Search

Lung cancer is the leading cause of cancer-related deaths worldwide. Recently, we have shown that Notch1 inhibition resulted in substantial cell death of non-small cell lung cancer (NSCLC) cells in vitro. New compounds targeting Notch signal transduction have been developed and are now being tested in clinical trials. However, the tumorigenic role of individual Notch receptors in vivo remains largely unclear. Using a KrasG12D-driven endogenous NSCLC mouse model, we analyzed the effect of conditional Notch1 and Notch2 receptor deletion on NSCLC tumorigenesis. Notch1 deficiency led to a reduced early tumor formation and lower activity of MAPK compared with the controls. Unexpectedly, Notch2 deletion resulted in a dramatically increased carcinogenesis and increased MAPK activity. These mice died significantly earlier due to rapidly growing tumor burden. We found that Notch1 regulates Ras/MAPK pathway via HES1-induced repression of the DUSP1 promoter encoding a phosphatase specifically suppressing pERK1/2. Interestingly, Notch1 but not Notch2 ablation leads to decreased HES1 and DUSP1 expression. However, Notch2-depleted tumors showed an appreciable increase in β-catenin expression, a known activator of HES1 and important lung cancer oncogene. Characteristically for β-catenin upregulation, we found that the majority of Notch2-deficient tumors revealed an undifferentiated phenotype as determined by their morphology, E-Cadherin and TTF1 expression levels. In addition, these carcinomas showed aggressive growth patterns with bronchus invasion and obstruction. Together, we show that Notch2 mediates differentiation and has tumor suppressor functions during lung carcinogenesis, whereas Notch1 promotes tumor initiation and progression. These data are further supported by immunohistochemical analysis of human NSCLC samples showing loss or downregulation of Notch2 compared with normal lung tissue. In conclusion, this is the first study characterizing the in vivo functions of Notch1 and Notch2 in KrasG12D-driven NSCLC tumorigenesis. These data highlight the clinical importance of a thorough understanding of Notch signaling especially with regard to Notch-targeted therapies. [ABSTRACT FROM AUTHOR]

Clinical decision-making is one of the most impactful parts of a physician's responsibilities and stands to benefit greatly from artificial intelligence solutions and large language models (LLMs) in particular. However, while LLMs have achieved excellent performance on medical licensing exams, these tests fail to assess many skills necessary for deployment in a realistic clinical decision-making environment, including gathering information, adhering to guidelines, and integrating into clinical workflows. Here we have created a curated dataset based on the Medical Information Mart for Intensive Care database spanning 2,400 real patient cases and four common abdominal pathologies as well as a framework to simulate a realistic clinical setting. We show that current state-of-the-art LLMs do not accurately diagnose patients across all pathologies (performing significantly worse than physicians), follow neither diagnostic nor treatment guidelines, and cannot interpret laboratory results, thus posing a serious risk to the health of patients. Furthermore, we move beyond diagnostic accuracy and demonstrate that they cannot be easily integrated into existing workflows because they often fail to follow instructions and are sensitive to both the quantity and order of information. Overall, our analysis reveals that LLMs are currently not ready for autonomous clinical decision-making while providing a dataset and framework to guide future studies., (© 2024. The Author(s).)

177 Lu-DOTATATE therapy is an effective treatment for advanced neuroendocrine tumors, despite its dose-limiting hematotoxicity. Herein, the significance of off-target splenic irradiation is unknown. Our study aims to identify predictive markers of peptide receptor radionuclide therapy-induced leukopenia. Methods: We retrospectively analyzed blood counts and imaging data of 88 patients with histologically confirmed, unresectable metastatic neuroendocrine tumors who received 177 Lu-DOTATATE treatment at our institution from February 2009 to July 2021. Inclusion criterium was a tumor uptake equivalent to or greater than that in the liver on baseline receptor imaging. We excluded patients with less than 24 mo of follow-up and those patients who received fewer than 4 treatment cycles, additional therapies, or blood transfusions during follow-up. Results: Our study revealed absolute and relative white blood cell counts and relative spleen volume reduction as independent predictors of radiation-induced leukopenia at 24 mo. However, a 30% decline in spleen volume 12 mo after treatment most accurately predicted patients proceeding to leukopenia at 24 mo (receiver operating characteristic area under the curve of 0.91, sensitivity of 0.93, and specificity of 0.90), outperforming all other parameters by far. Conclusion: Automated splenic volume assessments demonstrated superior predictive capabilities for the development of leukopenia in patients undergoing 177 Lu-DOTATATE treatment compared with conventional laboratory parameters. The reduction in spleen size proves to be a valuable, routinely available, and quantitative imaging-based biomarker for predicting radiation-induced leukopenia. This suggests potential clinical applications for risk assessment and management., (© 2024 by the Society of Nuclear Medicine and Molecular Imaging.)

