About me:
I am a Postdoctoral Researcher. I have previously worked as data scientist at the Department of Data Science in the Institute of Cancer Research. Before that, I completed my PhD at Technische Universität München. My main research interest is machine learning for clinical data analysis.
Also see my personal website for details.
Contact: sebastian (DOT) poelsterl (AT) med (DOT) uni-muenchen (DOT) de
Research Interests:
- Survival analysis
- Learning from unstructured data
- High-dimensional data, in particular genomics
- Convex optimisation
Awards:
- Won subchallenge of the Prostate Cancer DREAM Challenge
Publications:
Wolf, Tom Nuno; Bongratz, Fabian; Rickmann, Anne-Marie; Pölsterl, Sebastian; Wachinger, Christian Keep the Faith: Faithful Explanations in Convolutional Neural Networks for Case-Based Reasoning Proceedings Article Forthcoming In: AAAI Conference on Artificial Intelligence, 2024, Forthcoming. @inproceedings{nokey,Explaining predictions of black-box neural networks is crucial when applied to decision-critical tasks. Thus, attribution maps are commonly used to identify important image regions, despite prior work showing that humans prefer explanations based on similar examples. To this end, ProtoPNet learns a set of class-representative feature vectors (prototypes) for case-based reasoning. During inference, similarities of latent features to prototypes are linearly classified to form predictions and attribution maps are provided to explain the similarity. In this work, we evaluate whether architectures for case-based reasoning fulfill established axioms required for faithful explanations using the example of ProtoPNet. We show that such architectures allow the extraction of faithful explanations. However, we prove that the attribution maps used to explain the similarities violate the axioms. We propose a new procedure to extract explanations for trained ProtoPNets, named ProtoPFaith. Conceptually, these explanations are Shapley values, calculated on the similarity scores of each prototype. They allow to faithfully answer which prototypes are present in an unseen image and quantify each pixel's contribution to that presence, thereby complying with all axioms. The theoretical violations of ProtoPNet manifest in our experiments on three datasets (CUB-200-2011, Stanford Dogs, RSNA) and five architectures (ConvNet, ResNet, ResNet50, WideResNet50, ResNeXt50). Our experiments show a qualitative difference between the explanations given by ProtoPNet and ProtoPFaith. Additionally, we quantify the explanations with the Area Over the Perturbation Curve, on which ProtoPFaith outperforms ProtoPNet on all experiments by a factor >10**3. |
Wachinger, Christian; Wolf, Tom Nuno; Pölsterl, Sebastian Deep learning for the prediction of type 2 diabetes mellitus from neck-to-knee Dixon MRI in the UK biobank Journal Article In: Heliyon, vol. 9, no. 11, 2023. @article{nokey, |
Wolf, Tom Nuno; Pölsterl, Sebastian; Wachinger, Christian Don't PANIC: Prototypical Additive Neural Network for Interpretable Classification of Alzheimer's Disease Proceedings Article In: Information Processing in Medical Imaging (IPMI), 2023. @inproceedings{nokey,Alzheimer's disease (AD) has a complex and multifactorial etiology, which requires integrating information about neuroanatomy, genetics, and cerebrospinal fluid biomarkers for accurate diagnosis. Hence, recent deep learning approaches combined image and tabular information to improve diagnostic performance. However, the black-box nature of such neural networks is still a barrier for clinical applications, in which understanding the decision of a heterogeneous model is integral. We propose PANIC, a prototypical additive neural network for interpretable AD classification that integrates 3D image and tabular data. It is interpretable by design and, thus, avoids the need for post-hoc explanations that try to approximate the decision of a network. Our results demonstrate that PANIC achieves state-of-the-art performance in AD classification, while directly providing local and global explanations. Finally, we show that PANIC extracts biologically meaningful signatures of AD, and satisfies a set of desirable desiderata for trustworthy machine learning. Our implementation is available at https://github.com/ai-med/PANIC. |
Rickmann, Anne-Marie; Bongratz, Fabian; P"olsterl, Sebastian; Sarasua, Ignacio; Wachinger, Christian Joint Reconstruction and Parcellation of Cortical Surfaces Proceedings Article In: International Workshop on Machine Learning in Clinical Neuroimaging, pp. 3–12, Springer, 2022. @inproceedings{rickmann2022joint, |
