Arnd Dörfler

Yixing Huang, Christoph Bert, Stefan Fischer et al.

Due to data privacy constraints, data sharing among multiple centers is restricted. Continual learning, as one approach to peer-to-peer federated learning, can promote multicenter collaboration on deep learning algorithm development by sharing intermediate models instead of training data. This work aims to investigate the feasibility of continual learning for multicenter collaboration on an exemplary application of brain metastasis identification using DeepMedic. 920 T1 MRI contrast enhanced volumes are split to simulate multicenter collaboration scenarios. A continual learning algorithm, synaptic intelligence (SI), is applied to preserve important model weights for training one center after another. In a bilateral collaboration scenario, continual learning with SI achieves a sensitivity of 0.917, and naive continual learning without SI achieves a sensitivity of 0.906, while two models trained on internal data solely without continual learning achieve sensitivity of 0.853 and 0.831 only. In a seven-center multilateral collaboration scenario, the models trained on internal datasets (100 volumes each center) without continual learning obtain a mean sensitivity value of 0.699. With single-visit continual learning (i.e., the shared model visits each center only once during training), the sensitivity is improved to 0.788 and 0.849 without SI and with SI, respectively. With iterative continual learning (i.e., the shared model revisits each center multiple times during training), the sensitivity is further improved to 0.914, which is identical to the sensitivity using mixed data for training. Our experiments demonstrate that continual learning can improve brain metastasis identification performance for centers with limited data. This study demonstrates the feasibility of applying continual learning for peer-to-peer federated learning in multicenter collaboration.

Philipp Sommer, Yixing Huang, Christoph Bert et al.

Stereotactic radiotherapy (SRT) is one of the most important treatment for patients with brain metastases (BM). Conventionally, following SRT patients are monitored by serial imaging and receive salvage treatments in case of significant tumor growth. We hypothesized that using radiomics and machine learning (ML), metastases at high risk for subsequent progression could be identified during follow-up prior to the onset of significant tumor growth, enabling personalized follow-up intervals and early selection for salvage treatment. All experiments are performed on a dataset from clinical routine of the Radiation Oncology department of the University Hospital Erlangen (UKER). The classification is realized via the maximum-relevance minimal-redundancy (MRMR) technique and support vector machines (SVM). The pipeline leads to a classification with a mean area under the curve (AUC) score of 0.83 in internal cross-validation and allows a division of the cohort into two subcohorts that differ significantly in their median time to progression (low-risk metastasis (LRM): 17.3 months, high-risk metastasis (HRM): 9.6 months, p < 0.01). The classification performance is especially enhanced by the analysis of medical images from different points in time (AUC 0.53 -> AUC 0.74). The results indicate that risk stratification of BM based on radiomics and machine learning during post-SRT follow-up is possible with good accuracy and should be further pursued to personalize and improve post-SRT follow-up.

Ahmed Gomaa, Yixing Huang, Amr Hagag et al.

Background: This research aims to improve glioblastoma survival prediction by integrating MR images, clinical and molecular-pathologic data in a transformer-based deep learning model, addressing data heterogeneity and performance generalizability. Method: We propose and evaluate a transformer-based non-linear and non-proportional survival prediction model. The model employs self-supervised learning techniques to effectively encode the high-dimensional MRI input for integration with non-imaging data using cross-attention. To demonstrate model generalizability, the model is assessed with the time-dependent concordance index (Cdt) in two training setups using three independent public test sets: UPenn-GBM, UCSF-PDGM, and RHUH-GBM, each comprising 378, 366, and 36 cases, respectively. Results: The proposed transformer model achieved promising performance for imaging as well as non-imaging data, effectively integrating both modalities for enhanced performance (UPenn-GBM test-set, imaging Cdt 0.645, multimodal Cdt 0.707) while outperforming state-of-the-art late-fusion 3D-CNN-based models. Consistent performance was observed across the three independent multicenter test sets with Cdt values of 0.707 (UPenn-GBM, internal test set), 0.672 (UCSF-PDGM, first external test set) and 0.618 (RHUH-GBM, second external test set). The model achieved significant discrimination between patients with favorable and unfavorable survival for all three datasets (logrank p 1.9\times{10}^{-8}, 9.7\times{10}^{-3}, and 1.2\times{10}^{-2}). Conclusions: The proposed transformer-based survival prediction model integrates complementary information from diverse input modalities, contributing to improved glioblastoma survival prediction compared to state-of-the-art methods. Consistent performance was observed across institutions supporting model generalizability.

