arXiv subjects

Frank Tacke

Publications and source records attributed to Frank Tacke.

3 recordsLinked to original sources

arXiv2026

Ultra-wideband MRE of the human liver and spleen for viscoelastic model identification in hepatic inflammation

Magnetic resonance elastography (MRE) is established for noninvasive assessment of liver fibrosis. Conventional abdominal MRE is typically limited to 40-60 Hz. Lower frequencies remain largely unexplored, particularly with regard to hepatic inflammation. We developed ultra-wideband MRE covering 5-80 Hz to investigate frequency-resolved viscoelastic dispersion of the liver and spleen and to identify biomechanical markers of hepatic inflammation. Following phantom validation, nine healthy volunteers and nine patients with inflammatory liver disease were examined at 12 frequencies. Spatiotemporal phase unwrapping and frequency-adaptive wavefield preprocessing enabled reconstruction of shear wave speed (SWS), penetration rate (PR), and loss angle ($\phi$). Six rheological models were evaluated. The largest inflammation-associated changes were observed at frequencies below 20 Hz: $\phi$ increased by 63% (p<0.001), PR decreased by 37% (p=0.003), and SWS increased by 8% (p=0.008), indicating predominantly dissipative, rather than stiffness-related, changes and a shift toward fluid-like behavior with minor stiffness changes in the lower frequency regime. The rheological springpot model with serial dashpot provided the best fit and revealed distinct dispersion functions for liver and spleen. In patients, springpot elastic modulus increased (101%, p=0.001), while viscosity and springpot power-law exponent decreased (52%, p=0.002 and 58%, p<0.001) suggesting a shift from soft-fluid to stiff-solid liver properties. Ultra-wideband MRE revealed that inflammatory liver disease is associated with property shifts toward stronger dissipation and fluid-like behavior at low frequencies while displaying solid-like behavior at higher frequencies. Ultra-low frequency MRE may provide a diagnostic window into inflammation-associated liver viscoelasticity without full rheological modeling.

physics.med-ph↗

arXiv2025

Multimodal Deep Learning for Prediction of Progression-Free Survival in Patients with Neuroendocrine Tumors Undergoing 177Lu-based Peptide Receptor Radionuclide Therapy

Peptide receptor radionuclide therapy (PRRT) is an established treatment for metastatic neuroendocrine tumors (NETs), yet long-term disease control occurs only in a subset of patients. Predicting progression-free survival (PFS) could support individualized treatment planning. This study evaluates laboratory, imaging, and multimodal deep learning models for PFS prediction in PRRT-treated patients. In this retrospective, single-center study 116 patients with metastatic NETs undergoing 177Lu-DOTATOC were included. Clinical characteristics, laboratory values, and pretherapeutic somatostatin receptor positron emission tomography/computed tomographies (SR-PET/CT) were collected. Seven models were trained to classify low- vs. high-PFS groups, including unimodal (laboratory, SR-PET, or CT) and multimodal fusion approaches. Explainability was evaluated by feature importance analysis and gradient maps. Forty-two patients (36%) had short PFS (< 1 year), 74 patients long PFS (>1 year). Groups were similar in most characteristics, except for higher baseline chromogranin A (p = 0.003), elevated gamma-GT (p = 0.002), and fewer PRRT cycles (p < 0.001) in short-PFS patients. The Random Forest model trained only on laboratory biomarkers reached an AUROC of 0.59 +- 0.02. Unimodal three-dimensional convolutional neural networks using SR-PET or CT performed worse (AUROC 0.42 +- 0.03 and 0.54 +- 0.01, respectively). A multimodal fusion model laboratory values, SR-PET, and CT -augmented with a pretrained CT branch - achieved the best results (AUROC 0.72 +- 0.01, AUPRC 0.80 +- 0.01). Multimodal deep learning combining SR-PET, CT, and laboratory biomarkers outperformed unimodal approaches for PFS prediction after PRRT. Upon external validation, such models may support risk-adapted follow-up strategies.

cs.LG↗

arXiv2023

Leveraging weak complementary labels to improve semantic segmentation of hepatocellular carcinoma and cholangiocarcinoma in H&E-stained slides

In this paper, we present a deep learning segmentation approach to classify and quantify the two most prevalent primary liver cancers - hepatocellular carcinoma and intrahepatic cholangiocarcinoma - from hematoxylin and eosin (H&E) stained whole slide images. While semantic segmentation of medical images typically requires costly pixel-level annotations by domain experts, there often exists additional information which is routinely obtained in clinical diagnostics but rarely utilized for model training. We propose to leverage such weak information from patient diagnoses by deriving complementary labels that indicate to which class a sample cannot belong to. To integrate these labels, we formulate a complementary loss for segmentation. Motivated by the medical application, we demonstrate for general segmentation tasks that including additional patches with solely weak complementary labels during model training can significantly improve the predictive performance and robustness of a model. On the task of diagnostic differentiation between hepatocellular carcinoma and intrahepatic cholangiocarcinoma, we achieve a balanced accuracy of 0.91 (CI 95%: 0.86 - 0.95) at case level for 165 hold-out patients. Furthermore, we also show that leveraging complementary labels improves the robustness of segmentation and increases performance at case level.

cs.CV↗