Radiomics features can reveal hidden patterns in a tumor but usually lack an underlying biologic rationale. In this work, we aimed to investigate whether there is a correlation between radiomics features extracted from [ 18 F]FDG PET images and histologic expression patterns of a glycolytic marker, monocarboxylate transporter-4 (MCT4), in pancreatic cancer. Methods: A cohort of pancreatic ductal adenocarcinoma patients ( n = 29) for whom both tumor cross sections and [ 18 F]FDG PET/CT scans were available was used to develop an [ 18 F]FDG PET radiomics signature. By using immunohistochemistry for MCT4, we computed density maps of MCT4 expression and extracted pathomics features. Cluster analysis identified 2 subgroups with distinct MCT4 expression patterns. From corresponding [ 18 F]FDG PET scans, radiomics features that associate with the predefined MCT4 subgroups were identified. Results: Complex heat map visualization showed that the MCT4-high/heterogeneous subgroup was correlating with a higher MCT4 expression level and local variation. This pattern linked to a specific [ 18 F]FDG PET signature, characterized by a higher SUV mean and SUV max and second-order radiomics features, correlating with local variation. This MCT4-based [ 18 F]FDG PET signature of 7 radiomics features demonstrated prognostic value in an independent cohort of pancreatic cancer patients ( n = 71) and identified patients with worse survival. Conclusion: Our cross-modal pipeline allows the development of PET scan signatures based on immunohistochemical analysis of markers of a particular biologic feature, here demonstrated on pancreatic cancer using intratumoral MCT4 expression levels to select [ 18 F]FDG PET radiomics features. This study demonstrated the potential of radiomics scores to noninvasively capture intratumoral marker heterogeneity and identify a subset of pancreatic ductal adenocarcinoma patients with a poor prognosis., (© 2024 by the Society of Nuclear Medicine and Molecular Imaging.)

Objectives: Achieving a consensus on a definition for different aspects of radiomics workflows to support their translation into clinical usage. Furthermore, to assess the perspective of experts on important challenges for a successful clinical workflow implementation., Materials and Methods: The consensus was achieved by a multi-stage process. Stage 1 comprised a definition screening, a retrospective analysis with semantic mapping of terms found in 22 workflow definitions, and the compilation of an initial baseline definition. Stages 2 and 3 consisted of a Delphi process with over 45 experts hailing from sites participating in the German Research Foundation (DFG) Priority Program 2177. Stage 2 aimed to achieve a broad consensus for a definition proposal, while stage 3 identified the importance of translational challenges., Results: Workflow definitions from 22 publications (published 2012-2020) were analyzed. Sixty-nine definition terms were extracted, mapped, and semantic ambiguities (e.g., homonymous and synonymous terms) were identified and resolved. The consensus definition was developed via a Delphi process. The final definition comprising seven phases and 37 aspects reached a high overall consensus (> 89% of experts "agree" or "strongly agree"). Two aspects reached no strong consensus. In addition, the Delphi process identified and characterized from the participating experts' perspective the ten most important challenges in radiomics workflows., Conclusion: To overcome semantic inconsistencies between existing definitions and offer a well-defined, broad, referenceable terminology, a consensus workflow definition for radiomics-based setups and a terms mapping to existing literature was compiled. Moreover, the most relevant challenges towards clinical application were characterized., Critical Relevance Statement: Lack of standardization represents one major obstacle to successful clinical translation of radiomics. Here, we report a consensus workflow definition on different aspects of radiomics studies and highlight important challenges to advance the clinical adoption of radiomics., Key Points: Published radiomics workflow terminologies are inconsistent, hindering standardization and translation. A consensus radiomics workflow definition proposal with high agreement was developed. Publicly available result resources for further exploitation by the scientific community., (© 2024. The Author(s).)

Chronic obstructive pulmonary disease (COPD) is one of the leading causes of death. Recent studies have underlined the importance of non-contrast-enhanced chest CT scans not only for emphysema progression quantification, but for correlation with clinical outcomes as well. As about 40 percent of the 300 million CT scans per year are contrast-enhanced, no proper emphysema quantification is available in a one-stop-shop approach for patients with known or newly diagnosed COPD. Since the introduction of spectral imaging (e.g., dual-energy CT scanners), it has been possible to create virtual non-contrast-enhanced images (VNC) from contrast-enhanced images, making it theoretically possible to offer proper COPD imaging despite contrast enhancing. This study is aimed towards investigating whether these VNC images are comparable to true non-contrast-enhanced images (TNC), thereby reducing the radiation exposure of patients and usage of resources in hospitals. In total, 100 COPD patients with two scans, one with (VNC) and one without contrast media (TNC), within 8 weeks or less obtained by a spectral CT using dual-layer technology, were included in this retrospective study. TNC and VNC were compared according to their voxel-density histograms. While the comparison showed significant differences in the low attenuated volumes (LAVs) of TNC and VNC regarding the emphysema threshold of -950 Houndsfield Units (HU), the 15th and 10th percentiles of the LAVs used as a proxy for pre-emphysema were comparable. Upon further investigation, the threshold-based LAVs (-950 HU) of TNC and VNC were comparable in patients with a water equivalent diameter (DW) below 270 mm. The study concludes that VNC imaging may be a viable option for assessing emphysema progression in COPD patients, particularly those with a normal body mass index (BMI). Further, pre-emphysema was generally comparable between TNC and VNC. This approach could potentially reduce radiation exposure and hospital resources by making additional TNC scans obsolete.

