Dataset Information

Adaptive threshold segmentation of pituitary adenomas from FDG PET images for radiosurgery.

ABSTRACT: In this study we have attempted to optimize a PET based adaptive threshold seg- mentation method for delineating small tumors, particularly in a background of high tracer activity. The metabolic nature of pituitary adenomas and the constraints of MRI imaging in the postoperative setting to delineate these tumors during radio- surgical procedures motivated us to develop this method. Phantom experiments were done to establish a relationship between the threshold required for segmenting the PET images and the target size and the activity concentration within the target in relation to its background. The threshold was developed from multiple linear regression of the experimental data optimized for tumor sizes less than 4 cm3. We validated our method against the phantom target volumes with measured target to background ratios ranging from 1.6 to 14.58. The method was tested on ten retro- spective patients with residual growth hormone-secreting pituitary adenomas that underwent radiosurgery and compared against the volumes delineated by manual method. The predicted volumes against the true volume of the phantom inserts gave a correlation coefficient of 99% (p < 0.01). In the ten retrospective patients, the automatically segmented tumor volumes against volumes manually delineated by the clinicians had a correlation of 94% (p < 0.01). This adaptive threshold segmentation showed promising results in delineating tumor volumes in pituitary adenomas planned for stereotactic radiosurgery, particularly in the postoperative setting where MR and CT images may be associated with artifacts, provided opti- mization experiment is carried out.

SUBMITTER: Thomas T HM

PROVIDER: S-EPMC5711116 | biostudies-other | 2014 Nov

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Adaptive threshold segmentation of pituitary adenomas from FDG PET images for radiosurgery.

Thomas T Hannah M HM Devadhas Devakumar D Heck Danie K DK Chacko Ari G AG Rebekah Grace G Oommen Regi R Samuel E James J EJ

Journal of applied clinical medical physics 20141108 6

In this study we have attempted to optimize a PET based adaptive threshold seg- mentation method for delineating small tumors, particularly in a background of high tracer activity. The metabolic nature of pituitary adenomas and the constraints of MRI imaging in the postoperative setting to delineate these tumors during radio- surgical procedures motivated us to develop this method. Phantom experiments were done to establish a relationship between the threshold required for segmenting the PET ima ...[more]

PMID: 25493519

Similar Datasets

Project description:We introduce multiple-organ objective segmentation (MOOSE) software that generates subject-specific, multiorgan segmentation using data-centric artificial intelligence principles to facilitate high-throughput systemic investigations of the human body via whole-body PET imaging. Methods: Image data from 2 PET/CT systems were used in training MOOSE. For noncerebral structures, 50 whole-body CT images were used, 30 of which were acquired from healthy controls (14 men and 16 women), and 20 datasets were acquired from oncology patients (14 men and 6 women). Noncerebral tissues consisted of 13 abdominal organs, 20 bone segments, subcutaneous fat, visceral fat, psoas muscle, and skeletal muscle. An expert panel manually segmented all noncerebral structures except for subcutaneous fat, visceral fat, and skeletal muscle, which were semiautomatically segmented using thresholding. A majority-voting algorithm was used to generate a reference-standard segmentation. From the 50 CT datasets, 40 were used for training and 10 for testing. For cerebral structures, 34 18F-FDG PET/MRI brain image volumes were used from 10 healthy controls (5 men and 5 women imaged twice) and 14 nonlesional epilepsy patients (7 men and 7 women). Only 18F-FDG PET images were considered for training: 24 and 10 of 34 volumes were used for training and testing, respectively. The Dice score coefficient (DSC) was used as the primary metric, and the average symmetric surface distance as a secondary metric, to evaluate the automated segmentation performance. Results: An excellent overlap between the reference labels and MOOSE-derived organ segmentations was observed: 92% of noncerebral tissues showed DSCs of more than 0.90, whereas a few organs exhibited lower DSCs (e.g., adrenal glands [0.72], pancreas [0.85], and bladder [0.86]). The median DSCs of brain subregions derived from PET images were lower. Only 29% of the brain segments had a median DSC of more than 0.90, whereas segmentation of 60% of regions yielded a median DSC of 0.80-0.89. The results of the average symmetric surface distance analysis demonstrated that the average distance between the reference standard and the automatically segmented tissue surfaces (organs, bones, and brain regions) lies within the size of image voxels (2 mm). Conclusion: The proposed segmentation pipeline allows automatic segmentation of 120 unique tissues from whole-body 18F-FDG PET/CT images with high accuracy.