Ahmed Gomaa, Yixing Huang, Pluvio Stephan et al.

Accurate differentiation of pseudoprogression (PsP) from True Progression (TP) following radiotherapy (RT) in glioblastoma (GBM) patients is crucial for optimal treatment planning. However, this task remains challenging due to the overlapping imaging characteristics of PsP and TP. This study therefore proposes a multimodal deep-learning approach utilizing complementary information from routine anatomical MR images, clinical parameters, and RT treatment planning information for improved predictive accuracy. The approach utilizes a self-supervised Vision Transformer (ViT) to encode multi-sequence MR brain volumes to effectively capture both global and local context from the high dimensional input. The encoder is trained in a self-supervised upstream task on unlabeled glioma MRI datasets from the open BraTS2021, UPenn-GBM, and UCSF-PDGM datasets to generate compact, clinically relevant representations from FLAIR and T1 post-contrast sequences. These encoded MR inputs are then integrated with clinical data and RT treatment planning information through guided cross-modal attention, improving progression classification accuracy. This work was developed using two datasets from different centers: the Burdenko Glioblastoma Progression Dataset (n = 59) for training and validation, and the GlioCMV progression dataset from the University Hospital Erlangen (UKER) (n = 20) for testing. The proposed method achieved an AUC of 75.3%, outperforming the current state-of-the-art data-driven approaches. Importantly, the proposed approach relies on readily available anatomical MRI sequences, clinical data, and RT treatment planning information, enhancing its clinical feasibility. The proposed approach addresses the challenge of limited data availability for PsP and TP differentiation and could allow for improved clinical decision-making and optimized treatment plans for GBM patients.

Ahmed Gomaa, Annette Schwarz, Ludwig Singer et al.

Background: Differentiating radiation necrosis (RN) from tumor progression after stereotactic radiosurgery (SRS) remains a critical challenge in brain metastases. While histopathology represents the gold standard, its invasiveness limits feasibility. Conventional supervised deep learning approaches are constrained by scarce biopsy-confirmed training data. Self-supervised learning (SSL) overcomes this by leveraging the growing availability of large-scale unlabeled brain metastases imaging datasets. Methods: In a two-phase deep learning strategy inspired by the foundation model paradigm, a Vision Transformer (ViT) was pre-trained via SSL on 10,167 unlabeled multi-source T1CE MRI sub-volumes. The pre-trained ViT was then fine-tuned for RN classification using a two-channel input (T1CE MRI and segmentation masks) on the public MOLAB dataset (n=109) using 20% of datasets as same-center held-out test set. External validation was performed on a second-center test cohort (n=28). Results: The self-supervised model achieved an AUC of 0.916 on the same-center test set and 0.764 on the second center test set, surpassing the fully supervised ViT (AUC 0.624/0.496; p=0.001/0.008) and radiomics (AUC 0.807/0.691; p=0.005/0.014). Multimodal integration further improved performance (AUC 0.947/0.821; p=0.073/0.001). Attention map visualizations enabled interpretability showing the model focused on clinically relevant lesion subregions. Conclusion: Large-scale pre-training on increasingly available unlabeled brain metastases datasets substantially improves AI model performance. A two-phase multimodal deep learning strategy achieved high accuracy in differentiating radiation necrosis from tumor progression using only routine T1CE MRI and standard clinical data, providing an interpretable, clinically accessible solution that warrants further validation.

Yixing Huang, Christoph Bert, Philipp Sommer et al.

