Martin Dyrba

Martin Dyrba
Deutsches Zentrum für Neurodegenerative Erkrankungen | DZNE · RG Clinical Dementia Research - Integrated neuropsychiatric research clinic (Rostock)

Doctor of Medical Informatics
Junior Group Leader: Deep Learning Explainability

About

126
Publications
14,420
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
1,339
Citations
Additional affiliations
April 2011 - present
Deutsches Zentrum für Neurodegenerative Erkrankungen
Position
  • Researcher
September 2006 - September 2010
University of Rostock
Position
  • Student Assistant
Description
  • Programming of - Smart environments infrastructure/middleware; distributed screen sharing - Universal control software and universal remote control for smart meeting room - GPS sensor data annotation tool
Education
October 2005 - March 2011
University of Rostock
Field of study
  • Computer Science

Publications

Publications (126)
Article
Introduction: Amyloid-beta (Aβ) deposition and altered brain structure are the most relevant neuroimaging biomarkers for Alzheimer’s disease (AD). However, their spatial inconsistency was always confusing and misleading. Furthermore, the relationship between this spatial inconsistency and AD progression is unclear. Methods: The current study introd...
Article
Full-text available
Individuals with mild cognitive impairment (MCI) of different subtypes show distinct alterations in network patterns. The first aim of this study is to identify the subtypes of MCI by employing a regional radiomics similarity network (R2SN). The second aim is to characterize the abnormality patterns associated with the clinical manifestations of ea...
Article
Full-text available
Regional Radiomics Similarity Networks Reveal Distinct Subtypes in the Mild Cognitive Impairment In article number 2104538 by Kun Zhao, Yong Liu, Shuyu Li, and co‐workers, two mild cognitive impairment (MCI) subtypes are identified by employing nonnegative matrix factorization on the regional radiomics similarity network (R2SN). The two MCI subtype...
Preprint
Full-text available
Neuroimaging markers based on Magnetic Resonance Imaging (MRI) combined with various other measures (such as informative covariates, vascular risks, brain activity, neuropsychological test etc.,) might provide useful predictions of clinical outcomes during progression towards Alzheimer's disease (AD). The Bayesian approach aims to provide a trade-o...
Article
Full-text available
Purpose: To test whether correcting for unspecific signal from the cerebral white matter increases the sensitivity of amyloid-PET for early stages of cerebral amyloidosis. Methods: We analyzed 18F-Florbetapir-PET and cerebrospinal fluid (CSF) Aβ42 data from 600 older individuals enrolled in the Alzheimer’s Disease Neuroimaging Initiative (ADNI), in...
Article
Background: Inflammation has been described as a key pathogenic event In Alzheimer's disease (AD), downstream of amyloid and tau pathology. Preclinical and clinical data suggest that the cholinergic basal forebrain may moderate inflammatory response to different pathologies. Objective: To study the association of cholinergic basal forebrain volu...
Article
Full-text available
Background: Lipidomics may provide insight into biochemical processes driving Alzheimer's disease (AD) pathogenesis and ensuing clinical trajectories. Objective: To identify a peripheral lipidomics signature associated with AD pathology and investigate its potential to predict clinical progression. Methods: We used Bayesian elastic net regress...
Article
Although deep learning approaches achieve high diagnostic accuracy for detecting Alzheimer’s disease (AD) based on MRI scans, they are rarely applied in clinical research due to a lack of suitable methods for model comprehensibility and interpretability. Recent advances in convolutional neural networks (CNN) visualization algorithms may help to ove...
Conference Paper
In previous research, a data‐driven in vivo staging model of amyloid accumulation was established based on regional frequencies of amyloid positivity in 18F‐florbetapir‐PET scans of cognitively normal older adults. The model was subsequently validated in an independent cohort of participants with subjective memory complaints. In the current study w...
Article
Full-text available
Background Although convolutional neural networks (CNNs) achieve high diagnostic accuracy for detecting Alzheimer’s disease (AD) dementia based on magnetic resonance imaging (MRI) scans, they are not yet applied in clinical routine. One important reason for this is a lack of model comprehensibility. Recently developed visualization methods for deri...
