A Novel Transfer Learning Approach to Enhance Deep Neural Network Classification of Brain Functional Connectomes

Multiparametric MRI-based radiomics approach with deep transfer learning for preoperative prediction of Ki-67 status in sinonasal squamous cell carcinoma

Purpose

Based on comparison of different machine learning (ML) models, we developed the model that integrates traditional hand-crafted (HC) features and ResNet50 network-based deep transfer learning (DTL) features from multiparametric MRI to predict Ki-67 status in sinonasal squamous cell carcinoma (SNSCC).

Methods

Two hundred thirty-one SNSCC patients were retrospectively reviewed [training cohort (n = 185), test cohort (n = 46)]. Pathological grade, clinical, and MRI characteristics were analyzed to choose the independent predictor. HC and DTL radiomics features were extracted from fat-saturated T2-weighted imaging, contrast-enhanced T1-weighted imaging, and apparent diffusion coefficient map. Then, HC and DTL features were fused to formulate the deep learning-based radiomics (DLR) features. After feature selection and radiomics signature (RS) building, we compared the predictive ability of RS-HC, RS-DTL, and RS-DLR.

Results

No independent predictors were found based on pathological, clinical, and MRI characteristics. After feature selection, 42 HC and 10 DTL radiomics features were retained. The support vector machine (SVM), LightGBM, and ExtraTrees (ET) were the best classifier for RS-HC, RS-DTL, and RS-DLR. In the training cohort, the predictive ability of RS-DLR was significantly better than those of RS-DTL and RS-HC (p< 0.050); in the test set, the area under curve (AUC) of RS-DLR (AUC = 0.817) was also the highest, but there was no significant difference of the performance between DLR-RS and HC-RS.

Conclusions

Both the HC and DLR model showed favorable predictive efficacy for Ki-67 expression in patients with SNSCC. Especially, the RS-DLR model represented an opportunity to advance the prediction ability.

A Practical Roadmap to Implementing Deep Learning Segmentation in the Clinical Neuroimaging Research Workflow

BACKGROUND

Thanks to the proliferation of open-source tools, we are seeing an exponential growth of machine-learning applications, and its integration has become more accessible, particularly for segmentation tools in neuroimaging.

METHODS

This article explores a generalized methodology that harnesses these tools and aims/enables to expedite and enhance the reproducibility of clinical research. Herein, critical considerations include hardware, software, neural network training strategies, and data labeling guidelines. More specifically, we advocate an iterative approach to model training and transfer learning, focusing on internal validation and outlier handling early in the labeling process and fine-tuning later.

RESULTS

The iterative refinement process allows experts to intervene and improve model reliability while cutting down on their time spent in manual work. A seamless integration of the final model's predictions into clinical research is proposed to ensure standardized and reproducible results.

CONCLUSIONS

In short, this article provides a comprehensive framework for accelerating research using machine-learning techniques for image segmentation.

Supervised contrastive learning enhances graph convolutional networks for predicting neurodevelopmental deficits in very preterm infants using brain structural connectome

Very preterm (VPT) infants (born at less than 32 weeks gestational age) are at high risk for various adverse neurodevelopmental deficits. Unfortunately, most of these deficits cannot be accurately diagnosed until the age of 2-5 years old. Given the benefits of early interventions, accurate diagnosis and prediction soon after birth are urgently needed for VPT infants. Previous studies have applied deep learning models to learn the brain structural connectome (SC) to predict neurodevelopmental deficits in the preterm population. However, none of these models are specifically designed for graph-structured data, and thus may potentially miss certain topological information conveyed in the brain SC. In this study, we aim to develop deep learning models to learn the SC acquired at term-equivalent age for early prediction of neurodevelopmental deficits at 2 years corrected age in VPT infants. We directly treated the brain SC as a graph, and applied graph convolutional network (GCN) models to capture complex topological information of the SC. In addition, we applied the supervised contrastive learning (SCL) technique to mitigate the effects of the data scarcity problem, and enable robust training of GCN models. We hypothesize that SCL will enhance GCN models for early prediction of neurodevelopmental deficits in VPT infants using the SC. We used a regional prospective cohort of ∼280 VPT infants who underwent MRI examinations at term-equivalent age from the Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS). These VPT infants completed neurodevelopmental assessment at 2 years corrected age to evaluate cognition, language, and motor skills. Using the SCL technique, the GCN model achieved mean areas under the receiver operating characteristic curve (AUCs) in the range of 0.72∼0.75 for predicting three neurodevelopmental deficits, outperforming several competing models. Our results support our hypothesis that the SCL technique is able to enhance the GCN model in our prediction tasks.

The Use of Generative Adversarial Network and Graph Convolution Network for Neuroimaging-Based Diagnostic Classification

Functional connectivity (FC) obtained from resting-state functional magnetic resonance imaging has been integrated with machine learning algorithms to deliver consistent and reliable brain disease classification outcomes. However, in classical learning procedures, custom-built specialized feature selection techniques are typically used to filter out uninformative features from FC patterns to generalize efficiently on the datasets. The ability of convolutional neural networks (CNN) and other deep learning models to extract informative features from data with grid structure (such as images) has led to the surge in popularity of these techniques. However, the designs of many existing CNN models still fail to exploit the relationships between entities of graph-structure data (such as networks). Therefore, graph convolution network (GCN) has been suggested as a means for uncovering the intricate structure of brain network data, which has the potential to substantially improve classification accuracy. Furthermore, overfitting in classifiers can be largely attributed to the limited number of available training samples. Recently, the generative adversarial network (GAN) has been widely used in the medical field for its generative aspect that can generate synthesis images to cope with the problems of data scarcity and patient privacy. In our previous work, GCN and GAN have been designed to investigate FC patterns to perform diagnosis tasks, and their effectiveness has been tested on the ABIDE-I dataset. In this paper, the models will be further applied to FC data derived from more public datasets (ADHD, ABIDE-II, and ADNI) and our in-house dataset (PTSD) to justify their generalization on all types of data. The results of a number of experiments show the powerful characteristic of GAN to mimic FC data to achieve high performance in disease prediction. When employing GAN for data augmentation, the diagnostic accuracy across ADHD-200, ABIDE-II, and ADNI datasets surpasses that of other machine learning models, including results achieved with BrainNetCNN. Specifically, in ADHD, the accuracy increased from 67.74% to 73.96% with GAN, in ABIDE-II from 70.36% to 77.40%, and in ADNI, reaching 52.84% and 88.56% for multiclass and binary classification, respectively. GCN also obtains decent results, with the best accuracy in ADHD datasets at 71.38% for multinomial and 75% for binary classification, respectively, and the second-best accuracy in the ABIDE-II dataset (72.28% and 75.16%, respectively). Both GAN and GCN achieved the highest accuracy for the PTSD dataset, reaching 97.76%. However, there are still some limitations that can be improved. Both methods have many opportunities for the prediction and diagnosis of diseases.

