MCADNNet: Recognizing Stages of Cognitive Impairment Through Efficient Convolutional fMRI and MRI Neural Network Topology Models

Abstract

Mild cognitive impairment (MCI) represents the intermediate stage between normal cerebral aging and dementia associated with Alzheimer's disease (AD). Early diagnosis of MCI and AD through artificial intelligence has captured considerable scholarly interest; researchers hope to develop therapies capable of slowing or halting these processes. We developed a state-of-the-art deep learning algorithm based on an optimized convolutional neural network (CNN) topology called MCADNNet that simultaneously recognizes MCI, AD, and normally aging brains in adults over the age of 75 years, using structural and functional magnetic resonance imaging (fMRI) data. Following highly detailed preprocessing, fourdimensional (4D) fMRI and 3D MRI were decomposed to create 2D images using a lossless transformation, which enables maximum preservation of data details. The samples were shuffled and subject-level training and testing datasets were completely independent. The optimized MCADNNet was trained and extracted invariant and hierarchical features through convolutional layers followed by multi-classification in the last layer using a softmax layer. A decision-making algorithm was also designed to stabilize the outcome of the trained models. To measure the performance of classification, the accuracy rates for various pipelines were calculated before and after applying the decision-making algorithm. Accuracy rates of 99.77% ± 0.36% and 97.5% ± 1.16% were achieved for MRI and fMRI pipelines, respectively, after applying the decisionmaking algorithm. In conclusion, a cutting-edge and optimized topology called MCADNNet was designed and preceded a preprocessing pipeline; this was followed by a decision-making step that yielded the highest performance achieved for simultaneous classification of the three cohorts examined. INDEX TERMS Deep learning, classification, structural and functional magnetic resonance imaging, brain, Alzheimer's disease, MCI. I. INTRODUCTION A. COGNITIVE IMPAIRMENT Cognitive impairment is a general term referring to impairments in cognition among the domains of memory, learning, The associate editor coordinating the review of this manuscript and approving it for publication was Mohan Venkateshkumar .

The colors of our brain: an integrated approach for dimensionality reduction and explainability in fMRI through color coding (i-ECO)

“…Principal Components Analysis, Autoregression, Linear Embeddings, Autoencoders (Cordes & Nandy, 2006;Huang et al, 2018;Mannfolk et al, 2010;Pereira et al, 2009) being of difficult interpretation for clinicians and hard to generalize. Research in fMRI has been recently characterized by a higher reliance on Neural Networks (Suk et al, 2016) and Embeddings (Sidhu, 2019), with the most promising results coming from CNN (Meszlényi et al, 2017;Sarraf et al, 2019;Tahmassebi et al, 2018;Zhao et al, 2018), especially in the field of Computational Psychiatry (Ariyarathne et al, 2020;El Gazzar et al, 2019;Oh et al, 2019;Silva et al, 2021). Convolutional Neural Networks design follows biological research and the study of the receptive field by the visual cortex (Hubel & Wiesel, 1959), their first development establishing the groundwork for the field of computer vision (Denker et al, 1989;LeCun et al, 1989).…”

Predictive Modelling of Brain Disorders with Magnetic Resonance Imaging: A Systematic Review of Modelling Practices, Transparency, and Interpretability in the use of Convolutional Neural Networks

“…Additionally, relevant spatial information can be lost when modelling in 2D, even if multiple slices are taken as they may not be considered in unison during training. One paper making use of 2D slices provided code [48]. Additionally, ≈ 44% of studies (24/55) made use of multiple models for training and prediction, which in some cases translated to stacking, whereby the output of one trained model is passed to another for training as the input [49,50,51,52,53,54].…”

“…Thirty two out of 55 studies employed repeat experiments through cross validation or other means. A third of studies carrying out repeat experiments (10/32) reported only point estimates for their results, and 5 provided code [57,48,58,55,59]. Of the 14 studies that had both repeat experiments and considerations of interpretability, none detailed whether or not their saliency method was applied per fold or on a hold out test set, and this information was also not detailed where code was provided.…”

Automatic Localization and Discrete Volume Measurements of Hippocampi From MRI Data Using a Convolutional Neural Network

“…Also, a two phase, multimodel automatic brain tumour diagnosis system was developed using CNN in [44]. The ensemble system of a deep convolutional neural network (CNN) and transfer learning-based approach were proposed by a group of researchers in [27], [45], [46] to diagnose Alzheimer's disease from MRI and fMRI scans.…”

“…There are many different MRI analysis and clinical motivations to pursue various methods for segmentation, such as manual delineation, atlas/statistical-atlas-based methods [14]- [18], statistical parametric approach and/or statistical shape models [6], [19], [20], deformable morphometry-based approaches [21], Bayesian approaches [4], [10], patch-based methods [22], [23], machine learning-based approaches, and deep learning-based approaches [3], [8], [9], [24]- [27]. Segmentation of brain regions from MRI scans using any of these approaches does not strictly rely on the intensity information, rather, the intensity distribution of different subfields has a considerably overlapping intensity values.…”