Project description:PurposeFor pediatric lymphoma, quantitative FDG PET/CT imaging features such as metabolic tumor volume (MTV) are important for prognosis and risk stratification strategies. However, feature extraction is difficult and time-consuming in cases of high disease burden. The purpose of this study was to fully automate the measurement of PET imaging features in PET/CT images of pediatric lymphoma.Methods18F-FDG PET/CT baseline images of 100 pediatric Hodgkin lymphoma patients were retrospectively analyzed. Two nuclear medicine physicians identified and segmented FDG avid disease using PET thresholding methods. Both PET and CT images were used as inputs to a three-dimensional patch-based, multi-resolution pathway convolutional neural network architecture, DeepMedic. The model was trained to replicate physician segmentations using an ensemble of three networks trained with 5-fold cross-validation. The maximum SUV (SUVmax), MTV, total lesion glycolysis (TLG), surface-area-to-volume ratio (SA/MTV), and a measure of disease spread (Dmaxpatient) were extracted from the model output. Pearson's correlation coefficient and relative percent differences were calculated between automated and physician-extracted features.ResultsMedian Dice similarity coefficient of patient contours between automated and physician contours was 0.86 (IQR 0.78-0.91). Automated SUVmax values matched exactly the physician determined values in 81/100 cases, with Pearson's correlation coefficient (R) of 0.95. Automated MTV was strongly correlated with physician MTV (R = 0.88), though it was slightly underestimated with a median (IQR) relative difference of - 4.3% (- 10.0-5.7%). Agreement of TLG was excellent (R = 0.94), with median (IQR) relative difference of - 0.4% (- 5.2-7.0%). Median relative percent differences were 6.8% (R = 0.91; IQR 1.6-4.3%) for SA/MTV, and 4.5% (R = 0.51; IQR - 7.5-40.9%) for Dmaxpatient, which was the most difficult feature to quantify automatically.ConclusionsAn automated method using an ensemble of multi-resolution pathway 3D CNNs was able to quantify PET imaging features of lymphoma on baseline FDG PET/CT images with excellent agreement to reference physician PET segmentation. Automated methods with faster throughput for PET quantitation, such as MTV and TLG, show promise in more accessible clinical and research applications.

Project description:Background and aimAbout a third of Pituitary Neuroendocrine Tumors (PitNETs) may show aggressive behavior. Many efforts have been performed for identifying possible predictive factors to early determine the future behavior of PitNETs. Programmed cell death ligand 1 (PD-L1) expression was associated with a more aggressive biology in different solid tumors, but its role in PitNET is not well-established yet. Our study aims to analyze PD-L1 expression in a surgical cohort of PitNETs to determine its association with radiological invasion and pathology findings, as well as with long-term recurrence rates.MethodsWe performed a retrospective analysis in a series of 86 PitNETs. Clinical presentation and radiological features of the preoperative period were collected, as well as pathological data and follow-up data. The rate of PD-L1 expression was immunohistochemically evaluated and expressed as a tumor proportion score (TPS). We assessed its relationship with cavernous sinus invasion and Trouillas' classification as primary outcomes. Secondary outcomes included the TPS' relationship with histopathological markers of proliferation, hormonal expression, tumor size and long-term recurrence rates. We calculated the optimal cut-point for the primary outcomes while maximizing the product of the sensitivity and specificity and then we evaluated the significance of secondary outcomes with logistic regression analysis.ResultsEighty-six patients were included in the analysis; 50 cases were non-functional PitNETs. The TPS for PD-L1 showed a highly right-skewed distribution in our sample, as 30.2% of patients scored 0. Using Trouillas' classification, we found that "proliferative" cases have a significantly higher probability to express PD-L1 in more than 30% of tumor cells (OR: 5.78; CI 95%: 1.80-18.4). This same cut-point was also associated with p53 expression. A positive association was found between PD-L1 expression and GH expression (p = 0.001; OR: 5.44; CI 95%: 1.98-14.98), while an inverse relationship was found with FSH/LH expression (p = 0.014; OR = 0.27, CI 95%: 0.10-0.76). No association was found with CS invasion, tumor size, bone erosion or dura invasion. We could not find any association between PD-L1 expression and recurrence.ConclusionsPD-L1 expression was associated with proliferative grades of Trouillas' classification and p53 expression. We also confirmed a higher expression of PD-L1 in somatotroph tumors. Larger studies are necessary to investigate the relationship between PD-L1 expression and aggressive behaviors.

Dataset Information

Adaptive threshold segmentation of pituitary adenomas from FDG PET images for radiosurgery.

Publications

Adaptive threshold segmentation of pituitary adenomas from FDG PET images for radiosurgery.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets