Dataset Information

Automated liver lesion detection in ⁶⁸Ga DOTATATE PET/CT using a deep fully convolutional neural network.

ABSTRACT:

Background

Gastroenteropancreatic neuroendocrine tumors most commonly metastasize to the liver; however, high normal background ⁶⁸Ga-DOTATATE activity and high image noise make metastatic lesions difficult to detect. The purpose of this study is to develop a rapid, automated and highly specific method to identify ⁶⁸Ga-DOTATATE PET/CT hepatic lesions using a 2D U-Net convolutional neural network.

Methods

A retrospective study of ⁶⁸Ga-DOTATATE PET/CT patient studies (n = 125; 57 with ⁶⁸Ga-DOTATATE hepatic lesions and 68 without) was evaluated. The dataset was randomly divided into 75 studies for the training set (36 abnormal, 39 normal), 25 for the validation set (11 abnormal, 14 normal) and 25 for the testing set (11 abnormal, 14 normal). Hepatic lesions were physician annotated using a modified PERCIST threshold, and boundary definition by gradient edge detection. The 2D U-Net was trained independently five times for 100,000 iterations using a linear combination of binary cross-entropy and dice losses with a stochastic gradient descent algorithm. Performance metrics included: positive predictive value (PPV), sensitivity, F₁ score and area under the precision-recall curve (PR-AUC). Five different pixel area thresholds were used to filter noisy predictions.

Results

A total of 233 lesions were annotated with each abnormal study containing a mean of 4 ± 2.75 lesions. A pixel filter of 20 produced the highest mean PPV 0.94 ± 0.01. A pixel filter of 5 produced the highest mean sensitivity 0.74 ± 0.02. The highest mean F₁ score 0.79 ± 0.01 was produced with a 20 pixel filter. The highest mean PR-AUC 0.73 ± 0.03 was produced with a 15 pixel filter.

Conclusion

Deep neural networks can automatically detect hepatic lesions in ⁶⁸Ga-DOTATATE PET. Ongoing improvements in data annotation methods, increasing sample sizes and training methods are anticipated to further improve detection performance.

SUBMITTER: Wehrend J

PROVIDER: S-EPMC8487415 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Background[68 Ga]Ga-DOTATATE PET/CT is now recognised as the most sensitive functional imaging modality for the diagnosis of well-differentiated neuroendocrine tumours (NET) and can inform treatment with peptide receptor radionuclide therapy with [177Lu]Lu-DOTATATE. However, somatostatin receptor (SSTR) expression is not unique to NET, and therefore, [68 Ga]Ga-DOTATATE PET/CT may have oncological application in other tumours. Molecular profiling of gastrointestinal stromal tumours that lack activating somatic mutations in KIT or PDGFRA or so-called 'wild-type' GIST (wtGIST) has demonstrated that wtGIST and NET have overlapping molecular features and has encouraged exploration of shared therapeutic targets, due to a lack of effective therapies currently available for metastatic wtGIST.AimsTo investigate (i) the diagnostic role of [68 Ga]Ga-DOTATATE PET/CT; and, (ii) to investigate the potential of this imaging modality to guide treatment with [177Lu]Lu-DOTATATE in patients with wtGIST.Methods[68 Ga]Ga-DOTATATE PET/CT was performed on 11 patients with confirmed or metastatic wtGIST and one patient with a history of wtGIST and a mediastinal mass suspicious for metastatic wtGIST, who was subsequently diagnosed with a metachronous mediastinal paraganglioma. Tumour expression of somatostatin receptor subtype 2 (SSTR2) using immunohistochemistry was performed on 54 tumour samples including samples from 8/12 (66.6%) patients who took part in the imaging study and 46 tumour samples from individuals not included in the imaging study.Results[68 Ga]Ga-DOTATATE PET/CT imaging was negative, demonstrating that liver metastases had lower uptake than background liver for nine cases (9/12 cases, 75%) and heterogeneous uptake of somatostatin tracer was noted for two cases (16.6%) of wtGIST. However, [68 Ga]Ga-DOTATATE PET/CT demonstrated intense tracer uptake in a synchronous paraganglioma in one case and a metachronous paraganglioma in another case with wtGIST.ConclusionsOur data suggest that SSTR2 is not a diagnostic or therapeutic target in wtGIST. [68 Ga]Ga-DOTATATE PET/CT may have specific diagnostic utility in differentiating wtGIST from other primary tumours such as paraganglioma in patients with sporadic and hereditary forms of wtGIST.

