Improving the brain image resolution of generalized q-sampling MRI revealed by a three-dimensional CNN-based method.

Background: Understanding neural connections facilitates the neuroscience and cognitive behavioral research. There are many nerve fiber intersections in the brain that need to be observed, and the size is between 30 and 50 nanometers. Improving image resolution has become an important issue for mapping the neural connections non-invasively. Generalized q-sampling imaging (GQI) was used to reveal the fiber geometry of straight and crossing. In this work, we attempted to achieve super-resolution with a deep learning method on diffusion weighted imaging (DWI). Materials and methods: A three-dimensional super-resolution convolutional neural network (3D SRCNN) was utilized to achieve super-resolution on DWI. Then, generalized fractional anisotropy (GFA), normalized quantitative anisotropy (NQA), and the isotropic value of the orientation distribution function (ISO) mapping were reconstructed using GQI with super-resolution DWI. We also reconstructed the orientation distribution function (ODF) of brain fibers using GQI. Results: With the proposed super-resolution method, the reconstructed DWI was closer to the target image than the interpolation method. The peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) were also significantly improved. The diffusion index mapping reconstructed by GQI also had higher performance. The ventricles and white matter regions were much clearer. Conclusion: This super-resolution method can assist in postprocessing low-resolution images. With SRCNN, high-resolution images can be effectively and accurately generated. The method can clearly reconstruct the intersection structure in the brain connectome and has the potential to accurately describe the fiber geometry on a subvoxel scale.

Disrupted white matter network of brain structural connectomes in bipolar disorder patients revealed by q-ball imaging.

BACKGROUND: Structural and functional brain changes have been found to be associated with altered emotion and cognition in patients with bipolar disorder (BD). Widespread microstructural white matter abnormalities have been observed using traditional structural imaging in BD. q-Ball imaging (QBI) and graph theoretical analysis (GTA) improve the specificity and sensitivity and high accuracy of fiber tracking. We applied QBI and GTA to investigate and compare the structural connectivity alterations and network alterations in patients with and without BD. METHODS: Sixty-two patients with BD and 62 healthy controls (HCs) completed a MR scan. We evaluated the group differences in generalized fractional anisotropy (GFA) and normalized quantitative anisotropy (NQA) values by voxel-based statistical analysis with QBI. We also evaluated the group differences in topological parameters of GTA and subnetwork interconnections in network-based statistical analysis (NBS). RESULTS: The QBI indices in the BD group were significantly lower than those in the HC group in the corpus callosum, cingulate gyrus, and caudate. The GTA indices indicated that the BD group demonstrated less global integration and higher local segregation than the HC group, but they retained small-world properties. NBS evaluation showed that the majority of the more connected subnetworks in BD occurred in thalamo-temporal/parietal connectivity. CONCLUSION: Our findings supported white matter integrity with network alterations in BD.

Magnetic Resonance-Based Synthetic Computed Tomography Using Generative Adversarial Networks for Intracranial Tumor Radiotherapy Treatment Planning.

The purpose of this work is to develop a reliable deep-learning-based method that is capable of synthesizing needed CT from MRI for radiotherapy treatment planning. Simultaneously, we try to enhance the resolution of synthetic CT. We adopted pix2pix with a 3D framework, which is a conditional generative adversarial network, to map the MRI data domain into the CT data domain of our dataset. The original dataset contains paired MRI and CT images of 31 subjects; 26 pairs were used for model training and 5 were used for model validation. To identify the correctness of the synthetic CT of models, all of the synthetic CTs were calculated by the quantized image similarity formulas: cosine angle distance, Euclidean distance, mean square error, peak signal-to-noise ratio, and mean structural similarity. Two radiologists independently evaluated the satisfaction score, including spatial, detail, contrast, noise, and artifacts, for each imaging attribute. The mean (±standard deviation) of the structural similarity indices (CAD, L2 norm, MSE, PSNR, and MSSIM) between five real CT scans and the synthetic CT scans were 0.96 ± 0.015, 76.83 ± 12.06, 0.00118 ± 0.00037, 29.47 ± 1.35, and 0.84 ± 0.036, respectively. For synthetic CT, radiologists rated the results as evincing excellent satisfaction in spatial geometry and noise level, good satisfaction in contrast and artifacts, and fair imaging details. The similarity index and clinical evaluation results between synthetic CT and original CT guarantee the usability of the proposed method.

A CNN-Based Autoencoder and Machine Learning Model for Identifying Betel-Quid Chewers Using Functional MRI Features.

Betel quid (BQ) is one of the most commonly used psychoactive substances in some parts of Asia and the Pacific. Although some studies have shown brain function alterations in BQ chewers, it is virtually impossible for radiologists' to visually distinguish MRI maps of BQ chewers from others. In this study, we aimed to construct autoencoder and machine-learning models to discover brain alterations in BQ chewers based on the features of resting-state functional magnetic resonance imaging. Resting-state functional magnetic resonance imaging (rs-fMRI) was obtained from 16 BQ chewers, 15 tobacco- and alcohol-user controls (TA), and 17 healthy controls (HC). We used an autoencoder and machine learning model to identify BQ chewers among the three groups. A convolutional neural network (CNN)-based autoencoder model and supervised machine learning algorithm logistic regression (LR) were used to discriminate BQ chewers from TA and HC. Classifying the brain MRIs of HC, TA controls, and BQ chewers by conducting leave-one-out-cross-validation (LOOCV) resulted in the highest accuracy of 83%, which was attained by LR with two rs-fMRI feature sets. In our research, we constructed an autoencoder and machine-learning model that was able to identify BQ chewers from among TA controls and HC, which were based on data from rs-fMRI, and this might provide a helpful approach for tracking BQ chewers in the future.

