Search

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 (18)F-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 (18)F-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 (18)F-FDG PET/CT images with high accuracy.

Using bone‐avid radiotracers, cardiac transthyretin (TTR) amyloidosis can be diagnosed by scintigraphy, thus obviating endomyocardial biopsy. Radiotracer accumulation, however, may also be due to other causes. A 68‐year‐old male with acute myocardial infarction underwent recanalization of the left anterior descending coronary artery (LAD). Postinterventionally, transthoracic echocardiography showed hypokinesia of the septum and anterior wall and a thickened myocardium with granular sparkling appearance. Cardiac amyloidosis was suspected. A 99mTc‐3,3‐diphosphono‐1,2‐propanodicarboxylic acid whole‐body scan 4 days after LAD recanalization showed Perugini 2 myocardial tracer uptake. Monoclonal gammopathy was excluded, and cardiac TTR amyloidosis was diagnosed. Three months later, 99m‐Tc‐hydroxydiphosphate scan showed no myocardial tracer uptake. Cardiac magnetic resonance imaging revealed late gadolinium enhancement within the LAD supply area. No mutation of the TTR gene was found. Suspicion of amyloidosis should consider not only echocardiography but also history and clinical findings. Myocardial oedema due to reperfusion should be acknowledged as a differential diagnosis for cardiac uptake of bone‐avid radiotracers.

BACKGROUND The objective of this study was to characterize tumor activity and mineralization status in newly-detected multiple myeloma (MM) bone lesions using 2-18F-fluoro-2-deoxy-D-glucose (18F-FDG)-PET/CT and 18F-sodium fluoride (18F-NaF)-PET/CT before and after antitumor treatment. MATERIALS AND METHODS In this retrospective study, seven patients with histologically-verified MM were included (four women, three men; median age=57 years, standard deviation=11.23 years). PET/CT was performed with 18F-FDG and with 18F-NaF, both at baseline and after treatment. All patients had positive scans. Volumes of interest (VOIs) were drawn over all 18F-FDG-PET/CT-positive bone lesions, as well as the corresponding regions in 18F-NaF-PET/CT. For characterization of bone lesions, semi-quantitative standard uptake value (SUV) parameters were measured. RESULTS 18F-FDG-PET/CT in the seven patients detected 39 metabolically active lesions that were correlated with the corresponding sites in 18F-fluoride-PET/CT. Overall, the lesions showed a response to therapy, with a significant decrease in SUVmax on PET/CT using 18F-FDG (p

Simple Summary Breast cancer is the second most common diagnosed malignancy in women worldwide. In this study, we examine the feasibility of breast tumor characterization based on [18F]FDG-PET/CT images using machine learning (ML) approaches in combination with data-preprocessing techniques. ML prediction models for breast cancer detection and the identification of breast cancer receptor status, proliferation rate, and molecular subtypes were established and evaluated. Furthermore, the importance of most repeatable features was investigated. Results displayed high performance of malignant/benign tumor differentiation and triple negative tumor subtype ML models. We observed high repeatability of radiomic features for both high performing predictive models. Abstract Background: This study investigated the performance of ensemble learning holomic models for the detection of breast cancer, receptor status, proliferation rate, and molecular subtypes from [18F]FDG-PET/CT images with and without incorporating data pre-processing algorithms. Additionally, machine learning (ML) models were compared with conventional data analysis using standard uptake value lesion classification. Methods: A cohort of 170 patients with 173 breast cancer tumors (132 malignant, 38 benign) was examined with [18F]FDG-PET/CT. Breast tumors were segmented and radiomic features were extracted following the imaging biomarker standardization initiative (IBSI) guidelines combined with optimized feature extraction. Ensemble learning including five supervised ML algorithms was utilized in a 100-fold Monte Carlo (MC) cross-validation scheme. Data pre-processing methods were incorporated prior to machine learning, including outlier and borderline noisy sample detection, feature selection, and class imbalance correction. Feature importance in each model was assessed by calculating feature occurrence by the R-squared method across MC folds. Results: Cross validation demonstrated high performance of the cancer detection model (80% sensitivity, 78% specificity, 80% accuracy, 0.81 area under the curve (AUC)), and of the triple negative tumor identification model (85% sensitivity, 78% specificity, 82% accuracy, 0.82 AUC). The individual receptor status and luminal A/B subtype models yielded low performance (0.46–0.68 AUC). SUVmax model yielded 0.76 AUC in cancer detection and 0.70 AUC in predicting triple negative subtype. Conclusions: Predictive models based on [18F]FDG-PET/CT images in combination with advanced data pre-processing steps aid in breast cancer diagnosis and in ML-based prediction of the aggressive triple negative breast cancer subtype.