Wolf, Tom Nuno; Pölsterl, Sebastian; Wachinger, Christian DAFT: A Universal Module to Interweave Tabular Data and 3D Images in CNNs Journal Article In: NeuroImage, pp. 119505, 2022. @article{WOLF2022119505,Prior work on Alzheimer’s Disease (AD) has demonstrated that convolutional neural networks (CNNs) can leverage the high-dimensional image information for diagnosing patients. Beside such data-driven approaches, many established biomarkers exist and are typically represented as tabular data, such as demographics, genetic alterations, or laboratory measurements from cerebrospinal fluid. However, little research has focused on the effective integration of tabular data into existing CNN architectures to improve patient diagnosis. We introduce the Dynamic Affine Feature Map Transform (DAFT), a general-purpose module for CNNs that incites or represses high-level concepts learned from a 3D image by conditioning feature maps of a convolutional layer on both a patient’s image and tabular clinical information. This is achieved by using an auxiliary neural network that outputs a scaling factor and offset to dynamically apply an affine transformation to the feature maps of a convolutional layer. In our experiments on AD diagnosis and time-to-dementia prediction, we show that the DAFT is highly effective in combining 3D image and tabular information by achieving a mean balanced accuracy of 0.622 for diagnosis, and mean c-index of 0.748 for time-to-dementia prediction, thus outperforming all baseline methods. Finally, our extensive ablation study and empirical experiments reveal that the performance improvement due to the DAFT is robust with respect to many design choices. |
Bongratz, Fabian; Rickmann, Anne-Marie; Pölsterl, Sebastian; Wachinger, Christian Vox2Cortex: Fast Explicit Reconstruction of Cortical Surfaces from 3D MRI Scans with Geometric Deep Neural Networks Proceedings Article In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022. @inproceedings{Bongratz_2022_CVPR,The reconstruction of cortical surfaces from brain magnetic resonance imaging (MRI) scans is essential for quantitative analyses of cortical thickness and sulcal morphology. Although traditional and deep learning-based algorithmic pipelines exist for this purpose, they have two major drawbacks: lengthy runtimes of multiple hours (traditional) or intricate post-processing, such as mesh extraction and topology correction (deep learning-based). In this work, we address both of these issues and propose Vox2Cortex, a deep learning-based algorithm that directly yields topologically correct, three-dimensional meshes of the boundaries of the cortex. Vox2Cortex leverages convolutional and graph convolutional neural networks to deform an initial template to the densely folded geometry of the cortex represented by an input MRI scan. We show in extensive experiments on three brain MRI datasets that our meshes are as accurate as the ones reconstructed by state-of-the-art methods in the field, without the need for time- and resource-intensive post-processing. To accurately reconstruct the tightly folded cortex, we work with meshes containing about 168,000 vertices at test time, scaling deep explicit reconstruction methods to a new level. |
Ignacio Sarasua, Sebastian Pölsterl; Wachinger, Christian Hippocampal Representations for Deep Learning on Alzheimer's Disease Journal Article In: Scientific reports, vol. 12, no. 1, pp. 1-13, 2022, ISSN: 2045-2322. @article{sarasua2022hippocampal,Deep learning offers a powerful approach for analyzing hippocampal changes in Alzheimer's disease (AD) without relying on handcrafted features. Nevertheless, an input format needs to be selected to pass the image information to the neural network, which has wide ramifications for the analysis, but has not been evaluated yet. We compare five hippocampal representations (and their respective tailored network architectures) that span from raw images to geometric representations like meshes and point clouds. We performed a thorough evaluation for the prediction of AD diagnosis and time-to-dementia prediction with experiments on an independent test dataset. In addition, we evaluated the ease of interpretability for each representation-network pair. |
Ronge, Raphael; Nho, Kwangsik; Wachinger, Christian; Pölsterl, Sebastian Alzheimer's Disease Diagnosis via Deep Factorization Machine Models Conference Machine Learning in Medical Imaging (MLMI), 2021. @conference{Ronge2021,The current state-of-the-art deep neural networks (DNNs) for Alzheimer's Disease diagnosis use different biomarker combinations to classify patients, but do not allow extracting knowledge about the interactions of biomarkers. However, to improve our understanding of the disease, it is paramount to extract such knowledge from the learned model. In this paper, we propose a Deep Factorization Machine model that combines the ability of DNNs to learn complex relationships and the ease of interpretability of a linear model. The proposed model has three parts: (i) an embedding layer to deal with sparse categorical data, (ii) a Factorization Machine to efficiently learn pairwise interactions, and (iii) a DNN to implicitly model higher order interactions. In our experiments on data from the Alzheimer's Disease Neuroimaging Initiative, we demonstrate that our proposed model classifies cognitive normal, mild cognitive impaired, and demented patients more accurately than competing models. In addition, we show that valuable knowledge about the interactions among biomarkers can be obtained. |
