Ischaemic heart disease is among the most frequent causes of death. Early detection of myocardial pathologies can increase the benefit of therapy and reduce the number of lethal cases. Presence of myocardial scar is an indicator for developing ischaemic heart disease and can be detected with high diagnostic precision by magnetic resonance imaging. However, magnetic resonance imaging scanners are expensive and of limited availability. It is known that presence of myocardial scar has an impact on the well-established, reasonably low cost, and almost ubiquitously available electrocardiogram. However, this impact is non-specific and often hard to detect by a physician. We present an artificial intelligence based approach — namely a deep learning model — for the prediction of myocardial scar based on an electrocardiogram and additional clinical parameters. The model was trained and evaluated by applying 6-fold cross-validation to a dataset of 12-lead electrocardiogram time series together with clinical parameters. The proposed model for predicting the presence of scar tissue achieved an area under the curve score, sensitivity, specificity, and accuracy of 0.89, 70.0, 84.3, and 78.0%, respectively. This promisingly high diagnostic precision of our electrocardiogram-based deep learning models for myocardial scar detection may support a novel, comprehensible screening method.

In pathology, tissue images are evaluated using a light microscope, relying on the expertise and experience of pathologists. There is a great need for computational methods to quantify and standardize histological observations. Computational quantification methods become more and more essential to evaluate tissue images. In particular, the distribution of tumor cells and their microenvironment are of special interest. Here, we systematically investigated tumor cell properties and their spatial neighborhood relations by a new application of statistical analysis to whole slide images of Hodgkin lymphoma, a tumor arising in lymph nodes, and inflammation of lymph nodes called lymphadenitis. We considered properties of more than 400, 000 immunohistochemically stained, CD30-positive cells in 35 whole slide images of tissue sections from subtypes of the classical Hodgkin lymphoma, nodular sclerosis and mixed cellularity, as well as from lymphadenitis. We found that cells of specific morphology exhibited significantly favored and unfavored spatial neighborhood relations of cells in dependence of their morphology. This information is important to evaluate differences between Hodgkin lymph nodes infiltrated by tumor cells (Hodgkin lymphoma) and inflamed lymph nodes, concerning the neighborhood relations of cells and the sizes of cells. The quantification of neighborhood relations revealed new insights of relations of CD30-positive cells in different diagnosis cases. The approach is general and can easily be applied to whole slide image analysis of other tumor types. In pathology, histological diagnosis is still challenging, in particular, for tumor diseases. Pathologists diagnose the disease and its stage of development on the basis of evaluation and interpretation of images of tissue sections. The quantification of experimental data to support decisions of diagnosis and prognosis, applying bioinformatics methods, is an important issue. Here, we introduce a new, general approach to analyze tissue images of tumor and non-tumor patients and to evaluate the distribution of tumor cells in the tissue. Moreover, we consider neighborhood relations between immunostained cells of different cell morphology. We focus on a special type of lymph node tumor, the Hodgkin lymphoma, exploring the two main types of the classical Hodgkin lymphoma, the nodular sclerosis and the mixed cellularity, and the non-tumor case, the lymphadenitis, representing an inflammation of the lymph node. We considered more than 400, 000 cells immunohistochemically stained with CD30 in 35 whole slide images of tissue sections. We found that cells of specific morphology exhibited significant relations to cells of certain morphology as spatial nearest neighbor. We could show different neighborhood patterns of CD30-positive cells between tumor and non-tumor. The approach is general and can easily be applied to other tumor types.

The paper describes a mathematical model of the molecular switches of cell survival, apoptosis, and necroptosis in cellular signaling pathways initiated by tumor necrosis factor 1. Based on experimental findings in the literature, we constructed a Petri net model based on detailed molecular reactions of the molecular players, protein complexes, post-translational modifications, and cross talk. The model comprises 118 biochemical entities, 130 reactions, and 299 edges. We verified the model by evaluating invariant properties of the system at steady state and by in silico knockout analysis. Applying Petri net analysis techniques, we found 279 pathways, which describe signal flows from receptor activation to cellular response, representing the combinatorial diversity of functional pathways.120 pathways steered the cell to survival, whereas 58 and 35 pathways led to apoptosis and necroptosis, respectively. For 65 pathways, the triggered response was not deterministic and led to multiple possible outcomes. We investigated the in silico knockout behavior and identified important checkpoints of the TNFR1 signaling pathway in terms of ubiquitination within complex I and the gene expression dependent on NF-κB, which controls the caspase activity in complex II and apoptosis induction. Despite not knowing enough kinetic data of sufficient quality, we estimated system’s dynamics using a discrete, semi-quantitative Petri net model.

Summary: isiKnock is a new software that automatically conducts in silico knockouts for mathematical models of signaling pathways. The software allows for the prediction of the behavior of biological systems after single or multiple knockout. The implemented algorithm applies transition invariants and the novel concept of Manatee invariants. A knockout matrix visualizes the results. The tool enables the analysis of dependencies, for example, in signal flows from the receptor activation to the cell response at steady state. Availability and implementation: isiKnock is an open-source tool, freely available at http://www.bioinformatik.uni-frankfurt.de/tools/isiKnock/. It requires at least Java 8 and runs under Microsoft Windows, Linux, and Mac OS.

Evaluating explanations is an essential task in the field of eXplainable Artificial Intelligence (XAI) for time series. One approach to evaluate explanations is using ground-truth explanations that describe which features are expected to be relevant. However, defining such ground truth is particularly challenging for real-world time series classification tasks, where the temporal and spectral regions responsible for a prediction are rarely known. While some synthetic datasets provide ground-truth explanations, they cover only a limited range of time series properties, such as simple temporal patterns, single-label classification, and non-overlapping explanations. Existing datasets do not address multi-label time series classification, leaving this setting without a dedicated benchmark.To address this gap, we present a novel synthetic multi-label time series classification dataset designed to support the evaluation of XAI methods using ground-truth explanations. This dataset provides a controlled benchmark setting, in which multiple labels and their corresponding discriminative patterns can co-occur within a single time series sample. Discriminative patterns are located in both the time and frequency domains, enabling the evaluation of XAI methods across time, frequency, and time-frequency domains.The dataset contains 4000 univariate time series samples of 500 time steps each. Each sample contains between zero and three possible class labels defined by the presence of class-specific frequency windows. The three class labels may co-occur within a sample, resulting in eight possible combinations, each represented equally across 500 samples. In addition, the samples contain Gaussian noise and non-discriminative frequency components as distractors to simulate irrelevant signal components. Co-occurring class-specific frequency features overlap by 25% to 50% of the frequency windows, allowing the evaluation of explanation methods under feature overlap.In addition to the dataset, we provide trained models covering common time series classification architectures in order to support benchmarking and reproducible evaluation of XAI methods. The dataset and trained models are available on Zenodo, and the code used to generate the data and to train the models can be accessed via GitHub.