Delineating Transdiagnostic Subtypes in Neurodevelopmental Disorders via Contrastive Graph Machine Learning of Brain Connectivity Patterns

Neurodevelopmental disorders, such as Attention Deficit/Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD), are characterized by comorbidity and heterogeneity. Identifying distinct subtypes within these disorders can illuminate the underlying neurobiological and clinical characteristics, paving the way for more tailored treatments. We adopted a novel transdiagnostic approach across ADHD and ASD, using cutting-edge contrastive graph machine learning to determine subtypes based on brain network connectivity as revealed by resting-state functional magnetic resonance imaging. Our approach identified two generalizable subtypes characterized by robust and distinct functional connectivity patterns, prominently within the frontoparietal control network and the somatomotor network. These subtypes exhibited pronounced differences in major cognitive and behavioural measures. We further demonstrated the generalizability of these subtypes using data collected from independent study sites. Our data-driven approach provides a novel solution for parsing biological heterogeneity in neurodevelopmental disorders.

Discovering individual fingerprints in resting-state functional connectivity using deep neural networks

Non-negligible idiosyncrasy due to interindividual differences is an ongoing issue in resting-state functional MRI (rfMRI) analysis. We show that a deep neural network (DNN) can be employed for individual identification by learning important features from the time-varying functional connectivity (FC) of rfMRI in the Human Connectome Project. We employed the trained DNN to identify individuals from an independent dataset acquired at our institution. The results revealed that the DNN could successfully identify 300 individuals with an error rate of 2.9% using 15 s time-window and 870 individuals with an error rate of 6.7%. A trained DNN with nonlinear hidden layers led to the proposal of the "fingerprint of FC" (fpFC) as representative edges of individual FC. The fpFCs for individuals exhibited commonly important and individual-specific edges across time-window lengths (from 5 min to 15 s). Furthermore, the utility of our model for another group of subjects was validated, supporting the feasibility of our technique in the context of transfer learning. In conclusion, our study offers an insight into the discovery of the intrinsic mode of the human brain using whole-brain resting-state FC and DNNs.

A comprehensive analysis towards exploring the promises of AI-related approaches in autism research

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition that presents challenges in communication, social interaction, repetitive behaviour, and limited interests. Detecting ASD at an early stage is crucial for timely interventions and an improved quality of life. In recent times, Artificial Intelligence (AI) has been increasingly used in ASD research. The rise in ASD diagnoses is due to the growing number of ASD cases and the recognition of the importance of early detection, which leads to better symptom management. This study explores the potential of AI in identifying early indicators of autism, aligning with the United Nations Sustainable Development Goals (SDGs) of Good Health and Well-being (Goal 3) and Peace, Justice, and Strong Institutions (Goal 16). The paper aims to provide a comprehensive overview of the current state-of-the-art AI-based autism classification by reviewing recent publications from the last decade. It covers various modalities such as Eye gaze, Facial Expression, Motor skill, MRI/fMRI, and EEG, and multi-modal approaches primarily grouped into behavioural and biological markers. The paper presents a timeline spanning from the history of ASD to recent developments in the field of AI. Additionally, the paper provides a category-wise detailed analysis of the AI-based application in ASD with a diagrammatic summarization to convey a holistic summary of different modalities. It also reports on the successes and challenges of applying AI for ASD detection while providing publicly available datasets. The paper paves the way for future scope and directions, providing a complete and systematic overview for researchers in the field of ASD.

Neuroimaging genetics approaches to identify new biomarkers for the early diagnosis of autism spectrum disorder

Autism-spectrum disorders (ASDs) are developmental disabilities that manifest in early childhood and are characterized by qualitative abnormalities in social behaviors, communication skills, and restrictive or repetitive behaviors. To explore the neurobiological mechanisms in ASD, extensive research has been done to identify potential diagnostic biomarkers through a neuroimaging genetics approach. Neuroimaging genetics helps to identify ASD-risk genes that contribute to structural and functional variations in brain circuitry and validate biological changes by elucidating the mechanisms and pathways that confer genetic risk. Integrating artificial intelligence models with neuroimaging data lays the groundwork for accurate diagnosis and facilitates the identification of early diagnostic biomarkers for ASD. This review discusses the significance of neuroimaging genetics approaches to gaining a better understanding of the perturbed neurochemical system and molecular pathways in ASD and how these approaches can detect structural, functional, and metabolic changes and lead to the discovery of novel biomarkers for the early diagnosis of ASD.

Source Free Semi-Supervised Transfer Learning for Diagnosis of Mental Disorders on fMRI Scans

The high prevalence of mental disorders gradually poses a huge pressure on the public healthcare services. Deep learning-based computer-aided diagnosis (CAD) has emerged to relieve the tension in healthcare institutions by detecting abnormal neuroimaging-derived phenotypes. However, training deep learning models relies on sufficient annotated datasets, which can be costly and laborious. Semi-supervised learning (SSL) and transfer learning (TL) can mitigate this challenge by leveraging unlabeled data within the same institution and advantageous information from source domain, respectively. This work is the first attempt to propose an effective semi-supervised transfer learning (SSTL) framework dubbed S3TL for CAD of mental disorders on fMRI data. Within S3TL, a secure cross-domain feature alignment method is developed to generate target-related source model in SSL. Subsequently, we propose an enhanced dual-stage pseudo-labeling approach to assign pseudo-labels for unlabeled samples in target domain. Finally, an advantageous knowledge transfer method is conducted to improve the generalization capability of the target model. Comprehensive experimental results demonstrate that S3TL achieves competitive accuracies of 69.14%, 69.65%, and 72.62% on ABIDE-I, ABIDE-II, and ADHD-200 datasets, respectively. Furthermore, the simulation experiments also demonstrate the application potential of S3TL through model interpretation analysis and federated learning extension.

Multimodal Autism Spectrum Disorder Diagnosis Method Based on DeepGCN

Multimodal data play an important role in the diagnosis of brain diseases. This study constructs a whole-brain functional connectivity network based on functional MRI data, uses non-imaging data with demographic information to complement the classification task for diagnosing subjects, and proposes a multimodal and across-site WL-DeepGCN-based method for classification to diagnose autism spectrum disorder (ASD). This method is used to resolve the existing problem that deep learning ASD identification cannot efficiently utilize multimodal data. In the WL-DeepGCN, a weight-learning network is used to represent the similarity of non-imaging data in the latent space, introducing a new approach for constructing population graph edge weights, and we find that it is beneficial and robust to define pairwise associations in the latent space rather than the input space. We propose a graph convolutional neural network residual connectivity approach to reduce the information loss due to convolution operations by introducing residual units to avoid gradient disappearance and gradient explosion. Furthermore, an EdgeDrop strategy makes the node connections sparser by randomly dropping edges in the raw graph, and its introduction can alleviate the overfitting and oversmoothing problems in the DeepGCN training process. We compare the WL-DeepGCN model with competitive models based on the same topics and nested 10-fold cross-validation show that our method achieves 77.27% accuracy and 0.83 AUC for ASD identification, bringing substantial performance gains.