Article
Several observations suggest an impact of prematurity on the claustrum. First, the claustrum’s development appears to depend on transient subplate neurons of intra-uterine brain development, which are affected by prematurity. Second, the claustrum is the most densely connected region of the mammalian forebrain relative to its volume; due to its eff...
Article
Full-text available
Amyotrophic lateral sclerosis 8 (ALS8) is a predominantly lower motor neuron syndrome originally described in a Portuguese-Brazilian family, which originated from a common founder. ALS8 is caused by a VAPB mutation and extremely rare in Central Europe. We present a 51-year-old German man with ALS8 who had the P56S VAPB mutation independently of the...
Article
Full-text available
Background Half of all amyotrophic lateral sclerosis-frontotemporal spectrum disorder (ALS-FTSD) patients are classified as cognitively impaired, of which 10% have frontotemporal dementia (FTD), and an additional 40% suffer from a frontotemporal syndrome not severe enough to be described as dementia (cognitively impaired/ALSci). As changes in cereb...
Article
Full-text available
A structural covariance network (SCN) has been used successfully in structural magnetic resonance imaging (sMRI) studies. However, most SCNs have been constructed by a unitary marker that is insensitive for discriminating different disease phases. The aim of this study was to devise a novel regional radiomics similarity network (R2SN) that could pr...
Article
Full-text available
Background: Normal aging is associated with working memory decline. A decrease in working memory performance is associated with age-related changes in functional activation patterns in the dorsolateral prefrontal cortex (DLPFC). Cognitive training can improve cognitive performance in healthy older adults. We implemented a cognitive training study t...
Article
Full-text available
Background Previous research has described distinct subtypes of Alzheimer’s disease (AD) based on the differences in regional patterns of brain atrophy on MRI. We conducted a data-driven exploration of distinct AD neurodegeneration subtypes using FDG-PET as a sensitive molecular imaging marker of neurodegenerative processes. Methods Hierarchical c...
Chapter
Relevance maps derived from convolutional neural networks (CNN) indicate the influence of a particular image region on the decision of the CNN model. Individual maps are obtained for each single input 3D MRI image and various visualization options need to be adjusted to improve information content. In the use case of model prototyping and compariso...
Preprint
Full-text available
Background Previous research has described distinct subtypes of Alzheimer’s disease (AD) based on differences in regional patterns of brain atrophy on MRI. We conducted a data-driven exploration of distinct AD neurodegeneration subtypes using FDG-PET as a sensitive molecular imaging marker of neurodegenerative processes. Methods Hierarchical cluste...
Article
Full-text available
Background Cognitive decline has been found to be associated with gray matter atrophy and disruption of functional neural networks in Alzheimer’s disease (AD) in structural and functional imaging (fMRI) studies. Most previous studies have used single test scores of cognitive performance among monocentric cohorts. However, cognitive domain composite...
Preprint
Although convolutional neural networks (CNN) achieve high diagnostic accuracy for detecting Alzheimer's disease (AD) dementia based on magnetic resonance imaging (MRI) scans, they are not yet applied in clinical routine. One important reason for this is a lack of model comprehensibility. Recently developed visualization methods for deriving CNN rel...
Preprint
Background Structural covariance network (SCN) has been applied successfully to structural magnetic resonance imaging (MRI) study. However, most SCNs were constructed by the unitary marker, which was insensitive for the different disease phases. The aim of this study is to devise a novel regional radiomics similarity network (R2SN) that could provi...
Article
Background: TAR DNA-binding protein 43 (TDP-43) has been recognized as a frequent co-pathology of Alzheimer's disease (AD). The effect of the presence of TDP-43 pathology on in vivo measures of AD-related amyloid pathology using amyloid sensitive PET is still unresolved. Objective: To study the association of TDP-43 pathology with antemortem amy...
Article
Normal aging is associated with working memory decline. A decrease in working memory performance is associated with age‐related changes in functional activation patterns in the dorsolateral prefrontal cortex (DLPFC). Cognitive training can improve cognitive performance in healthy older adults. We implemented a cognitive training study to assess det...
