Neural Information Processing Research Articles

Region-Related Differences in Short-Term Synaptic Plasticity and Synaptotagmin-7 in the Male and Female Hippocampus of a Rat Model of Fragile X Syndrome

Fragile X syndrome (FXS) is an intellectual developmental disorder characterized, inter alia, by deficits in the short-term processing of neural information, such as sensory processing and working memory. The primary cause of FXS is the loss of fragile X messenger ribonucleoprotein (FMRP), which is profoundly involved in synaptic function and plasticity. Short-term synaptic plasticity (STSP) may play important roles in functions that are affected by FXS. Recent evidence points to the crucial involvement of the presynaptic calcium sensor synaptotagmin-7 (Syt-7) in STSP. However, how the loss of FMRP affects STSP and Syt-7 have been insufficiently studied. Furthermore, males and females are affected differently by FXS, but the underlying mechanisms remain elusive. The aim of the present study was to investigate possible changes in STSP and the expression of Syt-7 in the dorsal (DH) and ventral (VH) hippocampus of adult males and females in a Fmr1-knockout (KO) rat model of FXS. We found that the paired-pulse ratio (PPR) and frequency facilitation/depression (FF/D), two forms of STSP, as well as the expression of Syt-7, are normal in adult KO males, but the PPR is increased in the ventral hippocampus of KO females (6.4 ± 3.7 vs. 18.3 ± 4.2 at 25 ms in wild type (WT) and KO, respectively). Furthermore, we found no gender-related differences, but did find robust region-dependent difference in the STSP (e.g., the PPR at 50 ms: 50.0 ± 5.5 vs. 17.6 ± 2.9 in DH and VH of WT male rats; 53.1 ± 3.6 vs. 19.3 ± 4.6 in DH and VH of WT female rats; 48.1 ± 2.3 vs. 19.1 ± 3.3 in DH and VH of KO male rats; and 51.2 ± 3.3 vs. 24.7 ± 4.3 in DH and VH of KO female rats). AMPA receptors are similarly expressed in the two hippocampal segments of the two genotypes and in both genders. Also, basal excitatory synaptic transmission is higher in males compared to females. Interestingly, we found more than a twofold higher level of Syt-7, not synaptotagmin-1, in the dorsal compared to the ventral hippocampus in the males of both genotypes (0.43 ± 0.1 vs. 0.16 ± 0.02 in DH and VH of WT male rats, and 0.6 ± 0.13 vs. 0.23 ± 0.04 in DH and VH of KO male rats) and in the WT females (0.97 ± 0.23 vs. 0.31 ± 0.09 in DH and VH). These results point to the susceptibility of the female ventral hippocampus to FMRP loss. Importantly, the different levels of Syt-7, which parallel the higher score of the dorsal vs. ventral hippocampus on synaptic facilitation, suggest that Syt-7 may play a pivotal role in defining the striking differences in STSP along the long axis of the hippocampus.

Transcranial ultrasound stimulation effect in the redundant and synergistic networks consistent across macaques

Abstract Low-intensity transcranial ultrasound stimulation (TUS) is a non-invasive technique that safely alters neural activity, reaching deep brain areas with good spatial accuracy. We investigated the effects of TUS in macaques using a recent metric, the synergy minus redundancy rank gradient, that quantifies different kinds of neural information processing. We analyzed this high-order quantity on the fMRI data after TUS in two targets: the supplementary motor area (SMA-TUS) and the frontal polar cortex (FPC-TUS). The TUS produced specific changes at the limbic network at FPC-TUS and the motor network at SMA-TUS and altered the sensorimotor, temporal, and frontal networks in both targets, mostly consistent across macaques. Moreover, there was a reduction in the structural and functional coupling after both stimulations. Finally, the TUS changed the intrinsic high-order network topology, decreasing the modular organization of the redundancy at SMA-TUS and increasing the synergistic integration at FPC-TUS.

Comparing expert systems and their explainability through similarity

In our work, we propose the use of Representational Similarity Analysis (RSA) for explainable AI (XAI) approaches to enhance the reliability of XAI-based decision support systems. To demonstrate how similarity analysis of explanations can assess the output stability of post-hoc explainers, we conducted a computational evaluative study. This study addresses how our approach can be leveraged to analyze the stability of explanations amidst various changes in the ML pipeline. Our results show that modifications such as altered preprocessing or different ML models lead to changes in the explanations and illustrate the extent to which stability can suffer. Explanation similarity analysis enables practitioners to compare different explanation outcomes, thus monitoring stability in explanations. Alongside discussing the results and practical applications in operationalized ML, including both benefits and limitations, we also delve into insights from computational neuroscience and neural information processing.