Diagnosis of motor function injury based on near-infrared spectroscopy brain imaging (fNIRS) technology

Most clinical stroke patients may have difficulty moving, affecting their self-care ability and quality of life, and causing serious interference with the normal life and work of other family members. At present, in clinical literature, researchers provide functional training for patients with motor disorders through repeated and effective training, which can ultimately effectively promote the recovery of limb function. Therefore, the near-infrared spectroscopy imaging technology (fNIRS) used in this study combines the diagnosis of sports injury with the mechanism of brain function. FNIRS technology has many advantages, such as fast, and non-invasive, and has shown great value in detecting brain activity. Therefore, it has become a promising method in the biomedical field, especially in the field of brain science. Based on the clinical effects of sports injury treatment, fNIRS technology is used to detect the hemodynamic changes of hemoglobin circulation in the patient's brain tissue during training, and to detect the brain activity mechanism in the exercise mechanism, providing a basis for the clinical application of this method.

A novel collaborative self-supervised learning method for radiomic data

The computer-aided disease diagnosis from radiomic data is important in many medical applications. However, developing such a technique relies on labeling radiological images, which is a time-consuming, labor-intensive, and expensive process. In this work, we present the first novel collaborative self-supervised learning method to solve the challenge of insufficient labeled radiomic data, whose characteristics are different from text and image data. To achieve this, we present two collaborative pretext tasks that explore the latent pathological or biological relationships between regions of interest and the similarity and dissimilarity of information between subjects. Our method collaboratively learns the robust latent feature representations from radiomic data in a self-supervised manner to reduce human annotation efforts, which benefits the disease diagnosis. We compared our proposed method with other state-of-the-art self-supervised learning methods on a simulation study and two independent datasets. Extensive experimental results demonstrated that our method outperforms other self-supervised learning methods on both classification and regression tasks. With further refinement, our method will have the potential advantage in automatic disease diagnosis with large-scale unlabeled data available.

Sampling inequalities affect generalization of neuroimaging-based diagnostic classifiers in psychiatry

BACKGROUND

The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.

METHODS

Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies (n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.

RESULTS

A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient (G) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels (β =  - 2.75, p < .001, R²_adj = 0.40; r =  - .84, 95% CI: - .41 to - .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0-87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2-56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9-90.8%)/availability (80.88% of models, 95% CI: 77.3-84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF₁₀ > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.

CONCLUSIONS

Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.

A Semi-Supervised Graph Convolutional Network for Early Prediction of Motor Abnormalities in Very Preterm Infants

Approximately 32-42% of very preterm infants develop minor motor abnormalities. Earlier diagnosis soon after birth is urgently needed because the first two years of life represent a critical window of opportunity for early neuroplasticity in infants. In this study, we developed a semi-supervised graph convolutional network (GCN) model that is able to simultaneously learn the neuroimaging features of subjects and consider the pairwise similarity between them. The semi-supervised GCN model also allows us to combine labeled data with additional unlabeled data to facilitate model training. We conducted our experiments on a multisite regional cohort of 224 preterm infants (119 labeled subjects and 105 unlabeled subjects) who were born at 32 weeks or earlier from the Cincinnati Infant Neurodevelopment Early Prediction Study. A weighted loss function was applied to mitigate the impact of an imbalanced positive:negative (~1:2) subject ratio in our cohort. With only labeled data, our GCN model achieved an accuracy of 66.4% and an AUC of 0.67 in the early prediction of motor abnormalities, outperforming prior supervised learning models. By taking advantage of additional unlabeled data, the GCN model had significantly better accuracy (68.0%, p = 0.016) and a higher AUC (0.69, p = 0.029). This pilot work suggests that the semi-supervised GCN model can be utilized to aid early prediction of neurodevelopmental deficits in preterm infants.

Evaluation of Risk of Bias in Neuroimaging-Based Artificial Intelligence Models for Psychiatric Diagnosis: A Systematic Review

IMPORTANCE

Neuroimaging-based artificial intelligence (AI) diagnostic models have proliferated in psychiatry. However, their clinical applicability and reporting quality (ie, feasibility) for clinical practice have not been systematically evaluated.

OBJECTIVE

To systematically assess the risk of bias (ROB) and reporting quality of neuroimaging-based AI models for psychiatric diagnosis.

EVIDENCE REVIEW

PubMed was searched for peer-reviewed, full-length articles published between January 1, 1990, and March 16, 2022. Studies aimed at developing or validating neuroimaging-based AI models for clinical diagnosis of psychiatric disorders were included. Reference lists were further searched for suitable original studies. Data extraction followed the CHARMS (Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A closed-loop cross-sequential design was used for quality control. The PROBAST (Prediction Model Risk of Bias Assessment Tool) and modified CLEAR (Checklist for Evaluation of Image-Based Artificial Intelligence Reports) benchmarks were used to systematically evaluate ROB and reporting quality.

FINDINGS

A total of 517 studies presenting 555 AI models were included and evaluated. Of these models, 461 (83.1%; 95% CI, 80.0%-86.2%) were rated as having a high overall ROB based on the PROBAST. The ROB was particular high in the analysis domain, including inadequate sample size (398 of 555 models [71.7%; 95% CI, 68.0%-75.6%]), poor model performance examination (with 100% of models lacking calibration examination), and lack of handling data complexity (550 of 555 models [99.1%; 95% CI, 98.3%-99.9%]). None of the AI models was perceived to be applicable to clinical practices. Overall reporting completeness (ie, number of reported items/number of total items) for the AI models was 61.2% (95% CI, 60.6%-61.8%), and the completeness was poorest for the technical assessment domain with 39.9% (95% CI, 38.8%-41.1%).

CONCLUSIONS AND RELEVANCE

This systematic review found that the clinical applicability and feasibility of neuroimaging-based AI models for psychiatric diagnosis were challenged by a high ROB and poor reporting quality. Particularly in the analysis domain, ROB in AI diagnostic models should be addressed before clinical application.

Detection of autism spectrum disorder using graph representation learning algorithms and deep neural network, based on fMRI signals

Introduction

Can we apply graph representation learning algorithms to identify autism spectrum disorder (ASD) patients within a large brain imaging dataset? ASD is mainly identified by brain functional connectivity patterns. Attempts to unveil the common neural patterns emerged in ASD are the essence of ASD classification. We claim that graph representation learning methods can appropriately extract the connectivity patterns of the brain, in such a way that the method can be generalized to every recording condition, and phenotypical information of subjects. These methods can capture the whole structure of the brain, both local and global properties.

Methods

The investigation is done for the worldwide brain imaging multi-site database known as ABIDE I and II (Autism Brain Imaging Data Exchange). Among different graph representation techniques, we used AWE, Node2vec, Struct2vec, multi node2vec, and Graph2Img. The best approach was Graph2Img, in which after extracting the feature vectors representative of the brain nodes, the PCA algorithm is applied to the matrix of feature vectors. The classifier adapted to the features embedded in graphs is an LeNet deep neural network.

Results and discussion

Although we could not outperform the previous accuracy of 10-fold cross-validation in the identification of ASD versus control patients in this dataset, for leave-one-site-out cross-validation, we could obtain better results (our accuracy: 80%). The result is that graph embedding methods can prepare the connectivity matrix more suitable for applying to a deep network.