Sarasua, Ignacio; Pölsterl, Sebastian; Wachinger, Christian TransforMesh: A Transformer Network for Longitudinal modeling of Anatomical Meshes Conference Machine Learning in Medical Imaging (MLMI), 2021. @conference{sarasua2021transformesh,The longitudinal modeling of neuroanatomical changes related to Alzheimer's disease (AD) is crucial for studying the progression of the disease. To this end, we introduce TransforMesh, a spatio-temporal network based on transformers that models longitudinal shape changes on 3D anatomical meshes. While transformer and mesh networks have recently shown impressive performances in natural language processing and computer vision, their application to medical image analysis has been very limited. To the best of our knowledge, this is the first work that combines transformer and mesh networks. Our results show that TransforMesh can model shape trajectories better than other baseline architectures that do not capture temporal dependencies. Moreover, we also explore the capabilities of TransforMesh in detecting structural anomalies of the hippocampus in patients developing AD. |
Pölsterl, Sebastian; Wolf, Tom Nuno; Wachinger, Christian Combining 3D Image and Tabular Data via the Dynamic Affine Feature Map Transform Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI), 2021. @conference{Poelsterl2021-daft,Prior work on diagnosing Alzheimer’s disease from magnetic resonance images of the brain established that convolutional neural networks (CNNs) can leverage the high-dimensional image information for classifying patients. However, little research focused on how these models can utilize the usually low-dimensional tabular information, such as patient demographics or laboratory measurements. We introduce the Dynamic Affine Feature Map Transform (DAFT), a general-purpose module for CNNs that dynamically rescales and shifts the feature maps of a convolutional layer, conditional on a patient’s tabular clinical information. We show that DAFT is highly effective in combining 3D image and tabular information for diagnosis and time-to-dementia prediction, where it outperforms competing CNNs with a mean balanced accuracy of 0.622 and mean c-index of 0.748, respectively. Our extensive ablation study provides valuable insights into the architectural properties of DAFT. Our implementation is available at https://github.com/ai-med/DAFT. |
Pölsterl, Sebastian; Aigner, Christina; Wachinger, Christian Scalable, Axiomatic Explanations of Deep Alzheimer's Diagnosis from Heterogeneous Data Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI), 2021. @conference{Poelsterl2021-svehnn,Deep Neural Networks (DNNs) have an enormous potential to learn from complex biomedical data. In particular, DNNs have been used to seamlessly fuse heterogeneous information from neuroanatomy, genetics, biomarkers, and neuropsychological tests for highly accurate Alzheimer’s disease diagnosis. On the other hand, their black-box nature is still a barrier for the adoption of such a system in the clinic, where interpretability is absolutely essential. We propose Shapley Value Explanation of Heterogeneous Neural Networks (SVEHNN) for explaining the Alzheimer’s diagnosis made by a DNN from the 3D point cloud of the neuroanatomy and tabular biomarkers. Our explanations are based on the Shapley value, which is the unique method that satisfies all fundamental axioms for local explanations previously established in the literature. Thus, SVEHNN has many desirable characteristics that previous work on interpretability for medical decision making is lacking. To avoid the exponential time complexity of the Shapley value, we propose to transform a given DNN into a Lightweight Probabilistic Deep Network without re-training, thus achieving a complexity only quadratic in the number of features. In our experiments on synthetic and real data, we show that we can closely approximate the exact Shapley value with a dramatically reduced runtime and can reveal the hidden knowledge the network has learned from the data. |
Pölsterl, Sebastian; Wachinger, Christian Estimation of Causal Effects in the Presence of Unobserved Confounding in the Alzheimer's Continuum Conference IPMI: International Conference on Information Processing in Medical Imaging 2021, 2021. @conference{Poelsterl2021-causal-effects-in-ad,Studying the relationship between neuroanatomy and cognitive decline due to Alzheimer's has been a major research focus in the last decade. However, to infer cause-effect relationships rather than simple associations from observational data, we need to (i) express the causal relationships leading to cognitive decline in a graphical model, and (ii) ensure the causal effect of interest is identifiable from the collected data. We derive a causal graph from the current clinical knowledge on cause and effect in the Alzheimer's disease continuum, and show that identifiability of the causal effect requires all confounders to be known and measured. However, in complex neuroimaging studies, we neither know all potential confounders nor do we have data on them. To alleviate this requirement, we leverage the dependencies among multiple causes by deriving a substitute confounder via a probabilistic latent factor model. In our theoretical analysis, we prove that using the substitute confounder enables identifiability of the causal effect of neuroanatomy on cognition. We quantitatively evaluate the effectiveness of our approach on semi-synthetic data, where we know the true causal effects, and illustrate its use on real data on the Alzheimer's disease continuum, where it reveals important causes that otherwise would have been missed. |