Thermodynamic limits on neural information processing efficiency

Thermodynamic limits on neural information processing efficiency

Timing of acute cold exposure determines thermoregulatory responses - possible interactions between the thermal environment and circadian clocks

Introduction: Shift workers are frequently exposed to the harsh combination of night work in a low-temperature environment. In this project, we aimed to determine if acute induction of thermoregulatory markers UCP1 and FGF21 by cold exposure depend on the timing of the exposure, and to investigate if the neural processing of thermal information has a potential role. Methods: Six groups of C57BI/J6 mice (N=8) were habituated for 2 months to a 12 h light/ 12 h dark regiment at thermoneutral temperature 27 °C (TN). After this period, groups were acutely exposed to an ambient temperature of 15 °C (COLD) or a continuation of TN temperature. Temperature exposures of groups were applied at three different time-points of the day lasting for 8 hours before tissues were harvested at ZT4, ZT12 or ZT20 (i.e. 4, 12 or 20 hours after the light is on, respectively). Thermoregulatory markers were measured in plasma, brown adipose tissue (BAT), liver and the peripheral somatosensory neurons in dorsal root ganglion (DRG) using RTPCR and immunodetection. Furthermore, behaviour testing was used to evaluate thermal preference. Results: UCP1 expression increased in BAT from COLD mice, as compared to TN mice. A higher induction of UCP1 was observed at ZT12 than the other time points. Plasma levels of FGF21 were induced acutely in COLD mice at ZT4 and ZT12, compared to the time-matched TN controls, but COLD mice at ZT20 did not display increase of FGF21 plasma levels. Cold exposure affected mRNA expression patterns of core circadian clock components in BAT (Bmal1, Clock, Per1/2/3, Cry1/2, Nr1d1/2), but in liver these transcripts were not affected. Behavior testing showed that COLD mice developed cold tolerance behaviors at ZT4 but less at ZT12. Analyses of gene expression in the somatosensory DRG neurons revealed circadian oscillations of core components and gene regulation due to cold exposure. Summary: In a nutshell, data suggest that the timing of cold exposure might impact physiological and behavioural adaptations to cold environments. This was associated with diurnal variations of gene expression in DRG neurons. Supported with an internal grant at STAMI. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

Prediction error processing and sharpening of expected information across the face-processing hierarchy

The perception and neural processing of sensory information are strongly influenced by prior expectations. The integration of prior and sensory information can manifest through distinct underlying mechanisms: focusing on unexpected input, denoted as prediction error (PE) processing, or amplifying anticipated information via sharpened representation. In this study, we employed computational modeling using deep neural networks combined with representational similarity analyses of fMRI data to investigate these two processes during face perception. Participants were cued to see face images, some generated by morphing two faces, leading to ambiguity in face identity. We show that expected faces were identified faster and perception of ambiguous faces was shifted towards priors. Multivariate analyses uncovered evidence for PE processing across and beyond the face-processing hierarchy from the occipital face area (OFA), via the fusiform face area, to the anterior temporal lobe, and suggest sharpened representations in the OFA. Our findings support the proposition that the brain represents faces grounded in prior expectations.

Iterative kernel regression with preconditioning

Kernel methods are popular in nonlinear and nonparametric regression due to their solid mathematical foundations and optimal statistical properties. However, scalability remains the primary bottleneck in applying kernel methods to large-scale data regression analysis. This paper aims to improve the scalability of kernel methods. We combine Nyström subsampling and the preconditioned conjugate gradient method to solve regularized kernel regression. Our theoretical analysis indicates that achieving optimal convergence rates requires only [Formula: see text] memory and [Formula: see text] time (up to logarithmic factors). Numerical experiments show that our algorithm outperforms existing methods in time efficiency and prediction accuracy on large-scale datasets. Notably, compared to the FALKON algorithm [A. Rudi, L. Carratino and L. Rosasco, Falkon: An optimal large scale kernel method, in Advances in Neural Information Processing Systems (Curran Associates, 2017), pp. 3891–3901], which is known as the optimal large-scale kernel method, our method is more flexible (applicable to non-positive definite kernel functions) and has a lower algorithmic complexity. Additionally, our established theoretical analysis further relaxes the restrictive conditions on hyperparameters previously imposed in convergence analyses.

