the dosimetric parameters also demonstrate the high consistency of the manual OAR delineations and the auto segmentation results of the proposed method in brain tumor radiotherapy.ConclusionsThe geometric and dosimetric evaluations show the desirable performance of the proposed method on the application of OARs segmentations in brain tumor radiotherapy.     

A Review on Convolutional Neural Networks for Brain Tumor Segmentation: Methods,Datasets, Libraries,and Future Directions

ObjectivesAccurate and reliable segmentation of brain tumors from MRI images helps in planning an enhanced treatment and increases the life expectancy of patients. However, the manual segmentation of brain tumors is subjective and more prone to errors. Nonetheless, the recent advances in convolutional neural network (CNN)-based methods have exhibited outstanding potential in robust segmentation of brain tumors. This article comprehensively investigates recent advances in CNN-based methods for automatic segmentation of brain tumors from MRI images. It examines popular deep learning (DL) libraries/tools for an expeditious and effortless implementation of CNN models. Furthermore, a critical assessment of current DL architectures is delineated along with the scope of improvement.MethodsIn this work, more than 50 scientific papers from 2014-2020 are selected using Google Scholar and PubMed. Also, the leading journals related to our work along with proceedings from major conferences such as MICCAI, MIUA and ECCV are retrieved. This research investigated various annual challenges too related to this work including Multimodal Brain Tumor Segmentation Challenge (MICCAI BRATS) and Ischemic Stroke Lesion Segmentation Challenge (ISLES).ResultAfter a systematic literature search pertinent to the theme, we found that principally there exist three variations of CNN architecture for brain tumor segmentation: single-path and multi-path, fully convolutional, and cascaded CNNs. The respective performances of most automated methods based on CNN are appraised on the BraTS dataset, provided as a part of the MICCAI Multimodal Brain Tumor Segmentation challenge held annually since 2012.ConclusionNotwithstanding the remarkable potential of CNN-based methods, reliable and robust segmentation of brain tumors continues to be an intractable challenge. This is due to the intricate anatomy of the brain, variability in its appearance, and imperfection in image acquisition. Moreover, owing to the small size of MRI datasets, CNN-based methods cannot operate with their full capacity, as demonstrated with large scale datasets, such as ImageNet.     

A segmentation tool for pulmonary nodules in lung cancer screening: Testing and clinical usage

PurposeWith the future goal of defining a large dataset based on low-dose CT with labelled pulmonary lesions for lung cancer screening (LCS) research, the aim of this work is to propose and evaluate into a clinical context a tool for semi-automatic segmentation able to facilitate the process of labels collection from a LCS study (COSMOS, Continuous Observation of SMOking Subjects).MethodsConsidering a preliminary set of manual annotations, a segmentation model based on a 2D-Unet was trained from scratch. Contour quality of the final 2D-Unet was assessed on an internal test set of manual annotations and on a subset of the public available LIDC dataset used as external test set. The tool for semi-automatic segmentation was then designed integrating the tested model into a Graphical User Interface. According to the opinion of two clinical users, the percentage of lesions properly contoured through the tool was quantified (Acceptance Rate, AR). The variability between segmentations derived by the two readers was estimated as mean percentage of difference (MPD) between the two sets of volumes and comparing the likelihood of malignancy derived from Volume Doubling Time (VDT).ResultsPerformance in test sets were found similar (DICE ~ 0.75(0.15)). Accordingly, a good mean AR (80.1%) resulted from the two readers. Variability in terms of MPD was equal to 23.6% while 2.7% was the VDTs percentage of disagreement.ConclusionsA semi-automatic segmentation tool was developed and its applicability evaluated into a clinical context demonstrating the efficacy of the tool in facilitating the collection of labelled data.     