Project description:ContextThe 68 Ga-labelled somatostatin analogues (68 Ga-DOTA-SSAs) is becoming popular as an important diagnostic tool in neuroendocrine tumours as evidenced by a growing number of reports detailing institutional experience with various DOTA peptides. However, only few prospective studies have compared 68 Ga-DOTA-SSAs and somatostatin receptor scintigraphy (SRS) in gastroenteropancreatic neuroendocrine tumours (GEP-NETs) and pulmonary neuroendocrine tumours.ObjectiveThe aim of our prospective study was to perform head-to-head comparison between 68 Ga-DOTATATE PET/CT and standard imaging work-up (SI) that included multiphasic CT, liver MRI and SRS using single photon emission computed tomography.DesignIn this prospective study, the patients were enrolled only if they met any of the following inclusion criteria: (i) initial staging of a NETs without distant metastases on SI or neuroendocrine tumour with unknown primary on SI; (ii) restaging of NETs that could be treated by focused therapeutic interventions; (iii) elevated serum neuroendocrine hormones or peptides. The exclusion criteria was grade 3 GEP-NETs.ResultsThirty-two patients were enrolled in the study. Eleven patients (6 pancreas, 4 ileum, 1 duodenal) were included for initial evaluation and staging of NETs, 8 patients (5 pancreas, 1 ileal, 1 lung, 1 duodenal gastrinoma) for restaging, and 13 patients for elevated serum neuroendocrine biomarkers (5 ectopic Cushing's syndrome, 5 organic hypoglycaemia, 1 patient each with elevated vasoactive inhibitory peptide, chromogranin A and neuron-specific enolase). 68 Ga-DOTATATE PET/CT detected more primary tumours than SRS (15/18 vs 10/18: P = .074). The missed tumours on 68 Ga-DOTATATE PET/CT were located in the lung in two cases and duodenum in one case. For other anatomical regions (nodal and distant metastasis), no statistical difference was observed between imaging modalities using 68 Ga-DOTATATE PET/CT and SRS. Overall, 68 Ga-DOTATATE PET/CT+CT+MRI detected 31/33 of the involved regions (including primaries) (29 and 22 for 68 Ga-DOTATATE and SRS, respectively).ConclusionOur study shows that 68 Ga-DOTATATE PET/CT detected similar number of sites than combination of SRS, liver MRI and thoraco-abdominopelvic CT on region-based analysis. 68 Ga-DOTATATE PET/CT missed half of primary lung carcinoids with ectopic Cushing's syndrome.

Project description:BackgroundInflammation drives atherosclerotic plaque rupture. Although inflammation can be measured using fluorine-18-labeled fluorodeoxyglucose positron emission tomography ([18F]FDG PET), [18F]FDG lacks cell specificity, and coronary imaging is unreliable because of myocardial spillover.ObjectivesThis study tested the efficacy of gallium-68-labeled DOTATATE (68Ga-DOTATATE), a somatostatin receptor subtype-2 (SST2)-binding PET tracer, for imaging atherosclerotic inflammation.MethodsWe confirmed 68Ga-DOTATATE binding in macrophages and excised carotid plaques. 68Ga-DOTATATE PET imaging was compared to [18F]FDG PET imaging in 42 patients with atherosclerosis.ResultsTarget SSTR2 gene expression occurred exclusively in "proinflammatory" M1 macrophages, specific 68Ga-DOTATATE ligand binding to SST2 receptors occurred in CD68-positive macrophage-rich carotid plaque regions, and carotid SSTR2 mRNA was highly correlated with in vivo 68Ga-DOTATATE PET signals (r = 0.89; 95% confidence interval [CI]: 0.28 to 0.99; p = 0.02). 68Ga-DOTATATE mean of maximum tissue-to-blood ratios (mTBRmax) correctly identified culprit versus nonculprit arteries in patients with acute coronary syndrome (median difference: 0.69; interquartile range [IQR]: 0.22 to 1.15; p = 0.008) and transient ischemic attack/stroke (median difference: 0.13; IQR: 0.07 to 0.32; p = 0.003). 68Ga-DOTATATE mTBRmax predicted high-risk coronary computed tomography features (receiver operating characteristics area under the curve [ROC AUC]: 0.86; 95% CI: 0.80 to 0.92; p < 0.0001), and correlated with Framingham risk score (r = 0.53; 95% CI: 0.32 to 0.69; p <0.0001) and [18F]FDG uptake (r = 0.73; 95% CI: 0.64 to 0.81; p < 0.0001). [18F]FDG mTBRmax differentiated culprit from nonculprit carotid lesions (median difference: 0.12; IQR: 0.0 to 0.23; p = 0.008) and high-risk from lower-risk coronary arteries (ROC AUC: 0.76; 95% CI: 0.62 to 0.91; p = 0.002); however, myocardial [18F]FDG spillover rendered coronary [18F]FDG scans uninterpretable in 27 patients (64%). Coronary 68Ga-DOTATATE PET scans were readable in all patients.ConclusionsWe validated 68Ga-DOTATATE PET as a novel marker of atherosclerotic inflammation and confirmed that 68Ga-DOTATATE offers superior coronary imaging, excellent macrophage specificity, and better power to discriminate high-risk versus low-risk coronary lesions than [18F]FDG. (Vascular Inflammation Imaging Using Somatostatin Receptor Positron Emission Tomography [VISION]; NCT02021188).