A Synapse-Threshold Synergistic Learning Approach for Spiking Neural Networks

Spiking neural networks (SNNs) have demonstrated excellent capabilities in various intelligent scenarios. Most existing methods for training SNNs are based on the concept of synaptic plasticity; however, learning in the realistic brain also utilizes intrinsic non-synaptic mechanisms of neurons. The spike threshold of biological neurons is a critical intrinsic neuronal feature that exhibits rich dynamics on a millisecond timescale and has been proposed as an underlying mechanism that facilitates neural information processing. In this study, we develop a novel synergistic learning approach that involves simultaneously training synaptic weights and spike thresholds in SNNs. SNNs trained with synapse-threshold synergistic learning (STL-SNNs) achieve significantly superior performance on various static and neuromorphic datasets than SNNs trained with two degenerated single-learning models. During training, the synergistic learning approach optimizes neural thresholds, providing the network with stable signal transmission via appropriate firing rates. Further analysis indicates that STL-SNNs are robust to noisy data and exhibit low energy consumption for deep network structures. Additionally, the performance of STL-SNN can be further improved by introducing a generalized joint decision framework. Overall, our findings indicate that biologically plausible synergies between synaptic and intrinsic non-synaptic mechanisms may provide a promising approach for developing highly efficient SNN learning methods.

Newman-Watts-Strogatz topology in deep echo state networks for speech emotion recognition

Cortical neural connectivity exhibits a small-world (SW) network topology. However, its role in neural information processing remains unclear. In this study, we investigate the learning performance of a Deep Echo State Network (DeepESN) that incorporates the Newman-Watts-Strogatz (NWS) graph topology as a reservoir. The short path length of the SW topology allows input information to reach output nodes faster, and the clustered arrangement of this topology can expand the range of the echo state property. Additionally, in this paper, we seek to identify the best set of these key factors, such as the spectral radius of the reservoir’s weight matrix and the hyperparameters neighbors' and probabilities. We evaluate the NWS-DeepESNs on four Speech Emotion Recognition (SER) tasks, which are RAVDESS, Emo-db, SAVEE, and TESS databases. By examining the NWS-DeepESN’s performance on SER tasks, we aim to determine the role of the NWS graph topology in enhancing neural information processing. The obtained results exhibited significant improvements in recognition accuracies (weighted and unweighted) with (76.45%, 75.79%), (87.89%, 87.14%), (99%, 98.62%), and (63%, 62.24%), respectively, for RAVDESS, Emo-db, SAVEE, and TESS databases. The proposed approach improves emotion recognition capabilities and yields more accurate results compared to existing state-of-the-art approaches.

Greater up-modulation of intra-individual brain signal variability makes a high-load cognitive task more arduous for older adults

The extent to which brain responses are less distinctive across varying cognitive loads in older adults is referred to as neural dedifferentiation. Moment-to-moment brain signal variability, an emerging indicator, reveals not only the adaptability of an individual's brain as an inter-individual trait, but also the allocation of neural resources within an individual due to ever-changing task demands, thus shedding novel insight into the process of neural dedifferentiation. However, how the modulation of intra-individual brain signal variability reflects behavioral differences related to cognitively demanding tasks remains unclear. In this study, we employed functional near-infrared spectroscopy (fNIRS) imaging to capture the variability of brain signals, which was quantified by the standard deviation, during both the resting state and an n-back task (n = 1, 2, 3) in 57 healthy older adults. Using multivariate Partial Least Squares (PLS) analysis, we found that fNIRS signal variability increased from the resting state to the task and increased with working memory load in older adults. We further confirmed that greater fNIRS signal variability generally supported faster and more stable response time in the 2- and 3-back conditions. However, the intra-individual level analysis showed that the greater the up-modulation in fNIRS signal variability with cognitive loads, the more its accuracy decreases and mean response time increases, suggesting that a greater intra-individual brain signal variability up-modulation may reflect decreased efficiency in neural information processing. Taken together, our findings offer new insights into the nature of brain signal variability, suggesting that inter- and intra-individual brain signal variability may index distinct theoretical constructs.

Orthogonal neural encoding of targets and distractors supports multivariate cognitive control.