Background: Artificial intelligence (AI) models are increasingly used in the medical domain. However, as medical data is highly sensitive, special precautions to ensure its protection are required. The gold standard for privacy preservation is the introduction of differential privacy (DP) to model training. Prior work indicates that DP has negative implications on model accuracy and fairness, which are unacceptable in medicine and represent a main barrier to the widespread use of privacy-preserving techniques. In this work, we evaluated the effect of privacy-preserving training of AI models regarding accuracy and fairness compared to non-private training., Methods: We used two datasets: (1) A large dataset (N = 193,311) of high quality clinical chest radiographs, and (2) a dataset (N = 1625) of 3D abdominal computed tomography (CT) images, with the task of classifying the presence of pancreatic ductal adenocarcinoma (PDAC). Both were retrospectively collected and manually labeled by experienced radiologists. We then compared non-private deep convolutional neural networks (CNNs) and privacy-preserving (DP) models with respect to privacy-utility trade-offs measured as area under the receiver operating characteristic curve (AUROC), and privacy-fairness trade-offs, measured as Pearson's r or Statistical Parity Difference., Results: We find that, while the privacy-preserving training yields lower accuracy, it largely does not amplify discrimination against age, sex or co-morbidity. However, we find an indication that difficult diagnoses and subgroups suffer stronger performance hits in private training., Conclusions: Our study shows that - under the challenging realistic circumstances of a real-life clinical dataset - the privacy-preserving training of diagnostic deep learning models is possible with excellent diagnostic accuracy and fairness., (© 2024. The Author(s).)

When dealing with a newly emerging disease such as COVID-19, the impact of patient- and disease-specific factors (e.g., body weight or known co-morbidities) on the immediate course of the disease is largely unknown. An accurate prediction of the most likely individual disease progression can improve the planning of limited resources and finding the optimal treatment for patients. In the case of COVID-19, the need for intensive care unit (ICU) admission of pneumonia patients can often only be determined on short notice by acute indicators such as vital signs (e.g., breathing rate, blood oxygen levels), whereas statistical analysis and decision support systems that integrate all of the available data could enable an earlier prognosis. To this end, we propose a holistic, multimodal graph-based approach combining imaging and non-imaging information. Specifically, we introduce a multimodal similarity metric to build a population graph that shows a clustering of patients. For each patient in the graph, we extract radiomic features from a segmentation network that also serves as a latent image feature encoder. Together with clinical patient data like vital signs, demographics, and lab results, these modalities are combined into a multimodal representation of each patient. This feature extraction is trained end-to-end with an image-based Graph Attention Network to process the population graph and predict the COVID-19 patient outcomes: admission to ICU, need for ventilation, and mortality. To combine multiple modalities, radiomic features are extracted from chest CTs using a segmentation neural network. Results on a dataset collected in Klinikum rechts der Isar in Munich, Germany and the publicly available iCTCF dataset show that our approach outperforms single modality and non-graph baselines. Moreover, our clustering and graph attention increases understanding of the patient relationships within the population graph and provides insight into the network's decision-making process., (© 2023. The Author(s).)

Type 1 conventional dendritic cells (cDC1) can support T cell responses within tumors but whether this determines protective versus ineffective anti-cancer immunity is poorly understood. Here, we use imaging-based deep learning to identify intratumoral cDC1-CD8 + T cell clustering as a unique feature of protective anti-cancer immunity. These clusters form selectively in stromal tumor regions and constitute niches in which cDC1 activate TCF1 + stem-like CD8 + T cells. We identify a distinct population of immunostimulatory CCR7 neg cDC1 that produce CXCL9 to promote cluster formation and cross-present tumor antigens within these niches, which is required for intratumoral CD8 + T cell differentiation and expansion and promotes cancer immune control. Similarly, in human cancers, CCR7 neg cDC1 interact with CD8 + T cells in clusters and are associated with patient survival. Our findings reveal an intratumoral phase of the anti-cancer T cell response orchestrated by tumor-residing cDC1 that determines protective versus ineffective immunity and could be exploited for cancer therapy., Competing Interests: Declaration of interests The authors declare no competing interests., (Copyright © 2023 The Author(s). Published by Elsevier Inc. All rights reserved.)

Background: A quantitative assessment of pulmonary edema is important because the clinical severity can range from mild impairment to life threatening. A quantitative surrogate measure, although invasive, for pulmonary edema is the extravascular lung water index (EVLWI) extracted from the transpulmonary thermodilution (TPTD). Severity of edema from chest X-rays, to date is based on the subjective classification of radiologists. In this work, we use machine learning to quantitatively predict the severity of pulmonary edema from chest radiography., Methods: We retrospectively included 471 X-rays from 431 patients who underwent chest radiography and TPTD measurement within 24 h at our intensive care unit. The EVLWI extracted from the TPTD was used as a quantitative measure for pulmonary edema. We used a deep learning approach and binned the data into two, three, four and five classes increasing the resolution of the EVLWI prediction from the X-rays., Results: The accuracy, area under the receiver operating characteristic curve (AUROC) and Mathews correlation coefficient (MCC) in the binary classification models (EVLWI < 15, ≥ 15) were 0.93 (accuracy), 0.98 (AUROC) and 0.86(MCC). In the three multiclass models, the accuracy ranged between 0.90 and 0.95, the AUROC between 0.97 and 0.99 and the MCC between 0.86 and 0.92., Conclusion: Deep learning can quantify pulmonary edema as measured by EVLWI with high accuracy., (© 2023. The Author(s).)