Wachinger, Christian; Rieckmann, Anna; Pölsterl, Sebastian Detect and correct bias in multi-site neuroimaging datasets Journal Article In: Medical Image Analysis, vol. 67, pp. 101879, 2020. @article{wachinger2020detect,The desire to train complex machine learning algorithms and to increase the statistical power in association studies drives neuroimaging research to use ever-larger datasets. The most obvious way to increase sample size is by pooling scans from independent studies. However, simple pooling is often ill-advised as selection, measurement, and confounding biases may creep in and yield spurious correlations. In this work, we combine 35,320 magnetic resonance images of the brain from 17 studies to examine bias in neuroimaging. In the first experiment, Name That Dataset, we provide empirical evidence for the presence of bias by showing that scans can be correctly assigned to their respective dataset with 71.5% accuracy. Given such evidence, we take a closer look at confounding bias, which is often viewed as the main shortcoming in observational studies. In practice, we neither know all potential confounders nor do we have data on them. Hence, we model confounders as unknown, latent variables. Kolmogorov complexity is then used to decide whether the confounded or the causal model provides the simplest factorization of the graphical model. Finally, we present methods for dataset harmonization and study their ability to remove bias in imaging features. In particular, we propose an extension of the recently introduced ComBat algorithm to control for global variation across image features, inspired by adjusting for unknown population stratification in genetics. Our results demonstrate that harmonization can reduce dataset-specific information in image features. Further, confounding bias can be reduced and even turned into a causal relationship. However, harmonization also requires caution as it can easily remove relevant subject-specific information. Code is available at https://github.com/ai-med/Dataset-Bias. |
Pölsterl, Sebastian; Wachinger, Christian Adversarial Learned Molecular Graph Inference and Generation Proceedings Article In: European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD), 2020. @inproceedings{Poelsterl2020-almgig,Recent methods for generating novel molecules use graph representations of molecules and employ various forms of graph convolutional neural networks for inference. However, training requires solving an expensive graph isomorphism problem, which previous approaches do not address or solve only approximately. In this work, we propose ALMGIG, a likelihood-free adversarial learning framework for inference and de novo molecule generation that avoids explicitly computing a reconstruction loss. Our approach extends generative adversarial networks by including an adversarial cycle-consistency loss to implicitly enforce the reconstruction property. To capture properties unique to molecules, such as valence, we extend the Graph Isomorphism Network to multi-graphs. To quantify the performance of models, we propose to compute the distance between distributions of physicochemical properties with the 1-Wasserstein distance. We demonstrate that ALMGIG more accurately learns the distribution over the space of molecules than all baselines. Moreover, it can be utilized for drug discovery by efficiently searching the space of molecules using molecules' continuous latent representation. Our code is available at https://github.com/ai-med/almgig |
Senapati, Jyotirmay; Roy, Abhijit Guha; Pölsterl, Sebastian; Gutmann, Daniel; Gatidis, Sergios; Schlett, Christopher; Peters, Anette; Bamberg, Fabian; Wachinger, Christian Bayesian Neural Networks for Uncertainty Estimation of Imaging Biomarkers Proceedings Article In: Liu, Mingxia; Yan, Pingkun; Lian, Chunfeng; Cao, Xiaohuan (Ed.): Machine Learning in Medical Imaging, pp. 270–280, Springer International Publishing, Cham, 2020, ISBN: 978-3-030-59861-7. @inproceedings{10.1007/978-3-030-59861-7_28,Image segmentation enables to extract quantitative measures from scans that can serve as imaging biomarkers for diseases. However, segmentation quality can vary substantially across scans, and therefore yield unfaithful estimates in the follow-up statistical analysis of biomarkers. The core problem is that segmentation and biomarker analysis are performed independently. We propose to propagate segmentation uncertainty to the statistical analysis to account for variations in segmentation confidence. To this end, we evaluate four Bayesian neural networks to sample from the posterior distribution and estimate the uncertainty. We then assign confidence measures to the biomarker and propose statistical models for its integration in group analysis and disease classification. Our results for segmenting the liver in patients with diabetes mellitus clearly demonstrate the improvement of integrating biomarker uncertainty in the statistical inference. |