Leveraging Computational Intelligence Techniques for Diagnosing Degenerative Nerve Diseases: A Comprehensive Review, Open Challenges, and Future Research Directions

Degenerative nerve diseases such as Alzheimer's and Parkinson's diseases have always been a global issue of concern. Approximately 1/6th of the world's population suffers from these disorders, yet there are no definitive solutions to cure these diseases after the symptoms set in. The best way to treat these disorders is to detect them at an earlier stage. Many of these diseases are genetic; this enables machine learning algorithms to give inferences based on the patient's medical records and history. Machine learning algorithms such as deep neural networks are also critical for the early identification of degenerative nerve diseases. The significant applications of machine learning and deep learning in early diagnosis and establishing potential therapies for degenerative nerve diseases have motivated us to work on this review paper. Through this review, we covered various machine learning and deep learning algorithms and their application in the diagnosis of degenerative nerve diseases, such as Alzheimer's disease and Parkinson's disease. Furthermore, we also included the recent advancements in each of these models, which improved their capabilities for classifying degenerative nerve diseases. The limitations of each of these methods are also discussed. In the conclusion, we mention open research challenges and various alternative technologies, such as virtual reality and Big data analytics, which can be useful for the diagnosis of degenerative nerve diseases.

Interpreting mental state decoding with deep learning models

In mental state decoding, researchers aim to identify the set of mental states (e.g., experiencing happiness or fear) that can be reliably identified from the activity patterns of a brain region (or network). Deep learning (DL) models are highly promising for mental state decoding because of their unmatched ability to learn versatile representations of complex data. However, their widespread application in mental state decoding is hindered by their lack of interpretability, difficulties in applying them to small datasets, and in ensuring their reproducibility and robustness. We recommend approaching these challenges by leveraging recent advances in explainable artificial intelligence (XAI) and transfer learning, and also provide recommendations on how to improve the reproducibility and robustness of DL models in mental state decoding.

Diagnostic model optimization method for ADHD based on brain network analysis of resting-state fMRI images and transfer learning neural network

Introduction: Attention deficit and hyperactivity disorder (ADHD) is a common inherited disease of the nervous system whose cause(s) and pathogenesis remain unclear. Currently, the diagnosis of ADHD is mainly based on clinical experience and guidelines that have laid out some diagnostic standards. Our study aimed to apply a learning-based classification method to assist the ADHD diagnosis based on high-dimensional resting-state fMRI. Methods: Our study selected the ADHD-200 Peking dataset of resting-state fMRI, which has an ADHD patient (n = 142) group and a typically developing control (TDC) healthy control (n = 102) group. We first used Pearson and partial correlation coefficients to perform functional connectivity (FC) analysis between ROIs. Then, the Pearson and partial correlation coefficient matrices were concatenated into a dual-channel feature to build a dual data channel as input to the transfer learning neural network (TLNN) architecture. Finally, we transferred the pretrained model from the auxiliary domain to our target domain and fine-tuned it. Results: Based on the Pearson correlation coefficient, FC between ROIs was detected in 22 brain regions, including the fusiform gyrus, superior frontal gyrus, posterior superior temporal sulcus, inferior parietal lobule, anterior cingulate cortex, and parahippocampal gyrus. Based on the partial correlation coefficient, we found FC in the salient network, default network, sensory-motor network, dorsal attention network, and cerebellum network. With the TLNN architecture, we solved the problem of insufficient training data and improved the sensitivity of the classification method. When the VGG model (fine-tuned transfer strategy, 1,024 fully connected layers) was applied, the accuracy of TLNN classification ultimately reached 82%. Conclusion: Our study suggests that completing the training of the target domain by transferring the prior knowledge of the auxiliary domain is effective in solving the classification problem of small sample datasets. Based on prior knowledge of FC analysis, TLNN classification may assist ADHD diagnosis in a new way.

A Multiview Deep Learning Method for Brain Functional Connectivity Classification

The brain functional connectivity classification based on deep learning is a research hotspot nowadays. However, the classification performance is far behind the demand of clinical applications. To alleviate the problem, this paper proposes a multiview deep learning method for brain functional connectivity classification. Firstly, the proposed method adopts multiple brain atlases to identify brain regions and thereby builds different brain functional connectivity of different views. Secondly, it uses a multiview feature selection strategy to select out the most discriminative features of each view with the assistance of other views. Then, it trains a stacked autoencoder to extract deep features of the brain functional connectivity of each view. At last, it utilizes a multiview fusion strategy to take full advantage of complementary information of different views for brain functional connectivity classification. The proposed method has been compared with several deep learning-based brain functional connectivity classification methods on three public datasets of neuropsychiatric disorders. The experimental results have validated the superior performance of the proposed method.

Automatic autism spectrum disorder detection using artificial intelligence methods with MRI neuroimaging: A review

Autism spectrum disorder (ASD) is a brain condition characterized by diverse signs and symptoms that appear in early childhood. ASD is also associated with communication deficits and repetitive behavior in affected individuals. Various ASD detection methods have been developed, including neuroimaging modalities and psychological tests. Among these methods, magnetic resonance imaging (MRI) imaging modalities are of paramount importance to physicians. Clinicians rely on MRI modalities to diagnose ASD accurately. The MRI modalities are non-invasive methods that include functional (fMRI) and structural (sMRI) neuroimaging methods. However, diagnosing ASD with fMRI and sMRI for specialists is often laborious and time-consuming; therefore, several computer-aided design systems (CADS) based on artificial intelligence (AI) have been developed to assist specialist physicians. Conventional machine learning (ML) and deep learning (DL) are the most popular schemes of AI used for diagnosing ASD. This study aims to review the automated detection of ASD using AI. We review several CADS that have been developed using ML techniques for the automated diagnosis of ASD using MRI modalities. There has been very limited work on the use of DL techniques to develop automated diagnostic models for ASD. A summary of the studies developed using DL is provided in the Supplementary Appendix. Then, the challenges encountered during the automated diagnosis of ASD using MRI and AI techniques are described in detail. Additionally, a graphical comparison of studies using ML and DL to diagnose ASD automatically is discussed. We suggest future approaches to detecting ASDs using AI techniques and MRI neuroimaging.

Current and emerging artificial intelligence applications for pediatric abdominal imaging

Artificial intelligence (AI) uses computers to mimic cognitive functions of the human brain, allowing inferences to be made from generally large datasets. Traditional machine learning (e.g., decision tree analysis, support vector machines) and deep learning (e.g., convolutional neural networks) are two commonly employed AI approaches both outside and within the field of medicine. Such techniques can be used to evaluate medical images for the purposes of automated detection and segmentation, classification tasks (including diagnosis, lesion or tissue characterization, and prediction), and image reconstruction. In this review article we highlight recent literature describing current and emerging AI methods applied to abdominal imaging (e.g., CT, MRI and US) and suggest potential future applications of AI in the pediatric population.