Importance: Differentiating between malignant and benign etiology in large-bowel wall thickening on computed tomography (CT) images can be a challenging task. Artificial intelligence (AI) support systems can improve the diagnostic accuracy of radiologists, as shown for a variety of imaging tasks. Improvements in diagnostic performance, in particular the reduction of false-negative findings, may be useful in patient care., Objective: To develop and evaluate a deep learning algorithm able to differentiate colon carcinoma (CC) and acute diverticulitis (AD) on CT images and analyze the impact of the AI-support system in a reader study., Design, Setting, and Participants: In this diagnostic study, patients who underwent surgery between July 1, 2005, and October 1, 2020, for CC or AD were included. Three-dimensional (3-D) bounding boxes including the diseased bowel segment and surrounding mesentery were manually delineated and used to develop a 3-D convolutional neural network (CNN). A reader study with 10 observers of different experience levels was conducted. Readers were asked to classify the testing cohort under reading room conditions, first without and then with algorithmic support., Main Outcomes and Measures: To evaluate the diagnostic performance, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for all readers and reader groups with and without AI support. Metrics were compared using the McNemar test and relative and absolute predictive value comparisons., Results: A total of 585 patients (AD: n = 267, CC: n = 318; mean [SD] age, 63.2 [13.4] years; 341 men [58.3%]) were included. The 3-D CNN reached a sensitivity of 83.3% (95% CI, 70.0%-96.6%) and specificity of 86.6% (95% CI, 74.5%-98.8%) for the test set, compared with the mean reader sensitivity of 77.6% (95% CI, 72.9%-82.3%) and specificity of 81.6% (95% CI, 77.2%-86.1%). The combined group of readers improved significantly with AI support from a sensitivity of 77.6% to 85.6% (95% CI, 81.3%-89.3%; P < .001) and a specificity of 81.6% to 91.3% (95% CI, 88.1%-94.5%; P < .001). Artificial intelligence support significantly reduced the number of false-negative and false-positive findings (NPV from 78.5% to 86.4% and PPV from 80.9% to 90.8%; P < .001)., Conclusions and Relevance: The findings of this study suggest that a deep learning model able to distinguish CC and AD in CT images as a support system may significantly improve the diagnostic performance of radiologists, which may improve patient care.

Background: Pancreatic ductal adenocarcinoma (PDAC) lacks effective treatment options beyond chemotherapy. Although molecular subtypes such as classical and QM (quasi-mesenchymal)/basal-like with transcriptome-based distinct signatures have been identified, deduced therapeutic strategies and targets remain elusive. Gene expression data show enrichment of glycolytic genes in the more aggressive and therapy-resistant QM subtype. However, whether the glycolytic transcripts are translated into functional glycolysis that could further be explored for metabolic targeting in QM subtype is still not known., Methods: We used different patient-derived PDAC model systems (conventional and primary patient-derived cells, patient-derived xenografts (PDX), and patient samples) and performed transcriptional and functional metabolic analysis. These included RNAseq and Illumina HT12 bead array, in vitro Seahorse metabolic flux assays and metabolic drug targeting, and in vivo hyperpolarized [1- 13 C]pyruvate and [1- 13 C]lactate magnetic resonance spectroscopy (HP-MRS) in PDAC xenografts., Results: We found that glycolytic metabolic dependencies are not unambiguously functionally exposed in all QM PDACs. Metabolic analysis demonstrated functional metabolic heterogeneity in patient-derived primary cells and less so in conventional cell lines independent of molecular subtype. Importantly, we observed that the glycolytic product lactate is actively imported into the PDAC cells and used in mitochondrial oxidation in both classical and QM PDAC cells, although more actively in the QM cell lines. By using HP-MRS, we were able to noninvasively identify highly glycolytic PDAC xenografts by detecting the last glycolytic enzymatic step and prominent intra-tumoral [1- 13 C]pyruvate and [1- 13 C]lactate interconversion in vivo., Conclusion: Our study adds functional metabolic phenotyping to transcriptome-based analysis and proposes a functional approach to identify highly glycolytic PDACs as candidates for antimetabolic therapeutic avenues., (© 2022. The Author(s).)

The increasing prevalence of large-scale data collection in modern society represents a potential threat to individual privacy. Addressing this threat, for example through privacy-enhancing technologies (PETs), requires a rigorous definition of what exactly is being protected, that is, of privacy itself. In this work, we formulate an axiomatic definition of privacy based on quantifiable and irreducible information flows. Our definition synthesizes prior work from the domain of social science with a contemporary understanding of PETs such as differential privacy (DP). Our work highlights the fact that the inevitable difficulties of protecting privacy in practice are fundamentally information-theoretic. Moreover, it enables quantitative reasoning about PETs based on what they are protecting, thus fostering objective policy discourse about their societal implementation.

Invasive nonfunctioning (NF) pituitary neuroendocrine tumors (PitNETs) are non-resectable neoplasms associated with frequent relapses and significant comorbidities. As the current therapies of NF-PitNETs often fail, new therapeutic targets are needed. The observation that circulating angiopoietin-2 (ANGPT2) is elevated in patients with NF-PitNET and correlates with tumor aggressiveness prompted us to investigate the ANGPT2/TIE2 axis in NF-PitNETs in the GH3 PitNET cell line, primary human NF-PitNET cells, xenografts in zebrafish and mice, and in MENX rats, the only autochthonous NF-PitNET model. We show that PitNET cells express a functional TIE2 receptor and secrete bioactive ANGPT2, which promotes, besides angiogenesis, tumor cell growth in an autocrine and paracrine fashion. ANGPT2 stimulation of TIE2 in tumor cells activates downstream cell proliferation signals, as previously demonstrated in endothelial cells (ECs). Tie2 gene deletion blunts PitNETs growth in xenograft models, and pharmacological inhibition of Angpt2/Tie2 signaling antagonizes PitNETs in primary cell cultures, tumor xenografts in mice, and in MENX rats. Thus, the ANGPT2/TIE2 axis provides an exploitable therapeutic target in NF-PitNETs and possibly in other tumors expressing ANGPT2/TIE2. The ability of tumor cells to coopt angiogenic signals classically viewed as EC-specific expands our view on the microenvironmental cues that are essential for tumor progression., (© 2022 The Authors. Published under the terms of the CC BY 4.0 license.)