Background The blood flow (BF) is a critical determinant of organ functionality. Its assessment in the course of routine nuclear medicine examinations, including planar scintigraphy and positron emission tomography (PET), can be relevant for the diagnosis and monitoring of various conditions. The aim of this study was to investigate a new mathematical approach developed to estimate the organ BF from dynamic imaging data and to analyze if this method can be applied independent of the used radio tracer or imaging modality. The new approach uses the early phase of time activity curves extracted from animal and human dynamic scintigraphy and PET scans. Independence of tracer characteristics was evaluated with major oxygen-dependent organs (kidneys, liver, brain) of a mouse model. The approach was also applied on renal scans with two different imaging modalities from a representative cohort of 32 human subjects and compared to reference values. Results The mean organ-specific BF determined in the mouse model revealed no significant differences between the administered radiotracers and all calculated values corresponded to normal values (kidneys: 1.0-1.1 ml/min, liver: 1.4–1.6 ml/min, brain: 0.2 ml/min). In the human study cohort, the renal BFs from the two performed imaging modalities showed a good correlation (r = 0.61, p = 0.001) and a small significant difference (p = 0.047) among each other and good correlations to the reference value obtained from blood sampling (r = 0.79 and r = 0.52). Conclusions A mathematical approach was developed to assess the organ BF solely from dynamic imaging scans without the necessity of additional measurements. Preliminary data suggests that several radiotracers might be feasible to estimate the BF in major oxygen-dependent organs.

Background A method was developed to assess the kidney parameters glomerular filtration rate (GFR) and effective renal plasma flow (ERPF) from 2-deoxy-2-[18F]fluoro-d-glucose (FDG) concentration behavior in kidneys, measured with positron emission tomography (PET) scans. Twenty-four healthy adult subjects prospectively underwent dynamic simultaneous PET/magnetic resonance imaging (MRI) examination. Time activity curves (TACs) were obtained from the dynamic PET series, with the guidance of MR information. Patlak analysis was performed to determine the GFR, and based on integrals, ERPF was calculated. Results were compared to intra-individually obtained reference values determined from venous blood samples. Results Total kidney GFR and ERPF as estimated by dynamic PET/MRI were highly correlated to their reference values (r = 0.88/p

PURPOSE: In PIK3CA mutant breast cancer, downstream hyperactivation of the PI3K/AKT/mTOR pathway may be associated with increased glycolysis of cancer cells. The purpose of this study was to investigate the functional association of PIK3CA mutational status and tumor glycolysis in invasive ER+/HER2− early breast cancer. PROCEDURES: This institutional review board-approved retrospective study included a data set of 67 ER+/HER2−early breast cancer patients. All patients underwent 2-deoxy-2-[(18)F]fluoro-D-glucose positron emission tomography/x-ray computed tomography ([(18)F]FDG PET/CT) and clinico-pathologic assessments as part of a prospective study. For this retrospective analysis, pyrosequencing was used to detect PIK3CA mutations of exons 4, 7, 9, and 20. Tumor glucose metabolism was assessed semi-quantitatively with [(18)F]FDG PET/CT using maximum standardized uptake values (SUV(max)). SUV(max) values were corrected for the partial volume effect and metabolic tumor volume was calculated using the volume of interest automated lesion growing function 2D Tumor size, i.e., maximum tumor diameter was assessed on concurrent pre-treatment contrast-enhanced magnetic resonance imaging. RESULTS: PIK3CA mutations were present in 45% of all tumors. Mutations were associated with a small tumor diameter (p < 0.01) and with low nuclear grade (p = 0.04). Glycolytic activity was positively associated with nuclear grade (p = 0.01), proliferation (p = 0.002), regional lymph node metastasis (p = 0.015) and metabolic tumor volume (p = 0.001) but not with tumor size/T-stage. In invasive ductal carcinomas, median SUV(max) was increased in PIK3CA mutated compared to wild-type tumors; however, this increase did not reach statistical significance (p = 0.05). Multivariate analysis of invasive ductal carcinomas revealed [(18)F]FDG uptake to be independently associated with PIK3CA status (p = 0.002) and nuclear tumor grade (p = 0.046). Size, volume and regional nodal status had no influence on glycolytic activity. PIK3CA mutational status did not influence glycolytic metabolism in lobular carcinomas. Glycolytic activity and PIK3CA mutational status had no significant influence on Recurrence Free Survival or Disease Specific Survival. CONCLUSIONS: In ER+/HER2- invasive ductal carcinomas of the breast, glucose uptake is independently associated with PIK3CA mutations. Initial data suggest that [(18)F]FDG uptake reflects complex genomic alterations and may have the potential to be used as candidate biomarker for monitoring therapeutic response and resistance mechanisms in emerging therapies that target the PI3K/AKT/mTOR pathway.