Article
Cognitive decline in Alzheimer’s disease (AD) has been found associated with regional structural atrophy and functional disruption of neural networks, such as the default mode (DMN), visual (VN) and executive networks (EN) in structural and functional imaging (fMRI) studies, mostly using single test scores of cognitive performance among monocentric...
Poster
PET with the glucose analog FDG is a sensitive marker of neuronal dysfunction/degeneration. The aim of this study was to identify distinct metabolic subtypes in prodromal and dementia stages of AD by a data‐driven exploration of FDG‐PET and to compare the subtypes with respect to molecular biomarkers and clinical trajectory. First, we used hierarch...
Article
Although machine learning approaches achieve high diagnostic accuracy for detecting Alzheimer’s disease (AD) based on MRI scans, they are not applied in clinical routine due to a lack of suitable methods for model comprehensibility and interpretability. Recently developed visualization methods for convolutional neural networks (CNN) may fill this g...
Conference Paper
White matter hyperintensities (WMH), derived from T2‐weighted MRI, represent a sensitive imaging marker of cerebrovascular disease. To investigate WMH in large‐scale multicenter clinical studies, appropriate automated WMH segmentation algorithms are required. Therefore, prior evaluation of these methods is important, especially due to additional ch...
Article
Although machine learning approaches achieve high diagnostic accuracy when detecting Alzheimer’s disease (AD) based on MRI scans, they are rarely applied in clinical studies due to a lack of suitable methods for model comprehensibility and interpretability. Recent advances in convolutional neural networks (CNN) visualization algorithms and methods...
Preprint
Background: Previous research has described distinct subtypes of Alzheimer’s disease (AD) based on differences in regional patterns of brain atrophy on MRI. We conducted a data-driven exploration of distinct AD neurodegeneration subtypes using FDG-PET as a sensitive molecular imaging marker of neurodegenerative processes. Methods: Hierarchical clus...
Article
Full-text available
Objective To determine if PET‐based stages of regional amyloid deposition are associated with neuropathological phases of Aβ pathology. Methods We applied data‐driven regional frequency‐based and a‐priori striatum‐based PET staging approaches to ante‐mortem 18F‐Florbetapir‐PET scans of 30 cases from the Alzheimer’s Disease Neuroimaging Initiative...
Preprint
Full-text available
Deep Neural Networks - especially Convolutional Neural Network (ConvNet) has become the state-of-the-art for image classification, pattern recognition and various computer vision tasks. ConvNet has a huge potential in medical domain for analyzing medical data to diagnose diseases in an efficient way. Based on extracted features by ConvNet model fro...
Preprint
Full-text available
Convolutional neural networks (CNN) have become a powerful tool for detecting patterns in image data. Recent papers report promising results in the domain of disease detection using brain MRI data. Despite the high accuracy obtained from CNN models for MRI data so far, almost no papers provided information on the features or image regions driving t...
Article
Full-text available
Zusammenfassung Die limbisch prädominante altersassoziierte TDP-43(Transactivation response(TAR)-DNA-binding protein 43 kDa)-Enzephalopathie (LATE) wurde kürzlich als eigene neuropathologische Entität im Demenzspektrum charakterisiert. Neuropathologische Veränderungen im Sinne von LATE wurden zuvor bereits als Komorbidität der Alzheimer-Krankheit (...
Article
Full-text available
Alzheimer's disease (AD) is characterized by a sequence of pathological changes, which are commonly assessed in vivo using various brain imaging modalities such as magnetic resonance imaging (MRI) and positron emission tomography (PET). Currently, the most approaches to analyze statistical associations between regions and imaging modalities rely on...
Article
We tested the usefulness of a regional amyloid staging based on amyloid sensitive Positron Emission Tomography (PET) to predict conversion to cognitive impairment and dementia in preclinical and prodromal Alzheimer’s disease (AD). We analyzed 884 cases, including normal controls, and people with subjective cognitive decline or mild cognitive impair...
Chapter
Advances in medical imaging and convolutional neural networks (CNNs) have made it possible to achieve excellent diagnostic accuracy from CNNs comparable to human raters. However, CNNs are still not implemented in medical trials as they appear as a black box system and their inner workings cannot be properly explained. Therefore, it is essential to...
