PET and MRI Brain Dataset Containing Healthy and Disease Patients Data

The dataset comprised imaging data from 38 individuals, including 29 healthy controls and 9 patients diagnosed with neurological disorders. The patient cohort consisted of one individual with Hashimoto’s encephalopathy, one with Alzheimer’s disease, and the remaining individuals with epilepsy. All examinations were performed at Hautepierre Hospital, University Hospitals of Strasbourg, using a General Electric Healthcare SIGNA integrated PET/MRI scanner. Participants fasted for six hours before imaging, and blood glucose levels were confirmed to be below 1.3 g/L prior to radiotracer administration. Each participant received an intravenous injection of 2 MBq/kg of ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG), in accordance with the current recommendations of the European Association of Nuclear Medicine. Image acquisition began approximately 30 minutes after injection, and the total simultaneous PET/MRI acquisition time was 15 minutes. Time-of-flight PET images were acquired in three-dimensional mode using a 384 × 384 image matrix and an in-plane pixel size of 2.7 mm. The PET data were reconstructed using the iterative ordered-subset expectation maximization algorithm. The MRI protocol included three-dimensional T1-weighted gradient-echo, T2-weighted, fluid-attenuated inversion recovery, and arterial spin-labelling sequences, acquired at an isotropic spatial resolution of 1 mm. Data from the healthy participants were collected as part of clinical trial NCT053558573, with written informed consent and ethical approval for the ATLATEP study (No. SI 22.00287.000073; March 22, 2022). The use of patient clinico-radiological data was similarly authorized through informed consent and institutional review board approval for the INTARTEP study (IRB-2025-2).