Investigating the white matter correlates of reading performance: Evidence from Chinese children with reading difficulties.

PURPOSE: Reading comprehension is closely associated with word recognition, particularly at the early stage of reading development. This association is reflected in children with reading difficulties (RD) who demonstrate poor reading comprehension along with delayed word recognition or reduced recognition accuracy. Although the neural mechanisms underlying reading comprehension and word recognition are well studied, few has investigated the white matter (WM) structures that the two processes potentially share. METHODS: To explore the issue, behavioral scores (word recognition & reading comprehension) and diffusion spectrum imaging (DSI) were acquired from Chinese-speaking children with RD and their age-matched typically developing children. WM structures were measured with generalized fractional anisotropy and normalized quantitative anisotropy to optimize fiber tracking precision. RESULTS: The children with RD performed significantly poorer than the typically developing children in both behavioral tasks. Between group differences of WM structure were found in the right superior temporal gyrus, the left medial frontal gyrus, the left medial frontal gyrus, and the left caudate body. A significant association between reading comprehension and Chinese character recognition and the DSI indices were found in the corpus callosum. The findings demonstrated the microstructural difference between children with and without reading difficulties go beyond the well-established reading network. Further, the association between the WM integrity of the corpus callosum and the behavioral scores reveals the involvement of the WM structure in both tasks. CONCLUSION: It suggests the two reading-related skills have partially overlapped neural mechanism. Associating the corpus callosum with the reading skills leads to the reconsideration of the right hemisphere role in the typical reading process and, potentially, how it compensates for children with reading difficulties.

Convolutional Neural Network-Based Deep Learning Model for Predicting Differential Suicidality in Depressive Patients Using Brain Generalized q-Sampling Imaging.

OBJECTIVE: Suicide is a priority health problem. Suicide assessment depends on imperfect clinician assessment with minimal ability to predict the risk of suicide. Machine learning/deep learning provides an opportunity to detect an individual at risk of suicide to a greater extent than clinician assessment. The present study aimed to use deep learning of structural magnetic resonance imaging (MRI) to create an algorithm for detecting suicidal ideation and suicidal attempts. METHODS: We recruited 4 groups comprising a total of 186 participants: 33 depressive patients with suicide attempt (SA), 41 depressive patients with suicidal ideation (SI), 54 depressive patients without suicidal thoughts (DP), and 58 healthy controls (HCs). The confirmation of depressive disorder, SA and SI was based on psychiatrists' diagnosis and Mini-International Neuropsychiatric Interview (MINI) interviews. In the generalized q-sampling imaging (GQI) dataset, indices of generalized fractional anisotropy (GFA), the isotropic value of the orientation distribution function (ISO), and normalized quantitative anisotropy (NQA) were separately trained in convolutional neural network (CNN)-based deep learning and DenseNet models. RESULTS: From the results of 5-fold cross-validation, the best accuracies of the CNN classifier for predicting SA, SI, and DP against HCs were 0.916, 0.792, and 0.589, respectively. In SA-ISO, DenseNet outperformed the simple CNNs with a best accuracy from 5-fold cross-validation of 0.937. In SA-NQA, the best accuracy was 0.915. CONCLUSIONS: The results showed that a deep learning method based on structural MRI can effectively detect individuals at different levels of suicide risk, from depression to suicidal ideation and attempted suicide. Further studies from different populations, larger sample sizes, and prospective follow-up studies are warranted to confirm the utility of deep learning methods for suicide prevention and intervention.

An Autoencoder and Machine Learning Model to Predict Suicidal Ideation with Brain Structural Imaging.

It is estimated that at least one million people die by suicide every year, showing the importance of suicide prevention and detection. In this study, an autoencoder and machine learning model was employed to predict people with suicidal ideation based on their structural brain imaging. The subjects in our generalized q-sampling imaging (GQI) dataset consisted of three groups: 41 depressive patients with suicidal ideation (SI), 54 depressive patients without suicidal thoughts (NS), and 58 healthy controls (HC). In the GQI dataset, indices of generalized fractional anisotropy (GFA), isotropic values of the orientation distribution function (ISO), and normalized quantitative anisotropy (NQA) were separately trained in different machine learning models. A convolutional neural network (CNN)-based autoencoder model, the supervised machine learning algorithm extreme gradient boosting (XGB), and logistic regression (LR) were used to discriminate SI subjects from NS and HC subjects. After five-fold cross validation, separate data were tested to obtain the accuracy, sensitivity, specificity, and area under the curve of each result. Our results showed that the best pattern of structure across multiple brain locations can classify suicidal ideates from NS and HC with a prediction accuracy of 85%, a specificity of 100% and a sensitivity of 75%. The algorithms developed here might provide an objective tool to help identify suicidal ideation risk among depressed patients alongside clinical assessment.