Article
Full-text available
Background Dysfunction of the cholinergic basal forebrain (cBF) is associated with cognitive decline in Alzheimer’s disease (AD). Multimodal MRI allows for the investigation of cBF changes in-vivo. In this study we assessed alterations in cBF functional connectivity (FC), mean diffusivity (MD), and volume across the spectrum of AD. We further asses...
Article
Full-text available
BACKGROUND: Dysfunction of the cholinergic basal forebrain (cBF) is associated with cognitive decline in Alzheimer’s disease (AD). Multimodal MRI allows for the investigation of cBF changes in-vivo. In this study we assessed alterations in cBF functional connectivity (FC), mean diffusivity (MD), and volume across the spectrum of AD. We further asse...
Article
Full-text available
Introduction Subjective cognitive decline (SCD) can represent a preclinical stage of Alzheimer’s disease. Diffusion tensor imaging (DTI) could aid an early diagnosis, yet only few monocentric DTI studies in SCD have been conducted, reporting heterogeneous results. We investigated microstructural changes in SCD in a larger, multicentric cohort. Met...
Article
Full-text available
Diffusion changes as determined by diffusion tensor imaging are potential indicators of microstructural lesions in people with mild cognitive impairment (MCI), prodromal Alzheimer's disease (AD), and AD dementia. Here we extended the scope of analysis toward subjective cognitive complaints as a pre-MCI at risk stage of AD. In a cohort of 271 partic...
Poster
Full-text available
Introduction: • Although deep learning approaches achieve high diagnostic accuracy to automatically detect neurodegenerative diseases-such as Alzheimer's disease-based on MRI and PET, they are currently not part of clinically applied diagnostic systems. • The main reason for this lack of clinical use is the shortcoming in proper methods for model c...
Article
Full-text available
Background Current methods of amyloid PET interpretation based on the binary classification of global amyloid signal fail to identify early phases of amyloid deposition. A recent analysis of 18F-florbetapir PET data from the Alzheimer’s disease Neuroimaging Initiative cohort suggested a hierarchical four-stage model of regional amyloid deposition t...
Article
Full-text available
Introduction We examined the association between cerebrospinal fluid (CSF) biomarkers of Alzheimer's disease, neural novelty responses, and brain volume in predementia old age. Methods We conducted a cross-sectional analysis of the observational, multicentric DZNE-Longitudinal Cognitive Impairment and Dementia Study (DELCODE) study. Seventy-six pa...
Article
The cholinergic basal forebrain (CBF), comprising different groups of cortically projecting cholinergic neurons, plays a crucial role in higher cognitive processes and has been implicated in diverse neuropsychiatric disorders. A distinct corticotopic organization of CBF projections has been revealed in animal studies, but little is known about thei...
Article
Full-text available
Alzheimer’s disease is characterized by a cascade of pathological processes that can be assessed in vivo using different neuroimaging methods. Recent research suggests a systematic sequence of pathogenic events on a global biomarker level, but little is known about the associations and dependencies of distinct lesion patterns on a regional level. M...
Conference Paper
Several neuroimaging markers have been established for the early diagnosis of Alzheimer’s disease, among them amyloid-β deposition, glucose metabolism, and grey matter volume. Up to now, these imaging modalities were mostly analyzed separately from each other, and little is known about the regional interrelation and dependency of these markers. Gau...
Article
Background: Hippocampal mean diffusivity (MD) measured by diffusion-tensor imaging is a promising diagnostic marker for Mild Cognitive Impairment (MCI) and dementia. Its performance has yet to be evaluated in primary care patients, who vary systematically from patients visiting specialized care settings. Objective: We assessed the diagnostic acc...
Article
Background: Alterations of intrinsic networks from resting state fMRI (rs-fMRI) have been suggested as functional biomarkers of Alzheimer's disease (AD). Objective: To determine the diagnostic accuracy of multicenter rs-fMRI for prodromal and preclinical stages of AD. Methods: We determined rs-fMRI functional connectivity based on Pearson's co...