Intra-individual qualitative and quantitative comparison of [68Ga]Ga-DOTATATE PET/CT and PET/MRI.

Background: Somatostatin receptor (SSTR) positron emission tomography (PET) is a cornerstone of neuroendocrine tumor (NET) management. Hybrid PET/magnetic resonance imaging (MRI) is now available for NET-imaging, next to PET/computed tomography (CT). Objectives: To determine whether CT or MRI is the best hybrid partner for [68Ga]Ga-DOTATATE PET. Design: Monocentric, prospective study. Methods: Patients received a same-day [68Ga]Ga-DOTATATE PET/CT and subsequent PET/MRI, for suspicion of NET, (re)staging or peptide receptor radionuclide therapy-selection. The union (PETunion) of malignant lesions detected on PETCT and PETMRI was the reference standard. Concordance of detection of malignant lesions in an organ was measured between PETunion and CT and PETunion and MRI. Seven bins were used to categorize the number of malignant lesions, containing following ordinal variables: 0, 1, 2-5, 6-10, 11-20, >20 countable and diffuse/uncountable. The difference in number of malignant lesions was obtained as the difference in bin level ('Δbin') between PETunion and CT and PETunion and MRI with a Δbin closer to zero implying a higher concordance rate. Results: Twenty-nine patients were included. Primary tumors included 17 gastroenteropancreatic-NETs, 1 colon neuroendocrine carcinoma, 7 lung-NETs and 2 meningiomas. Patient level concordance with PETunion was 96% for MRI and 67% for CT (p = 0.039). Organ level concordance with PETunion was 74% for MRI and 40% for CT (p < 0.0001). In bone, there was a higher concordance rate for MRI compared to CT, 92% and 33%, respectively (p = 0.016). Overall, a mean Δbin of 0.5 ± 1.1 for PETunion/MRI and 1.4 ± 1.2 for PETunion/CT (p < 0.0001) was noted. In liver, a mean Δbin of 0.0 ± 1.1 for PETunion/MRI and 1.7 ± 1.2 for PETunion/CT was observed (p = 0.0078). In bone, a mean Δbin closer to zero was observed for PETunion/MRI compared to PETunion/CT, 0.6 ± 1.4 and 2.0 ± 1.5, respectively (p = 0.0098). Conclusions: Compared to SSTR PET/CT, SSTR PET/MRI had a higher patient and organ level concordance for malignant tumoral involvement and number of malignant lesions, with a clear added value in bone and liver specifically.

Prospective comparison of [18F]AlF-NOTA-octreotide PET/MRI to [68Ga]Ga-DOTATATE PET/CT in neuroendocrine tumor patients.

BACKGROUND: Fluorine-18-labeled SSAs have the potential to become the next-generation tracer in SSTR-imaging in neuroendocrine tumor (NET) patients given their logistical advantages over the current gold standard gallium-68-labeled SSAs. In particular, [18F]AlF-OC has already shown excellent clinical performance. We demonstrated in our previous report from our prospective multicenter trial that [18F]AlF-OC PET/CT outperforms [68Ga]Ga-DOTA-SSA, but histological confirmation was lacking due to ethical and practical reasons. In this second arm, we therefore aimed to provide evidence that the vast majority of [18F]AlF-OC PET lesions are in fact true NET lesions by analyzing their MR characteristics on simultaneously acquired MRI. We had a special interest in lesions solely detected by [18F]AlF-OC ("incremental lesions"). METHODS: Ten patients with a histologically confirmed neuroendocrine tumor (NET) and a standard-of-care [68Ga]Ga-DOTATATE PET/CT, performed within 3 months, were prospectively included. Patients underwent a whole-body PET/MRI (TOF, 3 T, GE Signa), 2 hours after IV injection of 4 MBq/kg [18F]AlF-OC. Positive PET lesions were evaluated for a corresponding lesion on MRI. The diagnostic performance of both PET tracers was evaluated by determining the detection ratio (DR) for each scan and the differential detection ratio (DDR) per patient. RESULTS: In total, 195 unique lesions were detected: 167 with [68Ga]Ga-DOTATATE and 193 with [18F]AlF-OC. The DR for [18F]AlF-OC was 99.1% versus 91.4% for [68Ga]Ga-DOTATATE, significant for non-inferiority testing (p = 0.0001). Out of these 193 [18F]AlF-OC lesions, 96.2% were confirmed by MRI to be NET lesions. Thirty-three incremental lesions were identified by [18F]AlF-OC, of which 91% were confirmed by MRI and considered true positives. CONCLUSION: The DR of [18F]AlF-OC was numerically higher and non-inferior to the DR of [68Ga]Ga-DOTATATE. [18F]AlF-OC lesions and especially incremental lesions were confirmed as true positives by MRI in more than 90% of lesions. Taken together, these data further validate [18F]AlF-OC as a new alternative for SSTR PET in clinical practice. Trial registration ClinicalTrials.gov: NCT04552847. Registered 17 September 2020, https://beta. CLINICALTRIALS: gov/study/NCT04552847.