Long-term health consequences in survivors of severe COVID-19 remain unclear. Eighteen COVID-19 patients admitted to the intensive care unit at the University Hospital Rechts der Isar, Munich, Germany, between 14 March and 23 June 2020, were prospectively followed-up at a median of 36, 75.5, 122 and 222 days after discharge. The health-related quality of life (HrQoL) (36-item Short Form Health Survey and St. George's Respiratory Questionnaire, SGRQ), cardiopulmonary function, laboratory parameters and chest imaging were assessed longitudinally. The HrQoL assessment revealed a reduced physical functioning, as well as increased SGRQ impact and symptoms scores that all improved over time but remained markedly impaired compared to the reference groups. The median radiological severity scores significantly declined; persistent abnormalities were found in 33.3% of the patients on follow-up. A reduced diffusion capacity was the most common abnormal pulmonary function parameter. The length of hospitalization correlated with role limitations due to physical problems, the SGRQ symptom and the impact score. In conclusion, in survivors of severe COVID-19, the pulmonary function and symptoms improve over time, but impairments in their physical function and diffusion capacity can persist over months. Longer follow-up studies with larger cohorts will be necessary to comprehensively characterize long-term sequelae upon severe COVID-19 and to identify patients at risk.

With the advent of deep learning algorithms, fully automated radiological image analysis is within reach. In spine imaging, several atlas- and shape-based as well as deep learning segmentation algorithms have been proposed, allowing for subsequent automated analysis of morphology and pathology. The first "Large Scale Vertebrae Segmentation Challenge" (VerSe 2019) showed that these perform well on normal anatomy, but fail in variants not frequently present in the training dataset. Building on that experience, we report on the largely increased VerSe 2020 dataset and results from the second iteration of the VerSe challenge (MICCAI 2020, Lima, Peru). VerSe 2020 comprises annotated spine computed tomography (CT) images from 300 subjects with 4142 fully visualized and annotated vertebrae, collected across multiple centres from four different scanner manufacturers, enriched with cases that exhibit anatomical variants such as enumeration abnormalities (n = 77) and transitional vertebrae (n = 161). Metadata includes vertebral labelling information, voxel-level segmentation masks obtained with a human-machine hybrid algorithm and anatomical ratings, to enable the development and benchmarking of robust and accurate segmentation algorithms., (© 2021. The Author(s).)

Correlation of in vivo imaging to histomorphological pathology in animal models requires comparative interdisciplinary expertise of different fields of medicine. From the morphological point of view, there is an urgent need to improve histopathological evaluation in animal model-based research to expedite translation into clinical applications. While different other fields of translational science were standardized over the last years, little was done to improve the pipeline of experimental pathology to ensure reproducibility based on pathological expertise in experimental animal models with respect to defined guidelines and classifications. Additionally, longitudinal analyses of preclinical models often use a variety of imaging methods and much more attention should be drawn to enable for proper co-registration of in vivo imaging methods with the ex vivo morphological read-outs. Here we present the development of the Comparative Experimental Pathology (CEP) unit embedded in the Institute of Pathology of the Technical University of Munich during the Collaborative Research Center 824 (CRC824) funding period together with selected approaches of histomorphological techniques for correlation of in vivo imaging to morphomolecular pathology., (© 2021. The Author(s).)

The success of deep learning in recent years has arguably been driven by the availability of large datasets for training powerful predictive algorithms. In medical applications however, the sensitive nature of the data limits the collection and exchange of large-scale datasets. Privacy-preserving and collaborative learning systems can enable the successful application of machine learning in medicine. However, collaborative protocols such as federated learning require the frequent transfer of parameter updates over a network. To enable the deployment of such protocols to a wide range of systems with varying computational performance, efficient deep learning architectures for resource-constrained environments are required. Here we present MoNet, a small, highly optimized neural-network-based segmentation algorithm leveraging efficient multi-scale image features. MoNet is a shallow, U-Net-like architecture based on repeated, dilated convolutions with decreasing dilation rates. We apply and test our architecture on the challenging clinical tasks of pancreatic segmentation in computed tomography (CT) images as well as brain tumor segmentation in magnetic resonance imaging (MRI) data. We assess our model's segmentation performance and demonstrate that it provides performance on par with compared architectures while providing superior out-of-sample generalization performance, outperforming larger architectures on an independent validation set, while utilizing significantly fewer parameters. We furthermore confirm the suitability of our architecture for federated learning applications by demonstrating a substantial reduction in serialized model storage requirement as a surrogate for network data transfer. Finally, we evaluate MoNet's inference latency on the central processing unit (CPU) to determine its utility in environments without access to graphics processing units. Our implementation is publicly available as free and open-source software., Competing Interests: The authors have declared that no competing interests exist.