In PIK3CA mutant breast cancer, downstream hyperactivation of the PI3K/AKT/mTOR pathway may be associated with increased glycolysis of cancer cells. The purpose of this study was to investigate the functional association of PIK3CA mutational status and tumor glycolysis in invasive ER+/HER2− early breast cancer. This institutional review board-approved retrospective study included a dataset of 67 ER+/HER2− early breast cancer patients. All patients underwent 2-deoxy-2-[18F]fluoro-D-glucose positron emission tomography/X-ray computed tomography ([18F]FDG PET/CT) and clinico-pathologic assessments as part of a prospective study. For this retrospective analysis, pyrosequencing was used to detect PIK3CA mutations of exons 4, 7, 9, and 20. Tumor glucose metabolism was assessed semi-quantitatively with [18F]FDG PET/CT using maximum standardized uptake values (SUVmax). SUVmax values were corrected for the partial volume effect, and metabolic tumor volume was calculated using the volume of interest automated lesion growing function 2D tumor size, i.e., maximum tumor diameter was assessed on concurrent pre-treatment contrast-enhanced magnetic resonance imaging. PIK3CA mutations were present in 45 % of all tumors. Mutations were associated with a small tumor diameter (p

Background Traditionally, isotope nephrography is considered as the method of choice to assess kidney function parameters in nuclear medicine. We propose a novel approach to determine the split function (SF), mean transit time (MTT), and outflow efficiency (OE) with 2-deoxy-2-[18F]fluoro-D-glucose (FDG) dynamic positron emission tomography (PET). Materials and methods Healthy adult subjects underwent dynamic simultaneous FDG-PET and magnetic resonance imaging (PET/MRI). Time-activity curves (TACs) of total kidneys, renal cortices, and the aorta were prospectively obtained from dynamic PET series. MRI images were used for anatomical correlation. The same individuals were subjected to dynamic renal Technetium-99 m-mercaptoacetyltriglycine (MAG3) scintigraphy and TACs of kidneys; the perirenal background and the left ventricle were determined. SF was calculated on the basis of integrals over the TACs, MTT was determined from renal retention functions after deconvolution analysis, and OE was determined from MTT. Values obtained from PET series were compared with scintigraphic parameters, which served as the reference. Results Twenty-four subjects underwent both examinations. Total kidney SF, MTT, and OE as estimated by dynamic PET/MRI correlated to their reference values by r = 0.75, r = 0.74 and r = 0.81, respectively, with significant difference (p

Aim We evaluated the clinical usefulness of 6-[18F]fluoro-3,4-dihydroxy-L-phenylalanine(18F-FDOPA)-positron-emission tomography (PET)/computed tomography (CT) in insulinoma detection with contrast enhancement, early acquisition time, and no carbidopa premedication. Patients and methods Twenty-six patients diagnosed with hyperinsulinemic hypoglycemia underwent an 18F-FDOPA PET/CT examination. Patients without carbidopa premedication and contrast-enhanced CT were included. Imaging findings were compared to the overall final diagnosis (histological findings). Results In 10 of 26 patients (eight women, two men; mean age=53 years; age range=30-94 years), a detected lesion was confirmed histologically as an insulinoma. 18F-FDOPA PET detected the tumor in five out of ten patients. Contrast-enhanced CT also detected the tumor in five out of ten. Overall, 18F-FDOPA PET/CT, with contrast enhancement and without carbidopa premedication, was able to detect the insulinoma in seven out of ten patients (70%). Conclusion Based on our data, 18F-DOPA PET/CT, with contrast enhancement and without carbidopa premedication, as a 'one-stop' diagnostic modality is a viable option for insulinoma detection.

To evaluate (18)F-fluorodihydroxyphenylalanine-positron emission tomography/ contrast-enhanced computed tomography ((18)F-DOPA PET/CT) for the detection of paragangliomas (PARA) without any patient selection, such as genetic predisposition for the development of these tumors, history of metastatic PARA or hormonal status.In this retrospective study, 28 consecutive patients (15 women, 13 men; mean age=46.4 years; age range=19-73 years), who were referred to our PET/CT center for the detection of clinically suspected PARA, were included. Final diagnosis was confirmed by histological reports of surgically proven lesions and/or clinical follow-up (including laboratory results and/or PET/CT follow-up).On a per-lesion basis (45 lesions) analysis, there was a sensitivity of 64.3% for CT, 73.8% for PET, 100% for PET/CT and a positive predictive value (PPV) of 93.1% for CT, 96.9% for PET and 100% for PET/CT. On a per-patient basis analysis, the sensitivity, specificity and accuracy for CT was 86.7%, 84.6% and 85.7%, respectively, and, for PET 80%, 100% and 89.3%, respectively, and, for PET/CT 100%.Based on our data, (18)F-DOPA PET/CT is a "one-stop diagnostic modality" for the assessment of patients with suspected PARA.