Negative effects of acorns on the wood mouse <Emphasis Type="Italic">Apodemus speciosus</Emphasis>

 Antinutritional effects of acorns and tannic acid on the Japanese wood mouse Apodemus speciosus were examined in the laboratory. The first feeding experiment was conducted for 15 days using three types of diet: control diet (laboratory chow for mice), acorns of Quercus serrata (QS), and acorns of Q. mongolica var. grosseserrata (QM), which differ in tannin content (control, tannin free; QS, 2.7% tannic acid equivalent; QM, 8.5%). Six and one of eight mice died in the QM and QS groups, respectively, whereas all mice survived in good health in the control group. Body weight in the QM and QS groups decreased as much as 23.6% and 16.8% in the first 5 days, respectively, whereas that in the control group did not change significantly. Dry matter intake in the QM group was 50.0% and 38.7% less than that in the control and QS, respectively. Apparent dry matter digestibility was not different among the diets, but apparent nitrogen digestibility did differ between the two acorn groups (QM, −17.5%; QS, 12.0%). The logistic regression analyses revealed that the survival of mice was synergistically influenced by both dry matter intake and apparent nitrogen digestibility. In the second experiment, wood mice fed the tannin-free formula diet, which is nutritionally matched to QS and QM acorns except for the tannin, did not suffer antinutritional effects, whereas mice fed the tannin-supplemented formula diets suffered body weight loss and negative nitrogen digestibility. These results indicate that the tannins in acorns could cause serious damage to the wood mouse, which may rely on acorns as a usual diet. Plausible hypotheses explaining how the wood mice could overcome the deleterious effects of the acorns are discussed. Received: May 9, 2002 / Accepted: December 6, 2002     

Fully Automated Enhanced Tumor Compartmentalization: Man vs. Machine Reloaded

ObjectiveComparison of a fully-automated segmentation method that uses compartmental volume information to a semi-automatic user-guided and FDA-approved segmentation technique.MethodsNineteen patients with a recently diagnosed and histologically confirmed glioblastoma (GBM) were included and MR images were acquired with a 1.5 T MR scanner. Manual segmentation for volumetric analyses was performed using the open source software 3D Slicer version 4.2.2.3 (www.slicer.org). Semi-automatic segmentation was done by four independent neurosurgeons and neuroradiologists using the computer-assisted segmentation tool SmartBrush® (referred to as SB), a semi-automatic user-guided and FDA-approved tumor-outlining program that uses contour expansion. Fully automatic segmentations were performed with the Brain Tumor Image Analysis (BraTumIA, referred to as BT) software. We compared manual (ground truth, referred to as GT), computer-assisted (SB) and fully-automated (BT) segmentations with regard to: (1) products of two maximum diameters for 2D measurements, (2) the Dice coefficient, (3) the positive predictive value, (4) the sensitivity and (5) the volume error.ResultsSegmentations by the four expert raters resulted in a mean Dice coefficient between 0.72 and 0.77 using SB. BT achieved a mean Dice coefficient of 0.68. Significant differences were found for intermodal (BT vs. SB) and for intramodal (four SB expert raters) performances. The BT and SB segmentations of the contrast-enhancing volumes achieved a high correlation with the GT. Pearson correlation was 0.8 for BT; however, there were a few discrepancies between raters (BT and SB 1 only). Additional non-enhancing tumor tissue extending the SB volumes was found with BT in 16/19 cases. The clinically motivated sum of products of diameters measure (SPD) revealed neither significant intermodal nor intramodal variations. The analysis time for the four expert raters was faster (1 minute and 47 seconds to 3 minutes and 39 seconds) than with BT (5 minutes).ConclusionBT and SB provide comparable segmentation results in a clinical setting. SB provided similar SPD measures to BT and GT, but differed in the volume analysis in one of the four clinical raters. A major strength of BT may its independence from human interactions, it can thus be employed to handle large datasets and to associate tumor volumes with clinical and/or molecular datasets ("-omics") as well as for clinical analyses of brain tumor compartment volumes as baseline outcome parameters. Due to its multi-compartment segmentation it may provide information about GBM subcompartment compositions that may be subjected to clinical studies to investigate the delineation of the target volumes for adjuvant therapies in the future.