Project description:Purpose To evaluate dynamic gallium 68 (68Ga) tetraazacyclododecane tetraacetic acid octreotate (DOTATATE) brain PET/MRI as an adjunct modality in meningioma, enabling multiparametric standardized uptake value (SUV) and Patlak net binding rate constant (Ki) imaging, and to optimize static acquisition period. Materials and Methods In this prospective study (ClinicalTrials.gov no. NCT04081701, DOMINO-START), 68Ga-DOTATATE PET/MRI-derived time-activity curves (TACs) were measured in 84 volumes of interest in 19 participants (mean age, 63 years; range, 36-89 years; 13 women; 2019-2021) with meningiomas. Region- and voxel-specific Ki were determined using Patlak analysis with a validated population-based reference tissue TAC model built from an independent data set of nine participants. Mean and maximum absolute and relative-to-superior-sagittal-sinus SUVs were extracted from the entire 50 minutes (SUV50) and last 10 minutes (SUV10) of acquisition. SUV versus Ki Spearman correlation, SUV and Ki meningioma versus posttreatment-change Mann-Whitney U tests, and SUV50 versus SUV10 Wilcoxon matched-pairs signed rank tests were performed. Results Absolute and relative maximum SUV50 demonstrated a strong positive correlation with Patlak Ki in meningioma (r = 0.82, P < .001 and r = 0.85, P < .001, respectively) and posttreatment-change lesions (r = 0.88, P = .007 and r = 0.83, P = .02, respectively). Patlak Ki images yielded higher lesion contrast by mitigating nonspecific background signal. All SUV50 and SUV10 metrics differed between meningioma and posttreatment-change regions (P < .001). Within the meningioma group, SUV10 attained higher mean scores than SUV50 (P < .001). Conclusion Combined SUV and Patlak K i 68Ga-DOTATATE PET/MRI enabled multiparametric evaluation of meningioma, offering the potential to enhance lesion contrast with Ki imaging and optimize the SUV measurement postinjection window. Keywords: Molecular Imaging-Clinical Translation, Neuro-Oncology, PET/MRI, Dynamic, Patlak ClinicalTrials.gov registration no. NCT04081701 © RSNA, 2022.

Project description:Background and purposeMeningiomas express the somatostatin receptor (SSTR), which normal bone and brain lack. PET imaging with SSTR ligands such as 68 Ga-DOTATATE have been recently shown to aid in the imaging and identification of menginiomas. We hypothesize that 68 Ga-DOTATATE PET/CT in conjunction with MRI aids in radiation (RT) target volume delineation and evaluating treatment response.Materials and methodsNineteen patients with meningiomas underwent 68 Ga-DOTATATE PET/CT and MRI for RT planning and/or post-treatment follow-up. Meningiomas were grade I (n = 9) or not biopsied (n = 8) and frequently involved base of skull (n = 10). Ten (53%) patients received post-operative RT and 9 (47%) received fractaionted RT. In the subgroup that underwent both pre- and post-RT 68 Ga-DOTATATE PET as well as MRI (n = 10), ROVER (ABX GmbH, Radeberg, Germany) adaptive thresholding software was utilized to measure total lesion activity (mean and max) before and after treatment. Tumor volume based on MRI was calculated before and after treatment. Total lesion activity and tumor volume changes were compared using Wilcoxon signed rank test.Results68 Ga-DOTATATE PET/CT identified intraosseous (n = 4, 22%), falcine (n = 5, 26%) and satellite lesions (n = 3, 19%) and clarified the diagnosis of meningioma, resulting in a change in management in three patients. Mean total lesion activity decreased 14.7% (median), from pre to post-RT 68 Ga-DOTATATE PET [range 97-8.5% (25-75%),S = - 26.5, p = 0.0039]. Max total lesion activity decreased 36% (median) over the same period [range 105-15% (25-75%), S = - 26.5 p = 0.0039]. In contrast, meningioma volumes based on MRI measurements did not significantly change per RECIST criteria and Wilcoxon signed rank test (S = - 3, p = 0.7422).Conclusion68 Ga-DOTATATE PET/CT helped confirm suspected diagnoses and delineate target volumes particularly when lesions involved osseous structures and the falx. Mean and max total tumor 68 Ga-DOTATATE activity on PET/CT decreased at three months following RT despite stable tumor volumes on MRI. Future studies are warranted to (1) assess the sensitivity and specificity of 68 Ga-DOTATATE PET/CT, (2) evaluate the impact of 68 Ga-DOTATATE PET/CT-based planning on treatment outcomes, and (3) assess the prognostic significance of these post-treatment imaging changes.

