MRI-PET dataset

该数据集包含了来自同一诊断阶段相同患者的MRI和PET扫描配对数据，共计2,843对。其中，MRI扫描为T1加权像，PET扫描使用的是18F-FDG，并且都经过了特定的预处理步骤。数据集规模为2,843对（训练集2,274对，验证集569对）。当前的任务是预测轻度认知障碍（MCI）进展为阿尔茨海默病（AD）的可能性。

This dataset contains 2,843 pairs of matched MRI and PET scans from the same patients at the same diagnostic stage. The MRI scans are T1-weighted images, the PET scans use 18F-FDG as the radiotracer, and all scans have undergone specific preprocessing steps. The dataset consists of 2,843 total scan pairs, with 2,274 pairs allocated to the training set and 569 pairs to the validation set. The current task is to predict the probability that mild cognitive impairment (MCI) progresses to Alzheimer's disease (AD).

搜集汇总

背景与挑战

背景概述

MRI-PET dataset是一个用于阿尔茨海默病多模态预测研究的数据集，包含MRI和PET的.nii.gz格式文件，用于训练生成模型和分类模型，支持GFE-Mamba模型进行MCI到AD的进展预测。