Brain metastases occur frequently in patients with metastatic cancer. Early and accurate detection of brain metastases is very essential for treatment planning and prognosis in radiation therapy. To improve brain metastasis detection performance with deep learning, a custom detection loss called volume-level sensitivity-specificity (VSS) is proposed, which rates individual metastasis detection sensitivity and specificity in (sub-)volume levels. As sensitivity and precision are always a trade-off in a metastasis level, either a high sensitivity or a high precision can be achieved by adjusting the weights in the VSS loss without decline in dice score coefficient for segmented metastases. To reduce metastasis-like structures being detected as false positive metastases, a temporal prior volume is proposed as an additional input of DeepMedic. The modified network is called DeepMedic+ for distinction. Our proposed VSS loss improves the sensitivity of brain metastasis detection for DeepMedic, increasing the sensitivity from 85.3% to 97.5%. Alternatively, it improves the precision from 69.1% to 98.7%. Comparing DeepMedic+ with DeepMedic with the same VSS loss, 44.4% of the false positive metastases are reduced in the high sensitivity model and the precision reaches 99.6% for the high specificity model. The mean dice coefficient for all metastases is about 0.81. With the ensemble of the high sensitivity and high specificity models, on average only 1.5 false positive metastases per patient needs further check, while the majority of true positive metastases are confirmed. The ensemble learning is able to distinguish high confidence true positive metastases from metastases candidates that require special expert review or further follow-up, being particularly well-fit to the requirements of expert support in real clinical practice.

Bernhard Stimpel, Christopher Syben, Tobias Würfl et al.

Hybrid X-ray and magnetic resonance (MR) imaging promises large potential in interventional medical imaging applications due to the broad variety of contrast of MRI combined with fast imaging of X-ray-based modalities. To fully utilize the potential of the vast amount of existing image enhancement techniques, the corresponding information from both modalities must be present in the same domain. For image-guided interventional procedures, X-ray fluoroscopy has proven to be the modality of choice. Synthesizing one modality from another in this case is an ill-posed problem due to ambiguous signal and overlapping structures in projective geometry. To take on these challenges, we present a learning-based solution to MR to X-ray projection-to-projection translation. We propose an image generator network that focuses on high representation capacity in higher resolution layers to allow for accurate synthesis of fine details in the projection images. Additionally, a weighting scheme in the loss computation that favors high-frequency structures is proposed to focus on the important details and contours in projection imaging. The proposed extensions prove valuable in generating X-ray projection images with natural appearance. Our approach achieves a deviation from the ground truth of only $6$% and structural similarity measure of $0.913\,\pm\,0.005$. In particular the high frequency weighting assists in generating projection images with sharp appearance and reduces erroneously synthesized fine details.

Bernhard Stimpel, Christopher Syben, Franziska Schirrmacher et al.

Deep learning-based image processing is capable of creating highly appealing results. However, it is still widely considered as a "blackbox" transformation. In medical imaging, this lack of comprehensibility of the results is a sensitive issue. The integration of known operators into the deep learning environment has proven to be advantageous for the comprehensibility and reliability of the computations. Consequently, we propose the use of the locally linear guided filter in combination with a learned guidance map for general purpose medical image processing. The output images are only processed by the guided filter while the guidance map can be trained to be task-optimal in an end-to-end fashion. We investigate the performance based on two popular tasks: image super resolution and denoising. The evaluation is conducted based on pairs of multi-modal magnetic resonance imaging and cross-modal computed tomography and magnetic resonance imaging datasets. For both tasks, the proposed approach is on par with state-of-the-art approaches. Additionally, we can show that the input image's content is almost unchanged after the processing which is not the case for conventional deep learning approaches. On top, the proposed pipeline offers increased robustness against degraded input as well as adversarial attacks.

Christopher Syben, Bernhard Stimpel, Jonathan Lommen et al.

