A 3D convolutional neural network to classify subjects as Alzheimer's disease, frontotemporal dementia or healthy controls using brain 18F-FDG PET.

With the arrival of disease-modifying drugs, neurodegenerative diseases will require an accurate diagnosis for optimal treatment. Convolutional neural networks are powerful deep learning techniques that can provide great help to physicians in image analysis. The purpose of this study is to introduce and validate a 3D neural network for classification of Alzheimer's disease (AD), frontotemporal dementia (FTD) or cognitively normal (CN) subjects based on brain glucose metabolism. Retrospective [18F]-FDG-PET scans of 199 CE, 192 FTD and 200 CN subjects were collected from our local database, Alzheimer's disease and frontotemporal lobar degeneration neuroimaging initiatives. Training and test sets were created using randomization on a 90 %-10 % basis, and training of a 3D VGG16-like neural network was performed using data augmentation and cross-validation. Performance was compared to clinical interpretation by three specialists in the independent test set. Regions determining classification were identified in an occlusion experiment and Gradient-weighted Class Activation Mapping. Test set subjects were age- and sex-matched across categories. The model achieved an overall 89.8 % accuracy in predicting the class of test scans. Areas under the ROC curves were 93.3 % for AD, 95.3 % for FTD, and 99.9 % for CN. The physicians' consensus showed a 69.5 % accuracy, and there was substantial agreement between them (kappa = 0.61, 95 % CI: 0.49-0.73). To our knowledge, this is the first study to introduce a deep learning model able to discriminate AD and FTD based on [18F]-FDG PET scans, and to isolate CN subjects with excellent accuracy. These initial results are promising and hint at the potential for generalization to data from other centers.

Myocardial, Retro-orbital, and Bilateral Testicular Metastases of an Ileal Neuroendocrine Tumor in 68 Ga-DOTATATE PET/CT.

ABSTRACT: A 71-year-old man with a newly discovered metastatic grade II neuroendocrine tumor of the terminal ileum was referred for a 68 Ga-DOTATATE PET/CT scan to stage the disease and assess suitability for PRRT (peptide receptor radionuclide therapy). The patient was known to have secondary nodal and bone/liver metastatic disease through prior morphological investigations. PET images revealed an atypical pattern of metastatic disease, showcasing secondary lesions in bilateral extraocular muscles, the myocardium, and both testes. The patient was pauci-symptomatic and only reported fatigue and diarrhea. Management involved lanreotide administration, and PRRT is being envisaged.

Optimization of Automated Radiosynthesis of Gallium-68-Labeled PSMA11 with Two [68Ge]Ge/[68Ga]Ga Generators: Fractional Elution or Prepurification?

Prostate cancer is one of the most common forms of cancer in men. An imaging technique for its diagnosis is [68Ga]-prostate-specific membrane antigen ([68Ga]Ga-PSMA-11) positron emission tomography (PET). To address the increasing demand for [68Ga]-labeled peptides and reduce the cost of radiosynthesis, it is therefore necessary to optimize the elution process of [68Ge]Ge/[68Ga]Ga generators. This study aims to identify the most effective approach for optimizing radiosynthesis using double elution in parallel of two [68Ge]Ge/[68Ga]Ga generators. Two methods have been tested: one using prepurification, and the other using fractionated elution. Five synthesis sequences were conducted using each method. The mean labeling yields for double elution with prepurification were 45.8 ± 29.4 (mean ± standard deviation) and none met the required criteria. The mean labeling yields for the fractionated double elution were 97.5 ± 1.9 (mean ± standard deviation) meeting the criteria, significantly superior to the prepurification method (p = 0.012), and similar to those of simple elution. This study showed that fractionated double elution from [68Ge]Ge/[68Ga]Ga generators produced a significantly higher labeling yield than double elution with prepurification, resulting in a larger activity recovered via radiosynthesis, thereby allowing more diagnostic tests to be performed.

Identifying and troubleshooting the pitfalls of ictal/interictal brain perfusion SPECT studies.

Epilepsy is a prevalent condition, and surgical intervention can benefit patients with refractory seizures. Single photon emission computed tomography (SPECT) using 99mTc-HMPAO or 99mTc-ECD provides assessment of regional cerebral blood flow and is the primary non-invasive approach for imaging brain perfusion in ictal and interictal states. Ictal/interictal SPECT is valuable in localising epileptogenic foci, particularly when MRI and electroencephalography are negative. However, to obtain accurate images reflecting brain perfusion in both states, meticulous preparation of the patient, timely radiotracer injection and close coordination between neurology and nuclear medicine teams are essential. Tracers also have inherent limitations, and patients may present with coexisting brain pathologies for which coregistration of SPECT images with MRI is recommended to improve diagnostic accuracy. Inconclusive SPECT findings may require repeating the exam or considering additional investigations. A comprehensive approach, considering various factors, is crucial for accurate interpretation of SPECT studies in presurgical epilepsy evaluations. This article provides a summary of the organisation and key challenges involved in conducting ictal/interictal SPECT studies, covering the entire process from a patient's hospital arrival to the integration of results within their presurgical pathway and using our experience of 182 patients over 10 years.

The relationship between glutamate, dopamine, and cortical gray matter: A simultaneous PET-MR study.

Prefrontal cortex has been shown to regulate striatal dopaminergic function via glutamatergic mechanisms in preclinical studies. Concurrent disruption of these systems is also often seen in neuropsychiatric disease. The simultaneous measurement of striatal dopamine signaling, cortical gray matter, and glutamate levels is therefore of major interest, but has not been previously reported. In the current study, twenty-eight healthy subjects underwent 2 simultaneous [11C]-( + )-PHNO PET-MRI scans, once after placebo and once after amphetamine in a double-blind randomized cross-over design, to measure striatal dopamine release, striatal dopamine receptor (D2/3R) availability, anterior cingulate glutamate+glutamine (Glx) levels, and cortical gray matter volumes at the same time. Voxel-based morphometry was used to investigate associations between neurochemical measures and gray matter volumes. Whole striatum D2/3R availability was positively associated with prefrontal cortex gray matter volume (pFWE corrected = 0.048). This relationship was mainly driven by associative receptor availability (pFWE corrected = 0.023). In addition, an interaction effect was observed between sensorimotor striatum D2/3R availability and anterior cingulate Glx, such that in individuals with greater anterior cingulate Glx concentrations, D2/3R availability was negatively associated with right frontal cortex gray matter volumes, while a positive D2/3R-gray matter association was observed in individuals with lower anterior cingulate Glx levels (pFWE corrected = 0.047). These results are consistent with the hypothesis that the prefrontal cortex is involved in regulation of striatal dopamine function. Furthermore, the observed associations raise the possibility that this regulation may be modulated by anterior cingulate glutamate concentrations.