A multi-label CNN model for the automatic detection and segmentation of gliomas using [18F]FET PET imaging.

PURPOSE: The aim of this study was to develop a convolutional neural network (CNN) for the automatic detection and segmentation of gliomas using [18F]fluoroethyl-L-tyrosine ([18F]FET) PET. METHODS: Ninety-three patients (84 in-house/7 external) who underwent a 20-40-min static [18F]FET PET scan were retrospectively included. Lesions and background regions were defined by two nuclear medicine physicians using the MIM software, such that delineations by one expert reader served as ground truth for training and testing the CNN model, while delineations by the second expert reader were used to evaluate inter-reader agreement. A multi-label CNN was developed to segment the lesion and background region while a single-label CNN was implemented for a lesion-only segmentation. Lesion detectability was evaluated by classifying [18F]FET PET scans as negative when no tumor was segmented and vice versa, while segmentation performance was assessed using the dice similarity coefficient (DSC) and segmented tumor volume. The quantitative accuracy was evaluated using the maximal and mean tumor to mean background uptake ratio (TBRmax/TBRmean). CNN models were trained and tested by a threefold cross-validation (CV) using the in-house data, while the external data was used for an independent evaluation to assess the generalizability of the two CNN models. RESULTS: Based on the threefold CV, the multi-label CNN model achieved 88.9% sensitivity and 96.5% precision for discriminating between positive and negative [18F]FET PET scans compared to a 35.3% sensitivity and 83.1% precision obtained with the single-label CNN model. In addition, the multi-label CNN allowed an accurate estimation of the maximal/mean lesion and mean background uptake, resulting in an accurate TBRmax/TBRmean estimation compared to a semi-automatic approach. In terms of lesion segmentation, the multi-label CNN model (DSC = 74.6 ± 23.1%) demonstrated equal performance as the single-label CNN model (DSC = 73.7 ± 23.2%) with tumor volumes estimated by the single-label and multi-label model (22.9 ± 23.6 ml and 23.1 ± 24.3 ml, respectively) closely approximating the tumor volumes estimated by the expert reader (24.1 ± 24.4 ml). DSCs of both CNN models were in line with the DSCs by the second expert reader compared with the lesion segmentations by the first expert reader, while detection and segmentation performance of both CNN models as determined with the in-house data were confirmed by the independent evaluation using external data. CONCLUSION: The proposed multi-label CNN model detected positive [18F]FET PET scans with high sensitivity and precision. Once detected, an accurate tumor segmentation and estimation of background activity was achieved resulting in an automatic and accurate TBRmax/TBRmean estimation, such that user interaction and potential inter-reader variability can be minimized.

Effect of sleeve gastrectomy and Roux-en-Y gastric bypass on gastrointestinal physiology.

BACKGROUND: Knowledge regarding the gastrointestinal physiology after sleeve gastrectomy and Roux-en-Y gastric bypass is urgently needed to understand, prevent and treat the nutritional and pharmacological complications of bariatric surgery. AIM: To investigate the effect of sleeve gastrectomy and Roux-en-Y gastric bypass on gastrointestinal motility (e.g., transit and pressure), pH, and intestinal bile acid concentration. MATERIAL AND METHODS: An exploratory cross-sectional study was performed in six participants living with obesity, six participants who underwent sleeve gastrectomy, and six participants who underwent Roux-en-Y gastric bypass. During the first visit, a wireless motility capsule (SmartPill©) was ingested after an overnight fast to measure gastrointestinal transit, pH, and pressure. During the second visit, a gastric emptying scintigraphy test of a nutritional drink labeled with 99mTc-colloid by a dual-head SPECT gamma camera was performed to measure gastric emptying half-time (GET1/2). During the third visit, two customized multiple lumen aspiration catheters were positioned to collect fasting and postprandial intestinal fluids to measure bile acid concentration. RESULTS: Immediate pouch emptying (P = 0.0007) and a trend for faster GET1/2 (P = 0.09) were observed in both bariatric groups. There was a tendency for a shorter orocecal transit in participants with sleeve gastrectomy and Roux-en-Y gastric bypass (P = 0.08). The orocecal segment was characterized by a higher 25th percentile pH (P = 0.004) and a trend for a higher median pH in both bariatric groups (P = 0.07). Fasting total bile acid concentration was 7.5-fold higher in the common limb after Roux-en-Y gastric bypass (P < 0.0001) and 3.5-fold higher in the jejunum after sleeve gastrectomy (P = 0.009) compared to obesity. Postprandial bile acid concentration was 3-fold higher in the jejunum after sleeve gastrectomy (P = 0.0004) and 6.5-fold higher in the common limb after Roux-en-Y gastric bypass (P < 0.0001) compared to obesity. CONCLUSION: The anatomical alterations of sleeve gastrectomy and Roux-en-Y gastric bypass have an important impact on gastrointestinal physiology. This data confirms changes in transit and pH and provides the first evidence for altered intraluminal bile acid concentration.