In this paper, we derive a neural network architecture based on an analytical formulation of the parallel-to-fan beam conversion problem following the concept of precision learning. The network allows to learn the unknown operators in this conversion in a data-driven manner avoiding interpolation and potential loss of resolution. Integration of known operators results in a small number of trainable parameters that can be estimated from synthetic data only. The concept is evaluated in the context of Hybrid MRI/X-ray imaging where transformation of the parallel-beam MRI projections to fan-beam X-ray projections is required. The proposed method is compared to a traditional rebinning method. The results demonstrate that the proposed method is superior to ray-by-ray interpolation and is able to deliver sharper images using the same amount of parallel-beam input projections which is crucial for interventional applications. We believe that this approach forms a basis for further work uniting deep learning, signal processing, physics, and traditional pattern recognition.

Bernhard Stimpel, Christopher Syben, Tobias Würfl et al.

The potential benefit of hybrid X-ray and MR imaging in the interventional environment is large due to the combination of fast imaging with high contrast variety. However, a vast amount of existing image enhancement methods requires the image information of both modalities to be present in the same domain. To unlock this potential, we present a solution to image-to-image translation from MR projections to corresponding X-ray projection images. The approach is based on a state-of-the-art image generator network that is modified to fit the specific application. Furthermore, we propose the inclusion of a gradient map in the loss function to allow the network to emphasize high-frequency details in image generation. Our approach is capable of creating X-ray projection images with natural appearance. Additionally, our extensions show clear improvement compared to the baseline method.

Franziska Schirrmacher, Thomas Köhler, Tobias Lindenberger et al.

This paper introduces an universal and structure-preserving regularization term, called quantile sparse image (QuaSI) prior. The prior is suitable for denoising images from various medical imaging modalities. We demonstrate its effectiveness on volumetric optical coherence tomography (OCT) and computed tomography (CT) data, which show different noise and image characteristics. OCT offers high-resolution scans of the human retina but is inherently impaired by speckle noise. CT on the other hand has a lower resolution and shows high-frequency noise. For the purpose of denoising, we propose a variational framework based on the QuaSI prior and a Huber data fidelity model that can handle 3-D and 3-D+t data. Efficient optimization is facilitated through the use of an alternating direction method of multipliers (ADMM) scheme and the linearization of the quantile filter. Experiments on multiple datasets emphasize the excellent performance of the proposed method.

Bernhard Stimpel, Christopher Syben, Tobias Würfl et al.

Hybrid imaging promises large potential in medical imaging applications. To fully utilize the possibilities of corresponding information from different modalities, the information must be transferable between the domains. In radiation therapy planning, existing methods make use of reconstructed 3D magnetic resonance imaging data to synthesize corresponding X-ray attenuation maps. In contrast, for fluoroscopic procedures only line integral data, i.e., 2D projection images, are present. The question arises which approaches could potentially be used for this MR to X-ray projection image-to-image translation. We examine three network architectures and two loss-functions regarding their suitability as generator networks for this task. All generators proved to yield suitable results for this task. A cascaded refinement network paired with a perceptual-loss function achieved the best qualitative results in our evaluation. The perceptual-loss showed to be able to preserve most of the high-frequency details in the projection images and, thus, is recommended for the underlying task and similar problems.

Christopher Syben, Bernhard Stimpel, Katharina Breininger et al.

In this paper, we present substantial evidence that a deep neural network will intrinsically learn the appropriate way to discretize the ideal continuous reconstruction filter. Currently, the Ram-Lak filter or heuristic filters which impose different noise assumptions are used for filtered back-projection. All of these, however, inhibit a fully data-driven reconstruction deep learning approach. In addition, the heuristic filters are not chosen in an optimal sense. To tackle this issue, we propose a formulation to directly learn the reconstruction filter. The filter is initialized with the ideal Ramp filter as a strong pre-training and learned in frequency domain. We compare the learned filter with the Ram-Lak and the Ramp filter on a numerical phantom as well as on a real CT dataset. The results show that the network properly discretizes the continuous Ramp filter and converges towards the Ram-Lak solution. In our view these observations are interesting to gain a better understanding of deep learning techniques and traditional analytic techniques such as Wiener filtering and discretization theory. Furthermore, this will allow fully trainable data-driven reconstruction deep learning approaches.