Background & Aims: The existence of different subtypes of pancreatic ductal adenocarcinoma (PDAC) and their correlation with patient outcome have shifted the emphasis on patient classification for better decision-making algorithms and personalized therapy. The contribution of mechanisms regulating the cancer stem cell (CSC) population in different subtypes remains unknown., Methods: Using RNA-seq, we identified B-cell CLL/lymphoma 3 (BCL3), an atypical nf-κb signaling member, as differing in pancreatic CSCs. To determine the biological consequences of BCL3 silencing in vivo and in vitro, we generated bcl3-deficient preclinical mouse models as well as murine cell lines and correlated our findings with human cell lines, PDX models, and 2 independent patient cohorts. We assessed the correlation of bcl3 expression pattern with clinical parameters and subtypes., Results: Bcl3 was significantly down-regulated in human CSCs. Recapitulating this phenotype in preclinical mouse models of PDAC via BCL3 genetic knockout enhanced tumor burden, metastasis, epithelial to mesenchymal transition, and reduced overall survival. Fluorescence-activated cell sorting analyses, together with oxygen consumption, sphere formation, and tumorigenicity assays, all indicated that BCL3 loss resulted in CSC compartment expansion promoting cellular dedifferentiation. Overexpression of BCL3 in human PDXs diminished tumor growth by significantly reducing the CSC population and promoting differentiation. Human PDACs with low BCL3 expression correlated with increased metastasis, and BCL3-negative tumors correlated with lower survival and nonclassical subtypes., Conclusions: We demonstrate that bcl3 impacts pancreatic carcinogenesis by restraining CSC expansion and by curtailing an aggressive and metastatic tumor burden in PDAC across species. Levels of BCL3 expression are a useful stratification marker for predicting subtype characterization in PDAC, thereby allowing for personalized therapeutic approaches., (Copyright © 2021 AGA Institute. Published by Elsevier Inc. All rights reserved.)

The successful training of deep learning models for diagnostic deployment in medical imaging applications requires large volumes of data. Such data cannot be procured without consideration for patient privacy, mandated both by legal regulations and ethical requirements of the medical profession. Differential privacy (DP) enables the provision of information-theoretic privacy guarantees to patients and can be implemented in the setting of deep neural network training through the differentially private stochastic gradient descent (DP-SGD) algorithm. We here present deepee, a free-and-open-source framework for differentially private deep learning for use with the PyTorch deep learning framework. Our framework is based on parallelised execution of neural network operations to obtain and modify the per-sample gradients. The process is efficiently abstracted via a data structure maintaining shared memory references to neural network weights to maintain memory efficiency. We furthermore offer specialised data loading procedures and privacy budget accounting based on the Gaussian Differential Privacy framework, as well as automated modification of the user-supplied neural network architectures to ensure DP-conformity of its layers. We benchmark our framework's computational performance against other open-source DP frameworks and evaluate its application on the paediatric pneumonia dataset, an image classification task and on the Medical Segmentation Decathlon Liver dataset in the task of medical image segmentation. We find that neural network training with rigorous privacy guarantees is possible while maintaining acceptable classification performance and excellent segmentation performance. Our framework compares favourably to related work with respect to memory consumption and computational performance. Our work presents an open-source software framework for differentially private deep learning, which we demonstrate in medical imaging analysis tasks. It serves to further the utilisation of privacy-enhancing techniques in medicine and beyond in order to assist researchers and practitioners in addressing the numerous outstanding challenges towards their widespread implementation.

Background: In the absence of PCR detection of SARS-CoV-2 RNA, accurate diagnosis of COVID-19 is challenging. Low-dose computed tomography (CT) detects pulmonary infiltrates with high sensitivity, but findings may be non-specific. This study assesses the diagnostic value of SARS-CoV-2 serology for patients with distinct CT features but negative PCR., Methods: IgM/IgG chemiluminescent immunoassay was performed for 107 patients with confirmed (group A: PCR + ; CT ±) and 46 patients with suspected (group B: repetitive PCR-; CT +) COVID-19, admitted to a German university hospital during the pandemic's first wave. A standardized, in-house CT classification of radiological signs of a viral pneumonia was used to assess the probability of COVID-19., Results: Seroconversion rates (SR) determined on day 5, 10, 15, 20 and 25 after symptom onset (SO) were 8%, 25%, 65%, 76% and 91% for group A, and 0%, 10%, 19%, 37% and 46% for group B, respectively; (p < 0.01). Compared to hospitalized patients with a non-complicated course (non-ICU patients), seroconversion tended to occur at lower frequency and delayed in patients on intensive care units. SR of patients with CT findings classified as high certainty for COVID-19 were 8%, 22%, 68%, 79% and 93% in group A, compared with 0%, 15%, 28%, 50% and 50% in group B (p < 0.01). SARS-CoV-2 serology established a definite diagnosis in 12/46 group B patients. In 88% (8/9) of patients with negative serology > 14 days after symptom onset (group B), clinico-radiological consensus reassessment revealed probable diagnoses other than COVID-19. Sensitivity of SARS-CoV-2 serology was superior to PCR > 17d after symptom onset., Conclusions: Approximately one-third of patients with distinct COVID-19 CT findings are tested negative for SARS-CoV-2 RNA by PCR rendering correct diagnosis difficult. Implementation of SARS-CoV-2 serology testing alongside current CT/PCR-based diagnostic algorithms improves discrimination between COVID-19-related and non-related pulmonary infiltrates in PCR negative patients. However, sensitivity of SARS-CoV-2 serology strongly depends on the time of testing and becomes superior to PCR after the 2 nd week following symptom onset.