The complex challenges of our mental life require us to coordinate multiple forms of neural information processing. Recent behavioural studies have found that people can coordinate multiple forms of attention, but the underlying neural control process remains obscure. We hypothesized that the brain implements multivariate control by independently monitoring feature-specific difficulty and independently prioritizing feature-specific processing. During functional MRI, participants performed a parametric conflict task that separately tags target and distractor processing. Consistent with feature-specific monitoring, univariate analyses revealed spatially segregated encoding of target and distractor difficulty in the dorsal anterior cingulate cortex. Consistent with feature-specific attentional priority, our encoding geometry analysis revealed overlapping but orthogonal representations of target and distractor coherence in the intraparietal sulcus. Coherence representations were mediated by control demands and aligned with both performance and frontoparietal activity, consistent with top-down attention. Together, these findings provide evidence for the neural geometry necessary to coordinate multivariate cognitive control.

Default mode network aberrance in subjects of alcohol and opioid use disorders during working memory task: An exploratory EEG microstates study

Background: Aberrance in switching from default mode network (DMN) to fronto-parietal network (FPN) is proposed to underlie working memory deficits in subjects with substance use disorders, which can be studied using neuro-imaging techniques during cognitive tasks. The current study used EEG to investigate pre-stimulus microstates during the performance of Sternberg’s working memory task in subjects with substance use disorders. Methods: 128-channel EEG was acquired and processed in ten age and gender-matched subjects, each with alcohol use disorder, opioid use disorder, and controls while they performed Sternberg’s task. Behavioral parameters, pre-stimulus EEG microstate, and underlying sources were analyzed and compared between subjects with substance use disorders and controls. Results: Both alcohol and opioid use disorder subjects had significantly lower accuracy (P < 0.01), while reaction times were significantly higher only in subjects of alcohol use disorder compared to controls (P < 0.01) and opioid use disorder (P < 0.01), reflecting working memory deficits of varying degrees in subjects with substance use disorders. Pre-stimulus EEG microstate revealed four topographic Maps 1-4: subjects of alcohol and opioid use disorder showing significantly lower mean duration of Map 3 (visual processing) and Map 2 (saliency and DMN switching), respectively, compared to controls (P < 0.05). Conclusion: Reduced mean durations in Map 3 and 2 in subjects of alcohol and opioid use disorder can underlie their poorer performance in Sternberg’s task. Furthermore, cortical sources revealed higher activity in both groups of substance use disorders in the parahippocampal gyrus- a hub of DMN; superior and middle temporal gyri associated with impulsivity; and insula that maintains balance between executive reflective system and impulsive system. EEG microstates can be used to envisage neural underpinnings implicated for working memory deficits in subjects of alcohol and opioid use disorders, reflected by aberrant switching between neural networks and information processing mechanisms.

Information Geometry Theoretic Measures for Characterizing Neural Information Processing from Simulated EEG Signals.

In this work, we explore information geometry theoretic measures for characterizing neural information processing from EEG signals simulated by stochastic nonlinear coupled oscillator models for both healthy subjects and Alzheimer's disease (AD) patients with both eyes-closed and eyes-open conditions. In particular, we employ information rates to quantify the time evolution of probability density functions of simulated EEG signals, and employ causal information rates to quantify one signal's instantaneous influence on another signal's information rate. These two measures help us find significant and interesting distinctions between healthy subjects and AD patients when they open or close their eyes. These distinctions may be further related to differences in neural information processing activities of the corresponding brain regions, and to differences in connectivities among these brain regions. Our results show that information rate and causal information rate are superior to their more traditional or established information-theoretic counterparts, i.e., differential entropy and transfer entropy, respectively. Since these novel, information geometry theoretic measures can be applied to experimental EEG signals in a model-free manner, and they are capable of quantifying non-stationary time-varying effects, nonlinearity, and non-Gaussian stochasticity presented in real-world EEG signals, we believe that they can form an important and powerful tool-set for both understanding neural information processing in the brain and the diagnosis of neurological disorders, such as Alzheimer's disease as presented in this work.

The effects of chronic morphine administration on spatial memory and microtubule dynamicity in male mice's brain

The examination of the influence of morphine on behavioral processes, specifically learning and memory, holds significant importance. Additionally, microtubule proteins play a pivotal role in cellular functions, and the dynamics of microtubules contribute to neural network connectivity, information processing, and memory storage. however, the molecular mechanism of morphine on microtubule dynamics, learning, and memory remains uncovered. In the present study, we examined the effects of chronic morphine administration on memory formation impairment and the kinetic alterations in microtubule proteins induced by morphine in mice. Chronic morphine administration at doses of 5 and 10 mg/kg dose-dependently decreased subjects' performance in spatial memory tasks, such as the Morris Water Maze and Y-maze spontaneous alternation behavior. Furthermore, morphine was found to stabilize microtubule structure, and increase polymerization, and total polymer mass. However, it simultaneously impaired microtubule dynamicity, stemming from structural changes in tubulin dimer structure. These findings emphasize the need for careful consideration of different doses when using morphine, urging a more cautious approach in the administration of this opioid medication.