Project description:Purpose68Ga-FAPI (fibroblast activation protein inhibitor) is a novel and highly promising radiotracer for PET/CT imaging. The aim of this retrospective analysis is to explore the potential of FAPI-PET/CT in gynecological malignancies. We assessed biodistribution, tumor uptake, and the influence of pre- or postmenopausal status on tracer accumulation in hormone-sensitive organs. Furthermore, a comparison with the current standard oncological tracer 18F-FDG was performed in selected cases.Patients and methodsA total of 31 patients (median age 59.5) from two centers with several gynecological tumors (breast cancer; ovarian cancer; cervical cancer; endometrial cancer; leiomyosarcoma of the uterus; tubal cancer) underwent 68Ga-FAPI-PET/CT. Out of 31 patients, 10 received an additional 18F-FDG scan within a median time interval of 12.5 days (range 1-76). Tracer uptake was quantified by standardized uptake values (SUV)max and (SUV)mean, and tumor-to-background ratio (TBR) was calculated (SUVmax tumor/ SUVmean organ). Moreover, a second cohort of 167 female patients with different malignancies was analyzed regarding their FAPI uptake in normal hormone-responsive organs: endometrium (n = 128), ovary (n = 64), and breast (n = 147). These patients were categorized by age as premenopausal (<35 years; n = 12), postmenopausal (>65 years; n = 68), and unknown menstrual status (35-65 years; n = 87), followed by an analysis of FAPI uptake of the pre- and postmenopausal group.ResultsIn 8 out of 31 patients, the primary tumor was present, and all 31 patients showed lesions suspicious for metastasis (n = 81) demonstrating a high mean SUVmax in both the primary (SUVmax 11.6) and metastatic lesions (SUVmax 9.7). TBR was significantly higher in 68Ga-FAPI compared to 18F-FDG for distant metastases (13.0 vs. 5.7; p = 0.047) and by trend for regional lymph node metastases (31.9 vs 27.3; p = 0.6). Biodistribution of 68Ga-FAPI-PET/CT presented significantly lower uptake or no significant differences in 15 out of 16 organs, compared to 18F-FDG-PET/CT. The highest uptake of all primary lesions was obtained in endometrial carcinomas (mean SUVmax 18.4), followed by cervical carcinomas (mean SUVmax 15.22). In the second cohort, uptake in premenopausal patients differed significantly from postmenopausal patients in endometrium (11.7 vs 3.9; p < 0.0001) and breast (1.8 vs 1.0; p = 0.004), whereas no significant difference concerning ovaries (2.8 vs 1.6; p = 0.141) was observed.ConclusionDue to high tracer uptake resulting in sharp contrasts in primary and metastatic lesions and higher TBRs than 18F-FDG-PET/CT, 68Ga-FAPI-PET/CT presents a promising imaging method for staging and follow-up of gynecological tumors. The presence or absence of the menstrual cycle seems to correlate with FAPI accumulation in the normal endometrium and breast. This first investigation of FAP ligands in gynecological tumor entities supports clinical application and further research in this field.

Dataset Information

Automated liver lesion detection in ⁶⁸Ga DOTATATE PET/CT using a deep fully convolutional neural network.

Background

Methods

Results

Conclusion

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets

Dataset Information

Automated liver lesion detection in 68Ga DOTATATE PET/CT using a deep fully convolutional neural network.

Background

Methods

Results

Conclusion

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets

Automated liver lesion detection in ⁶⁸Ga DOTATATE PET/CT using a deep fully convolutional neural network.