Data privacy mechanisms are essential for rapidly scaling medical training databases to capture the heterogeneity of patient data distributions toward robust and generalizable machine learning systems. In the current COVID-19 pandemic, a major focus of artificial intelligence (AI) is interpreting chest CT, which can be readily used in the assessment and management of the disease. This paper demonstrates the feasibility of a federated learning method for detecting COVID-19 related CT abnormalities with external validation on patients from a multinational study. We recruited 132 patients from seven multinational different centers, with three internal hospitals from Hong Kong for training and testing, and four external, independent datasets from Mainland China and Germany, for validating model generalizability. We also conducted case studies on longitudinal scans for automated estimation of lesion burden for hospitalized COVID-19 patients. We explore the federated learning algorithms to develop a privacy-preserving AI model for COVID-19 medical image diagnosis with good generalization capability on unseen multinational datasets. Federated learning could provide an effective mechanism during pandemics to rapidly develop clinically useful AI across institutions and countries overcoming the burden of central aggregation of large amounts of sensitive data.

BACKGROUNDPancreatic cancer is one of the deadliest cancers, with low long-term survival rates. Despite recent advances in treatment, it is important to identify and screen high-risk individuals for cancer prevention. Familial pancreatic cancer (FPC) accounts for 4%-10% of pancreatic cancers. Several germline mutations are related to an increased risk and might offer screening and therapy options. In this study, we aimed to identity of a susceptibility gene in a family with FPC.METHODSWhole exome sequencing and PCR confirmation was performed on the surgical specimen and peripheral blood of an index patient and her sister in a family with high incidence of pancreatic cancer, to identify somatic and germline mutations associated with familial pancreatic cancer. Compartment-specific gene expression data and immunohistochemistry were also queried.RESULTSThe identical germline mutation of the PALLD gene (NM&#95;001166108.1:c.G154A:p.D52N) was detected in the index patient with pancreatic cancer and the tumor tissue of her sister. Whole genome sequencing showed similar somatic mutation patterns between the 2 sisters. Apart from the PALLD mutation, commonly mutated genes that characterize pancreatic ductal adenocarcinoma were found in both tumor samples. However, the 2 patients harbored different somatic KRAS mutations (G12D and G12V). Healthy siblings did not have the PALLD mutation, indicating a disease-specific impact. Compartment-specific gene expression data and IHC showed expression in cancer-associated fibroblasts (CAFs).CONCLUSIONWe identified a germline mutation of the palladin (PALLD) gene in 2 siblings in Europe, affected by familial pancreatic cancer, with a significant overexpression in CAFs, suggesting that stromal palladin could play a role in the development, maintenance, and/or progression of pancreatic cancer.FUNDINGDFG SFB 1321.

Hyperpolarized 13 C nuclear magnetic resonance spectroscopy can characterize in vivo tissue metabolism, including preclinical models of cancer and inflammatory disease. Broad bandwidth radiofrequency excitation is often paired with free induction decay readout for spectral separation, but quantification of low-signal downstream metabolites using this method can be impeded by spectral peak overlap or when frequency separation of the detected peaks exceeds the excitation bandwidth. In this work, alternating frequency narrow bandwidth (250 Hz) slice-selective excitation was used for 13 C spectroscopy at 7 T in a subcutaneous xenograft rat model of human pancreatic cancer (PSN1) to improve quantification while measuring the dynamics of injected hyperpolarized [1- 13 C]lactate and its metabolite [1- 13 C]pyruvate. This method does not require sophisticated pulse sequences or specialized radiofrequency and gradient pulses, but rather uses nominally spatially offset slices to produce alternating frequency excitation with simpler slice-selective radiofrequency pulses. Additionally, point-resolved spectroscopy was used to calibrate the 13 C frequency from the thermal proton signal in the target region. This excitation scheme isolates the small [1- 13 C]pyruvate peak from the similar-magnitude tail of the much larger injected [1- 13 C]lactate peak, facilitates quantification of the [1- 13 C]pyruvate signal, simplifies data processing, and could be employed for other substrates and preclinical models.

The in vivo assessment of tissue metabolism represents a novel strategy for the evaluation of oncologic disease. Hepatocellular carcinoma (HCC) is a high-prevalence, high-mortality tumor entity often discovered at a late stage. Recent evidence indicates that survival differences depend on metabolic alterations in tumor tissue, with particular focus on glucose metabolism and lactate production. Here, we present an in vivo imaging technique for metabolic tumor phenotyping in rat models of HCC. Endogenous HCC was induced in Wistar rats by oral diethyl-nitrosamine administration. Peak lactate-to-alanine signal ratios (L/A) were assessed with hyperpolarized magnetic resonance spectroscopic imaging (HPMRSI) after [1- 13 C]pyruvate injection. Cell lines were derived from a subset of primary tumors, re-implanted in nude rats, and assessed in vivo with dynamic hyperpolarized magnetic resonance spectroscopy (HPMRS) after [1- 13 C]pyruvate injection and kinetic modelling of pyruvate metabolism, taking into account systemic lactate production and recirculation. For ex vivo validation, enzyme activity and metabolite concentrations were spectroscopically quantified in cell and tumor tissue extracts. Mean peak L/A was higher in endogenous HCC compared to non-tumorous tissue. Dynamic HPMRS revealed higher pyruvate-to-lactate conversion rates (k pl ) and lactate signal in subcutaneous tumors derived from high L/A tumor cells, consistent with ex vivo measurements of higher lactate dehydrogenase (LDH) levels in these cells. In conclusion, HPMRS and HPMRSI reveal distinct tumor phenotypes corresponding to differences in glycolytic metabolism in HCC tumor tissue.