Co-representation of Functional Brain Networks Is Shaped by Cortical Myeloarchitecture and Reveals Individual Behavioral Ability.

Large-scale functional networks are spatially distributed in the human brain. Despite recent progress in differentiating their functional roles, how the brain navigates the spatial coordination among them and the biological relevance of this coordination is still not fully understood. Capitalizing on canonical individualized networks derived from functional MRI data, we proposed a new concept, that is, co-representation of functional brain networks, to delineate the spatial coordination among them. To further quantify the co-representation pattern, we defined two indexes, that is, the co-representation specificity (CoRS) and intensity (CoRI), for separately measuring the extent of specific and average expression of functional networks at each brain location by using the data from both sexes. We found that the identified pattern of co-representation was anchored by cortical regions with three types of cytoarchitectural classes along a sensory-fugal axis, including, at the first end, primary (idiotypic) regions showing high CoRS, at the second end, heteromodal regions showing low CoRS and high CoRI, at the third end, paralimbic regions showing low CoRI. Importantly, we demonstrated the critical role of myeloarchitecture in sculpting the spatial distribution of co-representation by assessing the association with the myelin-related neuroanatomical and transcriptomic profiles. Furthermore, the significance of manifesting the co-representation was revealed in its prediction of individual behavioral ability. Our findings indicated that the spatial coordination among functional networks was built upon an anatomically configured blueprint to facilitate neural information processing, while advancing our understanding of the topographical organization of the brain by emphasizing the assembly of functional networks.

Enhancing Physical Education: A Dynamic Fuzzy Neural Network-Based Information Processing System Design

Enhancing Physical Education: A Dynamic Fuzzy Neural Network-Based Information Processing System Design

Coexisting discharge and synchronization of heterogeneous discrete neural network with crosstalk memristor synapses

Synaptic crosstalk, which occurs due to the overflow of neurotransmitters between neighboring synapses, holds a crucial position in shaping the discharge characteristics and signal transmission within nervous systems. In this paper, two memristors are employed to simulate biological neural synapses and bidirectionally couple Chialvo discrete neuron and Rulkov discrete neuron. Thus, a heterogeneous discrete neural network with memristor-synapse coupling is constructed that takes into account the crosstalk behavior between memristor synapses in the coupled state. The analysis demonstrates that the quantity and stability of fixed points within this neural network intimately depend on the strength of synaptic crosstalk. Additionally, through a thorough investigation of bifurcation diagrams, phase diagrams, Lyapunov exponents, and time sequences, we uncover the multi-stable state property exhibited by the neural network. This property manifests in the coexistence of diverse discharge behaviors, which vary significantly with the synaptic crosstalk strength. Intriguingly, the introduction of control parameter to state variable triggers offset boosting and the emergence of infinite stable states within the neural network. Furthermore, we conducted a comprehensive study to explore the influence of synaptic crosstalk strength on the synchronization behavior of the neural network, considering various coupling strengths, initial conditions, and parameters. Our analysis, which was founded on the phase difference and synchronization factor of neuronal discharge sequences, revealed that the neural network maintains phase synchronization despite the variations of the two crosstalk strengths. The insights gained from this paper provide significant support in elucidating the electrophysiological mechanisms underlying biological neural information processing and transmission. Especially, the coexisting discharge phenomenon in the neural network provides an electrophysiological theoretical foundation for the clinical symptoms and diagnosis of the same neurological disease among different individuals or at different stages. And the doctors can predict the progression and prognosis of neurological disease based on the patterns and characteristics of coexisting discharge in patients, enabling them to adopt appropriate intervention measures and monitoring plans. Therefore, the research on coexisting discharge in the neural system contributes to the comprehensive treatment of nervous system disease.

Molecular Mechanisms of AMPA Receptor Trafficking in the Nervous System.