A self-training deep neural network for early prediction of cognitive deficits in very preterm infants using brain functional connectome data

BACKGROUND

Deep learning has been employed using brain functional connectome data for evaluating the risk of cognitive deficits in very preterm infants. Although promising, training these deep learning models typically requires a large amount of labeled data, and labeled medical data are often very difficult and expensive to obtain.

OBJECTIVE

This study aimed to develop a self-training deep neural network (DNN) model for early prediction of cognitive deficits at 2 years of corrected age in very preterm infants (gestational age ≤32 weeks) using both labeled and unlabeled brain functional connectome data.

MATERIALS AND METHODS

We collected brain functional connectome data from 343 very preterm infants at a mean (standard deviation) postmenstrual age of 42.7 (2.5) weeks, among whom 103 children had a cognitive assessment at 2 years (i.e. labeled data), and the remaining 240 children had not received 2-year assessments at the time this study was conducted (i.e. unlabeled data). To develop a self-training DNN model, we built an initial student model using labeled brain functional connectome data. Then, we applied the trained model as a teacher model to generate pseudo-labels for unlabeled brain functional connectome data. Next, we combined labeled and pseudo-labeled data to train a new student model. We iterated this procedure to obtain the best student model for the early prediction task in very preterm infants.

RESULTS

In our cross-validation experiments, the proposed self-training DNN model achieved an accuracy of 71.0%, a specificity of 71.5%, a sensitivity of 70.4% and an area under the curve of 0.75, significantly outperforming transfer learning models through pre-training approaches.

CONCLUSION

We report the first self-training prognostic study in very preterm infants, efficiently utilizing a small amount of labeled data with a larger share of unlabeled data to aid the model training. The proposed technique is expected to facilitate deep learning with insufficient training data.

Machine learning for autism spectrum disorder diagnosis using structural magnetic resonance imaging: Promising but challenging

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder that affects approximately 1% of the population and causes significant burdens. ASD's pathogenesis remains elusive; hence, diagnosis is based on a constellation of behaviors. Structural magnetic resonance imaging (sMRI) studies have shown several abnormalities in volumetric and geometric features of the autistic brain. However, inconsistent findings prevented most contributions from being translated into clinical practice. Establishing reliable biomarkers for ASD using sMRI is crucial for the correct diagnosis and treatment. In recent years, machine learning (ML) and specifically deep learning (DL) have quickly extended to almost every sector, notably in disease diagnosis. Thus, this has led to a shift and improvement in ASD diagnostic methods, fulfilling most clinical diagnostic requirements. However, ASD discovery remains difficult. This review examines the ML-based ASD diagnosis literature over the past 5 years. A literature-based taxonomy of the research landscape has been mapped, and the major aspects of this topic have been covered. First, we provide an overview of ML's general classification pipeline and the features of sMRI. Next, representative studies are highlighted and discussed in detail with respect to methods, and biomarkers. Finally, we highlight many common challenges and make recommendations for future directions. In short, the limited sample size was the main obstacle; Thus, comprehensive data sets and rigorous methods are necessary to check the generalizability of the results. ML technologies are expected to advance significantly in the coming years, contributing to the diagnosis of ASD and helping clinicians soon.

Identifying Visual Attention Features Accurately Discerning Between Autism and Typically Developing: a Deep Learning Framework

Atypical visual attention is a hallmark of autism spectrum disorder (ASD). Identifying the attention features accurately discerning between people with ASD and typically developing (TD) at the individual level remains a challenge. In this study, we developed a new systematic framework combining high accuracy deep learning classification, deep learning segmentation, image ablation and a direct measurement of classification ability to identify the discriminative features for autism identification. Our two-stream model achieved the state-of-the-art performance with a classification accuracy of 0.95. Using this framework, two new categories of features, Food & drink and Outdoor-objects, were identified as discriminative attention features, in addition to the previously reported features including Center-object and Human-faces, etc. Altered attention to the new categories helps to understand related atypical behaviors in ASD. Importantly, the area under curve (AUC) based on the combined top-9 features identified in this study was 0.92, allowing an accurate classification at the individual level. We also obtained a small but informative dataset of 12 images with an AUC of 0.86, suggesting a potentially efficient approach for the clinical diagnosis of ASD. Together, our deep learning framework based on VGG-16 provides a novel and powerful tool to recognize and understand abnormal visual attention in ASD, which will, in turn, facilitate the identification of biomarkers for ASD.

Multi-Source Domain Adaptation Techniques for Mitigating Batch Effects: A Comparative Study

The past decade has seen an increasing number of applications of deep learning (DL) techniques to biomedical fields, especially in neuroimaging-based analysis. Such DL-based methods are generally data-intensive and require a large number of training instances, which might be infeasible to acquire from a single acquisition site, especially for data, such as fMRI scans, due to the time and costs that they demand. We can attempt to address this issue by combining fMRI data from various sites, thereby creating a bigger heterogeneous dataset. Unfortunately, the inherent differences in the combined data, known as batch effects, often hamper learning a model. To mitigate this issue, techniques such as multi-source domain adaptation [Multi-source Domain Adversarial Networks (MSDA)] aim at learning an effective classification function that uses (learned) domain-invariant latent features. This article analyzes and compares the performance of various popular MSDA methods [MDAN, Domain AggRegation Networks (DARN), Multi-Domain Matching Networks (MDMN), and Moment Matching for MSDA (M3SDA)] at predicting different labels (illness, age, and sex) of images from two public rs-fMRI datasets: ABIDE 1and ADHD-200. It also evaluates the impact of various conditions such as class imbalance, the number of sites along with a comparison of the degree of adaptation of each of the methods, thereby presenting the effectiveness of MSDA models in neuroimaging-based applications.

rs-fMRI and machine learning for ASD diagnosis: a systematic review and meta-analysis

Autism Spectrum Disorder (ASD) diagnosis is still based on behavioral criteria through a lengthy and time-consuming process. Much effort is being made to identify brain imaging biomarkers and develop tools that could facilitate its diagnosis. In particular, using Machine Learning classifiers based on resting-state fMRI (rs-fMRI) data is promising, but there is an ongoing need for further research on their accuracy and reliability. Therefore, we conducted a systematic review and meta-analysis to summarize the available evidence in the literature so far. A bivariate random-effects meta-analytic model was implemented to investigate the sensitivity and specificity across the 55 studies that offered sufficient information for quantitative analysis. Our results indicated overall summary sensitivity and specificity estimates of 73.8% and 74.8%, respectively. SVM stood out as the most used classifier, presenting summary estimates above 76%. Studies with bigger samples tended to obtain worse accuracies, except in the subgroup analysis for ANN classifiers. The use of other brain imaging or phenotypic data to complement rs-fMRI information seems promising, achieving higher sensitivities when compared to rs-fMRI data alone (84.7% versus 72.8%). Finally, our analysis showed AUC values between acceptable and excellent. Still, given the many limitations indicated in our study, further well-designed studies are warranted to extend the potential use of those classification algorithms to clinical settings.