Sarasua, Ignacio; Poelsterl, Sebastian; Wachinger, Christian Recalibration of Neural Networks for Point Cloud Analysis Proceedings Article In: 3DV, 2020. @inproceedings{sarasua2020recalibration,Spatial and channel re-calibration have become powerful concepts in computer vision. Their ability to capture long-range dependencies is especially useful for those networks that extract local features, such as CNNs. While re-calibration has been widely studied for image analysis, it has not yet been used on shape representations. In this work, we introduce re-calibration modules on deep neural networks for 3D point clouds. We propose a set of re-calibration blocks that extend Squeeze and Excitation blocks and that can be added to any network for 3D point cloud analysis that builds a global descriptor by hierarchically combining features from multiple local neighborhoods. We run two sets of experiments to validate our approach. First, we demonstrate the benefit and versatility of our proposed modules by incorporating them into three state-of-the-art networks for 3D point cloud analysis: PointNet++, DGCNN, and RSCNN. We evaluate each network on two tasks: object classification on ModelNet40, and object part segmentation on ShapeNet. Our results show an improvement of up to 1% in accuracy for ModelNet40 compared to the baseline method. In the second set of experiments, we investigate the benefits of re-calibration blocks on Alzheimer's Disease (AD) diagnosis. Our results demonstrate that our proposed methods yield a 2% increase in accuracy for diagnosing AD and a 2.3% increase in concordance index for predicting AD onset with time-to-event analysis. Concluding, re-calibration improves the accuracy of point cloud architectures, while only minimally increasing the number of parameters. |
Roy, Abhijit Guha; Siddiqui, Shayan; Pölsterl, Sebastian; Navab, Nassir; Wachinger, Christian 'Squeeze & Excite' Guided Few-Shot Segmentation of Volumetric Images Journal Article In: Medical Image Analysis, 2019. @article{few_shot_agr_2019, |
Pölsterl, Sebastian; Gutiérrez-Becker, Benjamı́n; Sarasua, Ignacio; Roy, Abhijit Guha; Wachinger, Christian An AutoML Approach for the Prediction of Fluid Intelligence from MRI-Derived Features Proceedings Article In: Adolescent Brain Cognitive Development Neurocognitive Prediction Challenge (ABCD-NP-Challenge), 2019. @inproceedings{Poelsterl2019-ABCDb,We propose an AutoML approach for the prediction of fluid intelligence from T1-weighted magnetic resonance images. We extracted 122 features from MRI scans and employed Sequential Model-based Algorithm Configuration to search for the best prediction pipeline, including the best data pre-processing and regression model. In total, we evaluated over 2600 prediction pipelines. We studied our final model by employing results from game theory in the form of Shapley values. Results indicate that predicting fluid intelligence from volume measurements is a challenging task with many challenges. We found that our final ensemble of 50 prediction pipelines associated larger parahippocampal gyrus volumes with lower fluid intelligence, and higher pons white matter volume with higher fluid intelligence. |
Pölsterl, Sebastian; Gutiérrez-Becker, Benjamı́n; Sarasua, Ignacio; Roy, Abhijit Guha; Wachinger, Christian Prediction of Fluid Intelligence from T1-Weighted Magnetic Resonance Images Proceedings Article In: Adolescent Brain Cognitive Development Neurocognitive Prediction Challenge (ABCD-NP-Challenge), 2019. @inproceedings{Poelsterl2019-ABCDa,We study predicting fluid intelligence of 9–10 year old children from T1-weighted magnetic resonance images. We extract volume and thickness measurements from MRI scans using FreeSurfer and the SRI24 atlas. We empirically compare two predictive models: (i) an ensemble of gradient boosted trees and (ii) a linear ridge regression model. For both, a Bayesian black-box optimizer for finding the best suitable prediction model is used. To systematically analyze feature importance our model, we employ results from game theory in the form of Shapley values. Our model with gradient boosting and FreeSurfer measures ranked third place among 24 submissions to the ABCD Neurocognitive Prediction Challenge. Our results on feature importance could be used to guide future research on the neurobiological mechanisms behind fluid intelligence in children. |