Synaptic plasticity enhances or reduces connections between neurons, affecting learning and memory. Postsynaptic AMPARs mediate greater than 90% of the rapid excitatory synaptic transmission in glutamatergic neurons. The number and subunit composition of AMPARs are fundamental to synaptic plasticity and the formation of entire neural networks. Accordingly, the insertion and functionalization of AMPARs at the postsynaptic membrane have become a core issue related to neural circuit formation and information processing in the central nervous system. In this review, we summarize current knowledge regarding the related mechanisms of AMPAR expression and trafficking. The proteins related to AMPAR trafficking are discussed in detail, including vesicle-related proteins, cytoskeletal proteins, synaptic proteins, and protein kinases. Furthermore, significant emphasis was placed on the pivotal role of the actin cytoskeleton, which spans throughout the entire transport process in AMPAR transport, indicating that the actin cytoskeleton may serve as a fundamental basis for AMPAR trafficking. Additionally, we summarize the proteases involved in AMPAR post-translational modifications. Moreover, we provide an overview of AMPAR transport and localization to the postsynaptic membrane. Understanding the assembly, trafficking, and dynamic synaptic expression mechanisms of AMPAR may provide valuable insights into the cognitive decline associated with neurodegenerative diseases.

Thalamic functional connectivity and sensorimotor processing in neurodevelopmental disorders.

One of the earliest neurobiological findings in autism has been the differences in the thalamocortical pathway connectivity, suggesting the vital role thalamus plays in human experience. The present functional MRI study investigated resting-state functional connectivity of the thalamus in 49 (autistic, ADHD, and neurotypical) young adults. All participants underwent structural MRI and eyes-open resting state functional MRI scans. After preprocessing the imaging data using Conn's connectivity toolbox, a seed-based functional connectivity analysis was conducted using bilateral thalamus as primary seeds. Autistic participants showed stronger thalamic connectivity, relative to ADHD and neurotypical participants, between the right thalamus and right precentral gyrus, right pars opercularis-BA44, right postcentral gyrus, and the right superior parietal lobule (RSPL). Autistic participants also showed significantly increased connectivity between the left thalamus and the right precentral gyrus. Furthermore, regression analyses revealed a significant relationship between autistic traits and left thalamic-precentral connectivity (R2 = 0.1113), as well as between autistic traits and right postcentral gyrus and RSPL connectivity (R2 = 0.1204) in autistic participants compared to ADHD. These findings provide significant insights into the role of thalamus in coordinating neural information processing and its alterations in neurodevelopmental disorders.

Two‐stage diffusion model deriving FDG‐PET from T1 Weighted Magnetic Resonance Images for diagnosis of Alzheimer’s disease

AbstractBackgroundAlzheimer’s Disease (AD) is a neurodegenerative disease which is characterized by cognitive and functional impairment, and defined by triad of Amyloid status, Tau pathology, and Neuroimaging marker of neurodegeneration or neuronal injury. An early diagnosis of AD can be challenging and it can be detected earlier in brain functional imaging than brain structural imaging.MethodWe develop a pipeline for synthesizing FDG‐PET/MRI from source MRI‐T1‐Weighted (MRI‐T1W) images based on denoising diffusion mechanism [1]. The pipeline includes a forward pass which encodes the source MRI‐T1WI images to gaussian noise space, and a backward pass which generates the corresponding FDG‐PET from the derived noises with repeated denoising steps. Furthermore, to validate the effectiveness of the synthesized FDG‐PET images, we also develop multiple types of classifiers and the experiments for both single input of MRI‐T1WI and joint input of MRI‐T1WI and synthesized FDG‐PET.ResultThe experiments are performed on 1036 of FDG‐PET/MRI pairs (286,474, 276 for Cognitive Normal, Mild Cognitive Impaiement and Alzheimer’s Disease each) are obtained from ADNI [2] datasets and display the promising results with low Mean Absolute Error (MAE) and high Structural Similarity Index Measure (SSIM) for the two stage diffusion denoising based FDG‐PET synthesis pipeline. The results for the additional classification experiments also show that in most cases the joint input enhances the classifiers' performance in accuracy and F1‐score. The results also indicate that the synthesized FDG‐PET images are not only visually similar to the ground truth images, but also improve the performance of machine learning based AD diagnosis.ConclusionThis work provides support for the notion that machine learning‐derived image analysis may be a useful approach to improving the diagnosis of AD.Reference:[1] Ho, J., Jain, A., & Abbeel, P. (2020). Denoising diffusion probabilistic models. Advances in Neural Information Processing Systems, 33, 6840‐6851.[2] Petersen, R., Aisen, P. S., Beckett, L. A., Donohue, M. C., Gamst, A. C., Harvey, D. J., … & Toga, A. W. (2010). Alzheimer’s disease neuroimaging initiative (ADNI). Neurology, 74(3), 201‐209.