Multi-level Functional Connectivity Fusion Classification Framework for Brain Disease Diagnosis

Brain disease diagnosis is a new hotspot in the cross research of artificial intelligence and neuroscience. Quantitative analysis of functional magnetic resonance imaging (fMRI) data can provide valuable biomarkers that contributes to clinical diagnosis, and the analysis of functional connectivity (FC) has become the primary method. However, previous studies mainly focus on brain disease classification based on the low-order FC features, ignoring the potential role of high-order functional relationships among brain regions. To solve this problem, this study proposed a novel multi-level FC fusion classification framework (MFC) for brain disease diagnosis. We firstly designed a deep neural network (DNN) model to extract and learn abstract feature representations for the constructed low-order and high-order FC patterns. Both unsupervised and supervised learning steps were performed during the DNN model training, and the prototype learning was introduced in the supervised fine-tuning to improve the intra-class compactness and inter-class separability of the feature representation. Then, we combined the learned multi-level abstract FC features and trained an ensemble classifier with a hierarchical stacking learning strategy for the brain disease classification. Systematic experiments were conducted on two real large-scale fMRI datasets. Results showed that the proposed MFC model obtained robust classification performance for different preprocessing pipelines, different brain parcellations, and different cross-validation schemes, suggesting the effectiveness and generality of the proposed MFC model. Overall, this study provides a promising solution to combine the informative low-order and high-order FC patterns to further promote the classification of brain diseases.

ConCeptCNN: A novel multi-filter convolutional neural network for the prediction of neurodevelopmental disorders using brain connectome

BACKGROUND

Deep convolutional neural network (CNN) and its derivatives have recently shown great promise in the prediction of brain disorders using brain connectome data. Existing deep CNN methods using single global row and column convolutional filters have limited ability to extract discriminative information from brain connectome for prediction tasks.

PURPOSE

This paper presents a novel deep Connectome-Inception CNN (ConCeptCNN) model, which is developed based on multiple convolutional filters. The proposed model is used to extract topological features from brain connectome data for neurological disorders classification and analysis.

METHODS

The ConCeptCNN uses multiple vector-shaped filters extract topological information from the brain connectome at different levels for complementary feature embeddings of brain connectome. The proposed model is validated using two datasets: the Neuro Bureau ADHD-200 dataset and the Cincinnati Early Prediction Study (CINEPS) dataset.

RESULTS

In a cross-validation experiment, the ConCeptCNN achieved a prediction accuracy of 78.7% for the detection of ADHD in adolescents and an accuracy of 81.6% for the prediction of cognitive deficits at 2-year corrected age in very preterm infants. In addition to the classification tasks, the ConCeptCNN identified several brain regions that are discriminative to neurodevelopmental disorders.

CONCLUSIONS

We compared the ConCeptCNN with several peer CNN methods. The results demonstrated that proposed model improves overall classification performance of neurodevelopmental disorders prediction tasks. This article is protected by copyright. All rights reserved.

Domain adaptation based on rough adjoint inconsistency and optimal transport for identifying autistic patients

BACKGROUND AND OBJECTIVE

Computer aided diagnosis technology has been widely used to diagnose autism spectrum disorder (ASD) from neural images. The performance of the model usually depends largely on a sufficient number of training samples that reflect the real sample distribution. Due to the lack of labelled neural images data, multisite data are often pooled together to expand the sample size. However, the heterogeneity among sites will inevitably lead to a decline in the generalization of models. To solve this problem, we propose a multisource unsupervised domain adaptation method using rough adjoint inconsistency and optimal transport.

METHODS

First, we define the concept of rough adjoint inconsistency and propose a double quantization method based on rough adjoint inconsistency and Dempster-Shafer (D-S) evidence theory to estimate the weight coefficient of each source domain to accurately describe the importance of each source domain to the target domain. Second, using optimal transport theory, we weaken the data distribution differences between domains and solve the problem of class imbalance by adjusting the sampling weights among classes.

RESULTS

The ASD recognition accuracy of the proposed method is improved on all eight tasks, which are 70.67%, 64.86%, 62.50%, 70.80%, 73.08%, 71.19%, 75.41% and 75.76%, respectively. Our proposed model achieves superior performance compared to traditional machine learning methods and other recently proposed deep learning model.

CONCLUSIONS

Our method demonstrates that the fusion of rough adjoint inconsistency and optimal transport can be a powerful tool for identifying ASD and quantifying the correlations between domains.

A multi-sequences MRI deep framework study applied to glioma classfication

Glioma is one of the most important central nervous system tumors, ranked 15th in the most common cancer for men and women. Magnetic Resonance Imaging (MRI) represents a common tool for medical experts to the diagnosis of glioma. A set of multi-sequences from an MRI is selected according to the severity of the pathology. Our proposed approach aims moreto create a computer-aided system that is capable of helping morethe expert diagnose the brain gliomas. moreWe propose a supervised learning regime based on a convolutional neural network based framework and transfer learning techniques. Our research morefocuses on the performance of different pre-trained deep learning models with respect to different MRI sequences. We highlight the best combinations of such model-MRI sequence couple for our specific task of classifying healthy brain against brain with glioma. moreWe also propose to visually analyze the extracted deep features for studying the existing relation of the MRI sequences and models. This interpretability analysis gives some hints for medical expert to understand the diagnosis made by the models. Our study is based on the well-known BraTS datasets including multi-sequence images and expert diagnosis.

Reproducible neuroimaging features for diagnosis of autism spectrum disorder with machine learning

Autism spectrum disorder (ASD) is the fourth most common neurodevelopmental disorder, with a prevalence of 1 in 160 children. Accurate diagnosis relies on experts, but such individuals are scarce. This has led to increasing interest in the development of machine learning (ML) models that can integrate neuroimaging features from functional and structural MRI (fMRI and sMRI) to help reveal central nervous system alterations characteristic of ASD. We optimized and compared the performance of 12 of the most popular and powerful ML models. Each was separately trained using 15 different combinations of fMRI and sMRI features and optimized with an unbiased model search. Deep learning models predicted ASD with the highest diagnostic accuracy and generalized well to other MRI datasets. Our model achieves state-of-the-art 80% area under the ROC curve (AUROC) in diagnosis on test data from the IMPAC dataset; and 86% and 79% AUROC on the external ABIDE I and ABIDE II datasets (with further improvement to 93% and 90% after supervised domain adaptation). The highest performing models identified reproducible putative biomarkers for accurate ASD diagnosis in accord with known ASD markers as well as novel cerebellar biomarkers. Such reproducibility lends credence to their tremendous potential for defining and using a set of truly generalizable ASD biomarkers that will advance scientific understanding of neuronal changes in ASD.

