Voxel-based morphometry in single subjects without a scanner-specific normal database using a convolutional neural network

Abstract

Objectives: Reliable detection of disease-specific atrophy in individual T1w-MRI by voxel-based morphometry (VBM) requires scanner-specific normal databases (NDB), which often are not available. The aim of this retrospective study was to design, train, and test a deep convolutional neural network (CNN) for single-subject VBM without the need for a NDB (CNN-VBM).

Materials and methods: The training dataset comprised 8945 T1w scans from 65 different scanners. The gold standard VBM maps were obtained by conventional VBM with a scanner-specific NDB for each of the 65 scanners. CNN-VBM was tested in an independent dataset comprising healthy controls (n = 37) and subjects with Alzheimer's disease (AD, n = 51) or frontotemporal lobar degeneration (FTLD, n = 30). A scanner-specific NDB for the generation of the gold standard VBM maps was available also for the test set. The technical performance of CNN-VBM was characterized by the Dice coefficient of CNN-VBM maps relative to VBM maps from scanner-specific VBM. For clinical testing, VBM maps were categorized visually according to the clinical diagnoses in the test set by two independent readers, separately for both VBM methods.

Results: The VBM maps from CNN-VBM were similar to the scanner-specific VBM maps (median Dice coefficient 0.85, interquartile range [0.81, 0.90]). Overall accuracy of the visual categorization of the VBM maps for the detection of AD or FTLD was 89.8% for CNN-VBM and 89.0% for scanner-specific VBM.

Conclusion: CNN-VBM without NDB provides a similar performance in the detection of AD- and FTLD-specific atrophy as conventional VBM.

Clinical relevance statement: A deep convolutional neural network for voxel-based morphometry eliminates the need of scanner-specific normal databases without relevant performance loss and, therefore, could pave the way for the widespread clinical use of voxel-based morphometry to support the diagnosis of neurodegenerative diseases.

Key points: • The need of normal databases is a barrier for widespread use of voxel-based brain morphometry. • A convolutional neural network achieved a similar performance for detection of atrophy than conventional voxel-based morphometry. • Convolutional neural networks can pave the way for widespread clinical use of voxel-based morphometry.

Keywords: Alzheimer disease; Brain mapping; Deep learning; Magnetic resonance imaging; Neural networks (computer).

Fig. 1

The proposed network for CNN-based single-subject VBM. The T1w-MRI scan is re-sampled to a 3D volume with cubic voxels of 1 mm edge length. The CNN operates patch-wise with a patch size of 160 × 160 × 160 voxels. It uses a fully convolutional encoder-decoder architecture with 3D convolutions, residual-block connections, and four reductions of the feature map size. The CNN generates the VBM map in four disjoint parts (GM, gray matter)

Fig. 2

Standardized display for visual interpretation of the thresholded VBM map. The display combines transversal slices of the thresholded VBM map overlaid to the individual T1w-MRI scan and a glass brain view of the thresholded VBM map in a single-page pdf document. The example shows the CNN-VBM map of a patient with semantic variant primary progressive aphasia

Fig. 3

Quantitative comparison of the VBM maps in the independent test set: box-and-whisker plots of the Dice similarity coefficient between scanner-specific VBM and CNN-VBM (left) and between scanner-specific VBM and multiple-scanner VBM (right)