Whole liver segmentation based on deep learning and manual adjustment for clinical use in SIRT.

PURPOSE: In selective internal radiation therapy (SIRT), an accurate total liver segmentation is required for activity prescription and absorbed dose calculation. Our goal was to investigate the feasibility of using automatic liver segmentation based on a convolutional neural network (CNN) for CT imaging in SIRT, and the ability of CNN to reduce inter-observer variability of the segmentation. METHODS: A multi-scale CNN was modified for liver segmentation for SIRT patients. The CNN model was trained with 139 datasets from three liver segmentation challenges and 12 SIRT patient datasets from our hospital. Validation was performed on 13 SIRT datasets and 12 challenge datasets. The model was tested on 40 SIRT datasets. One expert manually delineated the livers and adjusted the liver segmentations from CNN for 40 test SIRT datasets. Another expert performed the same tasks for 20 datasets randomly selected from the 40 SIRT datasets. The CNN segmentations were compared with the manual and adjusted segmentations from the experts. The difference between the manual segmentations was compared with the difference between the adjusted segmentations to investigate the inter-observer variability. Segmentation difference was evaluated through dice similarity coefficient (DSC), volume ratio (RV), mean surface distance (MSD), and Hausdorff distance (HD). RESULTS: The CNN segmentation achieved a median DSC of 0.94 with the manual segmentation and of 0.98 with the manually corrected CNN segmentation, respectively. The DSC between the adjusted segmentations is 0.98, which is 0.04 higher than the DSC between the manual segmentations. CONCLUSION: The CNN model achieved good liver segmentations on CT images of good image quality, with relatively normal liver shapes and low tumor burden. 87.5% of the 40 CNN segmentations only needed slight adjustments for clinical use. However, the trained model failed on SIRT data with low dose or contrast, lesions with large density difference from their surroundings, and abnormal liver position and shape. The abovementioned scenarios were not adequately represented in the training data. Despite this limitation, the current CNN is already a useful clinical tool which improves inter-observer agreement and therefore contributes to the standardization of the dosimetry. A further improvement is expected when the CNN will be trained with more data from SIRT patients.

Combined brain and spinal FDG PET allows differentiation between ALS and ALS mimics.

PURPOSE: Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disorder with on average a 1-year delay between symptom onset and diagnosis. Studies have demonstrated the value of [18F]-FDG PET as a sensitive diagnostic biomarker, but the discriminatory potential to differentiate ALS from patients with symptoms mimicking ALS has not been investigated. We investigated the combination of brain and spine [18F]-FDG PET-CT for differential diagnosis between ALS and ALS mimics in a real-life clinical diagnostic setting. METHODS: Patients with a suspected diagnosis of ALS (n = 98; 64.8 ± 11 years; 61 M) underwent brain and spine [18F]-FDG PET-CT scans. In 62 patients, ALS diagnosis was confirmed (67.8 ± 10 years; 35 M) after longitudinal follow-up (average 18.1 ± 8.4 months). In 23 patients, another disease was diagnosed (ALS mimics, 60.9 ± 12.9 years; 17 M) and 13 had a variant motor neuron disease, primary lateral sclerosis (PLS; n = 4; 53.6 ± 2.5 years; 2 M) and progressive muscular atrophy (PMA; n = 9; 58.4 ± 7.3 years; 7 M). Spine metabolism was determined after manual and automated segmentation. VOI- and voxel-based comparisons were performed. Moreover, a support vector machine (SVM) approach was applied to investigate the discriminative power of regional brain metabolism, spine metabolism and the combination of both. RESULTS: Brain metabolism was very similar between ALS mimics and ALS, whereas cervical and thoracic spine metabolism was significantly different (in standardised uptake values; cervical: ALS 2.1 ± 0.5, ALS mimics 1.9 ± 0.4; thoracic: ALS 1.8 ± 0.3, ALS mimics 1.5 ± 0.3). As both brain and spine metabolisms were very similar between ALS mimics and PLS/PMA, groups were pooled for accuracy analyses. Mean discrimination accuracy was 65.4%, 80.0% and 81.5%, using only brain metabolism, using spine metabolism and using both, respectively. CONCLUSION: The combination of brain and spine FDG PET-CT with SVM classification is useful as discriminative biomarker between ALS and ALS mimics in a real-life clinical setting.