Transfer Learning Approaches for Neuroimaging Analysis: A Scoping Review

Deep learning algorithms have been moderately successful in diagnoses of diseases by analyzing medical images especially through neuroimaging that is rich in annotated data. Transfer learning methods have demonstrated strong performance in tackling annotated data. It utilizes and transfers knowledge learned from a source domain to target domain even when the dataset is small. There are multiple approaches to transfer learning that result in a range of performance estimates in diagnosis, detection, and classification of clinical problems. Therefore, in this paper, we reviewed transfer learning approaches, their design attributes, and their applications to neuroimaging problems. We reviewed two main literature databases and included the most relevant studies using predefined inclusion criteria. Among 50 reviewed studies, more than half of them are on transfer learning for Alzheimer's disease. Brain mapping and brain tumor detection were second and third most discussed research problems, respectively. The most common source dataset for transfer learning was ImageNet, which is not a neuroimaging dataset. This suggests that the majority of studies preferred pre-trained models instead of training their own model on a neuroimaging dataset. Although, about one third of studies designed their own architecture, most studies used existing Convolutional Neural Network architectures. Magnetic Resonance Imaging was the most common imaging modality. In almost all studies, transfer learning contributed to better performance in diagnosis, classification, segmentation of different neuroimaging diseases and problems, than methods without transfer learning. Among different transfer learning approaches, fine-tuning all convolutional and fully-connected layers approach and freezing convolutional layers and fine-tuning fully-connected layers approach demonstrated superior performance in terms of accuracy. These recent transfer learning approaches not only show great performance but also require less computational resources and time.

Functional Connectivity-Based Prediction of Autism on Site Harmonized ABIDE Dataset

OBJECTIVE

The larger sample sizes available from multi-site publicly available neuroimaging data repositories makes machine-learning based diagnostic classification of mental disorders more feasible by alleviating the curse of dimensionality. However, since multi-site data are aggregated post-hoc, i.e. they were acquired from different scanners with different acquisition parameters, non-neural inter-site variability may mask inter-group differences that are at least in part neural in origin. Hence, the advantages gained by the larger sample size in the context of machine-learning based diagnostic classification may not be realized.

METHODS

We address this issue using harmonization of multi-site neuroimaging data using the ComBat technique, which is based on an empirical Bayes formulation to remove inter-site differences in data distributions, to improve diagnostic classification accuracy. Specifically, we demonstrate this using ABIDE (Autism Brain Imaging Data Exchange) multi-site data for classifying individuals with Autism from healthy controls using resting state fMRI-based functional connectivity data.

RESULTS

Our results show that higher classification accuracies across multiple classification models can be obtained (especially for models based on artificial neural networks) from multi-site data post harmonization with the ComBat technique as compared to without harmonization, outperforming earlier results from existing studies using ABIDE. Furthermore, our network ablation analysis facilitated important insights into autism spectrum disorder pathology and the connectivity in networks shown to be important for classification covaried with verbal communication impairments in Autism.

CONCLUSION

Multi-site data harmonization using ComBat improves neuroimaging-based diagnostic classification of mental disorders.

SIGNIFICANCE

ComBat has the potential to make AI-based clinical decision-support systems more feasible in psychiatry.

Multi-site clustering and nested feature extraction for identifying autism spectrum disorder with resting-state fMRI

Brain functional connectivity (FC) derived from resting-state functional magnetic resonance imaging (rs-fMRI) has been widely employed to study neuropsychiatric disorders such as autism spectrum disorder (ASD). Existing studies usually suffer from (1) significant data heterogeneity caused by different scanners or studied populations in multiple sites, (2) curse of dimensionality caused by millions of voxels in each fMRI scan and a very limited number (tens or hundreds) of training samples, and (3) poor interpretability, which hinders the identification of reproducible disease biomarkers. To this end, we propose a Multi-site Clustering and Nested Feature Extraction (MC-NFE) method for fMRI-based ASD detection. Specifically, we first divide multi-site training data into ASD and healthy control (HC) groups. To model inter-site heterogeneity within each category, we use a similarity-driven multiview linear reconstruction model to learn latent representations and perform subject clustering within each group. We then design a nested singular value decomposition (SVD) method to mitigate inter-site heterogeneity and extract FC features by learning both local cluster-shared features across sites within each category and global category-shared features across ASD and HC groups, followed by a linear support vector machine (SVM) for ASD detection. Experimental results on 609 subjects with rs-fMRI from the ABIDE database with 21 imaging sites suggest that the proposed MC-NFE outperforms several state-of-the-art methods in ASD detection. The most discriminative FCs identified by the MC-NFE are mainly located in default mode network, salience network, and cerebellum region, which could be used as potential biomarkers for fMRI-based ASD analysis.

Convolutional Neural Network With Graphical Lasso to Extract Sparse Topological Features for Brain Disease Classification

The functional connectivity provides new insights into the mechanisms of the human brain at network-level, which has been proved to be an effective biomarker for brain disease classification. Recently, machine learning methods have played an important role in functional connectivity classification, among which convolutional neural network (CNN) based methods become a new hot topic since they can extract topological features in the brain network. However, the conventional CNN-based methods haven't taken sparse connectivity patterns (SCPs) of the human brain into consideration, which may lead to redundancy of the topological features, and limit their performance and generalization. To solve it, we propose a novel CNN-based model with graphical Lasso (CNNGLasso) to extract sparse topological features for brain disease classification. First, we develop a novel graphical Lasso model for revealing the SCPs at group-level. Then, the SCPs are used to guide the topological feature extraction. Finally, the obtained sparse topological features are used to classify the patients from normal controls. The experiment results on the ABIDE dataset demonstrate that the CNNGLasso outperforms the others on various performances. Besides, the abnormal brain regions derived from the trained model are consistent with the previous investigations, which further proves the application prospect of the CNNGLasso.

A Variational Encoder Framework for Decoding Behavior Choices from Neural Data

In this paper, using an adversarial variational encoder model, we propose a two-step data-driven approach to extract cross-subject feature representations from neural activity in order to decode subjects' behavior choices. First, various characteristics of the recorded behavior are computed and passed as features to a clustering model in order to categorize different behavior choices in each trial and create labels for the data. Then, we utilize a variational encoder to learn the latent space mappings from neural activity. An attached adversary network is used in a discriminative setting to detach the subject's individuality from the representations. Recorded cortical activity from Thy1-GCaMP6s transgenic mice during a motivational licking experiment was used in this study. Experimental results demonstrate the capabilities of the proposed method in extracting discriminative representations from neural data to decode behavior by achieving an average classification accuracy of 88.8% across subjects.

Deep learning for neuroimaging-based diagnosis and rehabilitation of Autism Spectrum Disorder: A review

Accurate diagnosis of Autism Spectrum Disorder (ASD) followed by effective rehabilitation is essential for the management of this disorder. Artificial intelligence (AI) techniques can aid physicians to apply automatic diagnosis and rehabilitation procedures. AI techniques comprise traditional machine learning (ML) approaches and deep learning (DL) techniques. Conventional ML methods employ various feature extraction and classification techniques, but in DL, the process of feature extraction and classification is accomplished intelligently and integrally. DL methods for diagnosis of ASD have been focused on neuroimaging-based approaches. Neuroimaging techniques are non-invasive disease markers potentially useful for ASD diagnosis. Structural and functional neuroimaging techniques provide physicians substantial information about the structure (anatomy and structural connectivity) and function (activity and functional connectivity) of the brain. Due to the intricate structure and function of the brain, proposing optimum procedures for ASD diagnosis with neuroimaging data without exploiting powerful AI techniques like DL may be challenging. In this paper, studies conducted with the aid of DL networks to distinguish ASD are investigated. Rehabilitation tools provided for supporting ASD patients utilizing DL networks are also assessed. Finally, we will present important challenges in the automated detection and rehabilitation of ASD and propose some future works.