The differentiation of autoimmune pancreatitis (AIP) and pancreatic ductal adenocarcinoma (PDAC) poses a relevant diagnostic challenge and can lead to misdiagnosis and consequently poor patient outcome. Recent studies have shown that radiomics-based models can achieve high sensitivity and specificity in predicting both entities. However, radiomic features can only capture low level representations of the input image. In contrast, convolutional neural networks (CNNs) can learn and extract more complex representations which have been used for image classification to great success. In our retrospective observational study, we performed a deep learning-based feature extraction using CT-scans of both entities and compared the predictive value against traditional radiomic features. In total, 86 patients, 44 with AIP and 42 with PDACs, were analyzed. Whole pancreas segmentation was automatically performed on CT-scans during the portal venous phase. The segmentation masks were manually checked and corrected if necessary. In total, 1411 radiomic features were extracted using PyRadiomics and 256 features (deep features) were extracted using an intermediate layer of a convolutional neural network (CNN). After feature selection and normalization, an extremely randomized trees algorithm was trained and tested using a two-fold shuffle-split cross-validation with a test sample of 20% ( n = 18) to discriminate between AIP or PDAC. Feature maps were plotted and visual difference was noted. The machine learning (ML) model achieved a sensitivity, specificity, and ROC-AUC of 0.89 ± 0.11, 0.83 ± 0.06, and 0.90 ± 0.02 for the deep features and 0.72 ± 0.11, 0.78 ± 0.06, and 0.80 ± 0.01 for the radiomic features. Visualization of feature maps indicated different activation patterns for AIP and PDAC. We successfully trained a machine learning model using deep feature extraction from CT-images to differentiate between AIP and PDAC. In comparison to traditional radiomic features, deep features achieved a higher sensitivity, specificity, and ROC-AUC. Visualization of deep features could further improve the diagnostic accuracy of non-invasive differentiation of AIP and PDAC.

Evasion of programmed cell death represents a critical form of oncogene addiction in cancer cells. Understanding the molecular mechanisms underpinning cancer cell survival despite the oncogenic stress could provide a molecular basis for potential therapeutic interventions. Here we explore the role of pro-survival genes in cancer cell integrity during clonal evolution in non-small cell lung cancer (NSCLC). We identify gains of MCL-1 at high frequency in multiple independent NSCLC cohorts, occurring both clonally and subclonally. Clonal loss of functional TP53 is significantly associated with subclonal gains of MCL-1. In mice, tumour progression is delayed upon pharmacologic or genetic inhibition of MCL-1. These findings reveal that MCL-1 gains occur with high frequency in lung adenocarcinoma and can be targeted therapeutically.

One of the major challenges in using pancreatic cancer patient-derived organoids (PDOs) in precision oncology is the time from biopsy to functional characterization. This is particularly true for endoscopic ultrasound-guided fine-needle aspiration biopsies, typically resulting in specimens with limited tumor cell yield. Here, we tested conditioned media of individual PDOs for cell-free DNA to detect driver mutations already early on during the expansion process to accelerate the genetic characterization of PDOs as well as subsequent functional testing. Importantly, genetic alterations detected in the PDO supernatant, collected as early as 72 hours after biopsy, recapitulate the mutational profile of the primary tumor, indicating suitability of this approach to subject PDOs to drug testing in a reduced time frame. In addition, we demonstrated that this workflow was practicable, even in patients for whom the amount of tumor material was not sufficient for molecular characterization by established means. Together, our findings demonstrate that generating PDOs from very limited biopsy material permits molecular profiling and drug testing. With our approach, this can be achieved in a rapid and feasible fashion with broad implications in clinical practice.

Background: To develop a supervised machine learning (ML) algorithm predicting above- versus below-median overall survival (OS) from diffusion-weighted imaging-derived radiomic features in patients with pancreatic ductal adenocarcinoma (PDAC)., Methods: One hundred two patients with histopathologically proven PDAC were retrospectively assessed as training cohort, and 30 prospectively accrued and retrospectively enrolled patients served as independent validation cohort (IVC). Tumors were segmented on preoperative apparent diffusion coefficient (ADC) maps, and radiomic features were extracted. A random forest ML algorithm was fit to the training cohort and tested in the IVC. Histopathological subtype of tumor samples was assessed by immunohistochemistry in 21 IVC patients. Individual radiomic feature importance was evaluated by assessment of tree node Gini impurity decrease and recursive feature elimination. Fisher's exact test, 95% confidence intervals (CI), and receiver operating characteristic area under the curve (ROC-AUC) were used., Results: The ML algorithm achieved 87% sensitivity (95% IC 67.3-92.7), 80% specificity (95% CI 74.0-86.7), and ROC-AUC 90% for the prediction of above- versus below-median OS in the IVC. Heterogeneity-related features were highly ranked by the model. Of the 21 patients with determined histopathological subtype, 8/9 patients predicted to experience below-median OS exhibited the quasi-mesenchymal subtype, whilst 11/12 patients predicted to experience above-median OS exhibited a non-quasi-mesenchymal subtype (p < 0.001)., Conclusion: ML application to ADC radiomics allowed OS prediction with a high diagnostic accuracy in an IVC. The high overlap of clinically relevant histopathological subtypes with model predictions underlines the potential of quantitative imaging in PDAC pre-operative subtyping and prognosis.