Predicting Intelligence Based on Cortical WM/GM Contrast, Cortical Thickness and Volumetry

Abstract

We propose a four-layer fully-connected neural network (FNN) for predicting fluid intelligence scores from T1-weighted MR images for the ABCD-challenge. In addition to the volumes of brain structures, the FNN uses cortical WM/GM contrast and cortical thickness at 78 cortical regions. These last two measurements were derived from the T1-weighted MR images using cortical surfaces produced by the CIVET pipeline. The age and gender of the subjects and the scanner manufacturer are also used as features for the learning algorithm. This yielded 283 features provided to the FNN with two hidden layers of 20 and 15 nodes. The method was applied to the data from the ABCD study. Trained with a training set of 3736 subjects, the proposed method achieved a MSE of 71.596 and a correlation of 0.151 in the validation set of 415 subjects. For the final submission, the model was trained with 3568 subjects and it achieved a MSE of 94.0270 in the test set comprised of 4383 subjects.

Can deep learning predict human intelligence from structural brain MRI?

“…As we mentioned in section that several studies used sMRI-based regional brain volumes as features in different machine learning methods to predict intelligence scores 4,[6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][27][28][29][30][31] . These studies used ∼8,500 healthy subjects for model training and then predicted the residual PIQ score of more than 3,500 adolescents with a mean square error (MSE) ranging from 86 to 103 (for a range of true residual PIQ score of [−40, 30]), or a correlation of 10% (p < 0.05).…”

“…Recent studies have started to use the machine especially deep learning to predict intelligence for individuals but left many open questions. In over 20 studies corresponding to a 2019 Grand Challenge, the predicted fluid intelligence had a mean square error ranging from 86 to 103 (for a range of true residual fluid intelligence score of [−40, 30]) [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] . The modest accuracy may suggest the need for more sophisticated deep learning algorithms.…”

Can deep learning predict human intelligence from structural brain MRI?

“…We also computed the residual scores to account for the influence of covariates (see below). Deep learning strategies that predict absolute and residual scores have been used in other machine learning studies 4,[6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][35][36][37][38][39] .…”

“…Recent studies have started to use machine learning, especially deep learning, to predict intelligence for individuals, but left many open questions. In more than 20 studies corresponding to a 2019 Grand Challenge, predicted fluid intelligence had mean square errors ranging from 86 to 103 (for a range of the true residual fluid intelligence score of [−40, 30]) [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] . The modest accuracy suggests the need for more sophisticated deep learning algorithms.…”

Inferring Neurocognition and Intelligence using Brain MRI

“…Cortical metrics are often combined with regional volumes to predict neurocognition. Studies using the ABCD dataset predicted the residual fluid intelligence score of more than 4500 adolescents with an MSE ranging from 93 to 95 (for a range of true residual fluid intelligence score of [-40, 30]), as summarized in [84][85][86][87][88][89] Significant positive correlations were also observed between the Reynolds intellectual assessment scales (RIAS) composite IQ scores and cortical gray matter volumes in the orbitofrontal gyrus (BAs 11, 12) (r = 0.41; p = 0.03), transverse temporal gyri (BAs 41, 42) (r = 0.42; p = 0.02), left superior temporal gyrus (BA 22) (r = 0.41; p = 0.04), and right anterior cingulate gyrus (BAs 24, 32, 33) (r = 0.42; p = 0.03). 90 The local gyrification and surface area in the superior parietal (BA 7), left supramarginal (BA 40), left caudal middle frontal (BA 22), left parsopercularis (BA 44), left inferior temporal (BA 20), right inferior and middle temporal (BA 21), right medial orbitofrontal (BAs 11, 12), and right rostral middle frontal (BA 10) regions are also found correlated to gF (r = 0.29; p < 0.001) and (r = 0.22; p < 0.001), respectively, and to gC (r = 0.28; p < 0.001) and (r = 0.28; p < 0.001), respectively, on a healthy young dataset (age = 21-35 years).…”