Autism Spectrum Disorder Studies Using fMRI Data and Machine Learning: A Review

Machine learning methods have been frequently applied in the field of cognitive neuroscience in the last decade. A great deal of attention has been attracted to introduce machine learning methods to study the autism spectrum disorder (ASD) in order to find out its neurophysiological underpinnings. In this paper, we presented a comprehensive review about the previous studies since 2011, which applied machine learning methods to analyze the functional magnetic resonance imaging (fMRI) data of autistic individuals and the typical controls (TCs). The all-round process was covered, including feature construction from raw fMRI data, feature selection methods, machine learning methods, factors for high classification accuracy, and critical conclusions. Applying different machine learning methods and fMRI data acquired from different sites, classification accuracies were obtained ranging from 48.3% up to 97%, and informative brain regions and networks were located. Through thorough analysis, high classification accuracies were found to usually occur in the studies which involved task-based fMRI data, single dataset for some selection principle, effective feature selection methods, or advanced machine learning methods. Advanced deep learning together with the multi-site Autism Brain Imaging Data Exchange (ABIDE) dataset became research trends especially in the recent 4 years. In the future, advanced feature selection and machine learning methods combined with multi-site dataset or easily operated task-based fMRI data may appear to have the potentiality to serve as a promising diagnostic tool for ASD.

Machine learning with neuroimaging data to identify autism spectrum disorder: a systematic review and meta-analysis

PURPOSE

Autism Spectrum Disorder (ASD) is diagnosed through observation or interview assessments, which is time-consuming, subjective, and with questionable validity and reliability. Thus, we aimed to evaluate the role of machine learning (ML) with neuroimaging data to provide a reliable classification of ASD.

METHODS

A systematic search of PubMed, Scopus, and Embase was conducted to identify relevant publications. Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) was used to assess the studies' quality. A bivariate random-effects model meta-analysis was employed to evaluate the pooled sensitivity, the pooled specificity, and the diagnostic performance through the hierarchical summary receiver operating characteristic (HSROC) curve of ML with neuroimaging data in classifying ASD. Meta-regression was also performed.

RESULTS

Forty-four studies (5697 ASD and 6013 typically developing individuals [TD] in total) were included in the quantitative analysis. The pooled sensitivity for differentiating ASD from TD individuals was 86.25 95% confidence interval [CI] (81.24, 90.08), while the pooled specificity was 83.31 95% CI (78.12, 87.48) with a combined area under the HSROC (AUC) of 0.889. Higgins I² (> 90%) and Cochran's Q (p < 0.0001) suggest a high degree of heterogeneity. In the bivariate model meta-regression, a higher pooled specificity was observed in studies not using a brain atlas (90.91 95% CI [80.67, 96.00], p = 0.032). In addition, a greater pooled sensitivity was seen in studies recruiting both males and females (89.04 95% CI [83.84, 92.72], p = 0.021), and combining imaging modalities (94.12 95% [85.43, 97.76], p = 0.036).

CONCLUSION

ML with neuroimaging data is an exciting prospect in detecting individuals with ASD but further studies are required to improve its reliability for usage in clinical practice.

Application of Artificial Intelligence in the MRI Classification Task of Human Brain Neurological and Psychiatric Diseases: A Scoping Review

Artificial intelligence (AI) for medical imaging is a technology with great potential. An in-depth understanding of the principles and applications of magnetic resonance imaging (MRI), machine learning (ML), and deep learning (DL) is fundamental for developing AI-based algorithms that can meet the requirements of clinical diagnosis and have excellent quality and efficiency. Moreover, a more comprehensive understanding of applications and opportunities would help to implement AI-based methods in an ethical and sustainable manner. This review first summarizes recent research advances in ML and DL techniques for classifying human brain magnetic resonance images. Then, the application of ML and DL methods to six typical neurological and psychiatric diseases is summarized, including Alzheimer's disease (AD), Parkinson's disease (PD), major depressive disorder (MDD), schizophrenia (SCZ), attention-deficit/hyperactivity disorder (ADHD), and autism spectrum disorder (ASD). Finally, the limitations of the existing research are discussed, and possible future research directions are proposed.

Detection of Autism Spectrum Disorder in Children Using Machine Learning Techniques

Autism Spectrum Disorder (ASD) is a neurological disorder which might have a lifelong impact on the language learning, speech, cognitive, and social skills of an individual. Its symptoms usually show up in the developmental stages, i.e., within the first two years after birth, and it impacts around 1% of the population globally [https://www.autism-society.org/whatis/facts-and-statistics/. Accessed 25 Dec 2019]. ASD is mainly caused by genetics or by environmental factors; however, its conditions can be improved by detecting and treating it at earlier stages. In the current times, clinical standardized tests are the only methods which are being used, to diagnose ASD. This not only requires prolonged diagnostic time but also faces a steep increase in medical costs. To improve the precision and time required for diagnosis, machine learning techniques are being used to complement the conventional methods. We have applied models such as Support Vector Machines (SVM), Random Forest Classifier (RFC), Naïve Bayes (NB), Logistic Regression (LR), and KNN to our dataset and constructed predictive models based on the outcome. The main objective of our paper is to thus determine if the child is susceptible to ASD in its nascent stages, which would help streamline the diagnosis process. Based on our results, Logistic Regression gives the highest accuracy for our selected dataset.

ASD-SAENet: A Sparse Autoencoder, and Deep-Neural Network Model for Detecting Autism Spectrum Disorder (ASD) Using fMRI Data

Autism spectrum disorder (ASD) is a heterogenous neurodevelopmental disorder which is characterized by impaired communication, and limited social interactions. The shortcomings of current clinical approaches which are based exclusively on behavioral observation of symptomology, and poor understanding of the neurological mechanisms underlying ASD necessitates the identification of new biomarkers that can aid in study of brain development, and functioning, and can lead to accurate and early detection of ASD. In this paper, we developed a deep-learning model called ASD-SAENet for classifying patients with ASD from typical control subjects using fMRI data. We designed and implemented a sparse autoencoder (SAE) which results in optimized extraction of features that can be used for classification. These features are then fed into a deep neural network (DNN) which results in superior classification of fMRI brain scans more prone to ASD. Our proposed model is trained to optimize the classifier while improving extracted features based on both reconstructed data error and the classifier error. We evaluated our proposed deep-learning model using publicly available Autism Brain Imaging Data Exchange (ABIDE) dataset collected from 17 different research centers, and include more than 1,035 subjects. Our extensive experimentation demonstrate that ASD-SAENet exhibits comparable accuracy (70.8%), and superior specificity (79.1%) for the whole dataset as compared to other methods. Further, our experiments demonstrate superior results as compared to other state-of-the-art methods on 12 out of the 17 imaging centers exhibiting superior generalizability across different data acquisition sites and protocols. The implemented code is available on GitHub portal of our lab at: https://github.com/pcdslab/ASD-SAENet.