Pölsterl, Sebastian; Sarasua, Ignacio; Gutiérrez-Becker, Benjamín; Wachinger, Christian A Wide and Deep Neural Network for Survival Analysis from Anatomical Shape and Tabular Clinical Data Proceedings Article In: Data and Machine Learning Advances with Multiple Views Workshop, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD), 2019. @inproceedings{Poelsterl2019-WideAndDeep,We introduce a wide and deep neural network for prediction of progression from patients with mild cognitive impairment to Alzheimer's disease. Information from anatomical shape and tabular clinical data (demographics, biomarkers) are fused in a single neural network. The network is invariant to shape transformations and avoids the need to identify point correspondences between shapes. To account for right censored time-to-event data, i.e., when it is only known that a patient did not develop Alzheimer's disease up to a particular time point, we employ a loss commonly used in survival analysis. Our network is trained end-to-end to combine information from a patient's hippocampus shape and clinical biomarkers. Our experiments on data from the Alzheimer's Disease Neuroimaging Initiative demonstrate that our proposed model is able to learn a shape descriptor that augments clinical biomarkers and outperforms a deep neural network on shape alone and a linear model on common clinical biomarkers. |
Wachinger, Christian; Becker, Benjamin Gutierrez; Rieckmann, Anna; Pölsterl, Sebastian Quantifying Confounding Bias in Neuroimaging Datasets with Causal Inference Book Section In: Medical Image Computing and Computer Aided Intervention, 2019. @incollection{wachinger2019causal,Neuroimaging datasets keep growing in size to address increasingly complex medical questions. However, even the largest datasets today alone are too small for training complex machine learning models. A potential solution is to increase sample size by pooling scans from several datasets. In this work, we combine 12,207 MRI scans from 15 studies and show that simple pooling is often ill-advised due to introducing various types of biases in the training data. First, we systematically define these biases. Second, we detect bias by experimentally showing that scans can be correctly assigned to their respective dataset with 73.3% accuracy. Finally, we propose to tell causal from confounding factors by quantifying the extent of confounding and causality in a single dataset using causal inference. We achieve this by finding the simplest graphical model in terms of Kolmogorov complexity. As Kolmogorov complexity is not directly computable, we employ the minimum description length to approximate it. We empirically show that our approach is able to estimate plausible causal relationships from real neuroimaging data. |
Roy, Abhijit Guha; Siddiqui, Shayan; Pölsterl, Sebastian; Navab, Nassir; Wachinger, Christian Braintorrent: A peer-to-peer environment for decentralized federated learning Journal Article In: arXiv preprint arXiv:1905.06731, 2019. @article{roy2019braintorrent, |
Narazani, Marla; Sarasua, Ignacio; Pölsterl, Sebastian; Lizarraga, Aldana; Yakushev, Igor; Wachinger, Christian Is a PET all you need? A multi-modal study for Alzheimer's disease using 3D CNNs Proceedings Article In: Medical Image Computing and Computer-Assisted Intervention (MICCAI), 0000. @inproceedings{Narazani2022,Alzheimer’s Disease (AD) is the most common form of dementia and often difficult to diagnose due to the multifactorial etiology of dementia. Recent works on neuroimaging-based computer-aided diagnosis with deep neural networks (DNNs) showed that fusing structural magnetic resonance images (sMRI) and fluorodeoxyglucose positron emission tomography (FDG-PET) leads to improved accuracy in a study population of healthy controls and subjects with AD. However, this result conflicts with the established clinical knowledge that FDG-PET better captures AD-specific pathologies than sMRI. Therefore, we propose a framework for the systematic evaluation of multi-modal DNNs and critically re-evaluate single- and multi-modal DNNs based on FDG-PET and sMRI for binary healthy vs. AD, and three-way healthy/mild cognitive impairment/AD classification. Our experiments demonstrate that a single-modality network using FDG-PET performs better than MRI (accuracy 0.91 vs 0.87) and does not show improvement when combined. This conforms with the established clinical knowledge on AD biomarkers, but raises questions about the true benefit of multi-modal DNNs. We argue that future work on multi-modal fusion should systematically assess the contribution of individual modalities following our proposed evaluation framework. Finally, we encourage the community to go beyond healthy vs. AD classification and focus on differential diagnosis of dementia, where fusing multi-modal image information conforms with a clinical need. |
Sarasua, Ignacio; Pölsterl, Sebastian; Wachinger, Christian CASHformer: Cognition Aware SHape Transformer for Longitudinal Analysis Proceedings Article In: Medical Image Computing and Computer-Assisted Intervention (MICCAI), 0000. @inproceedings{Sarasua2022b, |
You must be logged in to post a comment.