arXiv subjects

Tim Kacprowski

Publications and source records attributed to Tim Kacprowski.

5 recordsLinked to original sources

arXiv2026

Configurable Semantic Chunking for Biomedical Information Extraction in Retrieval-Augmented Generation

BioMedRAG introduced retrieval-augmented generation with a learned chunk scorer for biomedical information extraction. However, it relies on fixed-size chunking which can fragment semantic evidence. We propose a configurable semantic chunking framework that addresses this limitation by combining entity-preserving windows, trigger-centered chunking, proposition-first extraction, tiered trigger prioritization, and hierarchical relation resolution. The framework integrates with BioMedRAG by replacing only the chunk construction stage while preserving the embedding model, learned chunk scorer, generator, and evaluation protocol. We evaluate the framework on biomedical relation extraction benchmarks (GM-CIHT, DDI, ChemProt) and adverse event classification (ADE). On GM-CIHT, the full hybrid configuration achieves 82.6% F1, improving over the fixed-size baseline (74.2% F1) by 8.4 points under our experimental setup. Cross-dataset analysis shows that semantic chunking improves extraction datasets with explicit relation cues, such as GM-CIHT and DDI, while fixed chunking remains competitive or stronger for dense biochemical extraction and binary classification settings such as ChemProt and ADE. By externalizing chunking logic into configuration files, the framework provides an interpretable and adaptable alternative to rigid fixed-size chunking for biomedical RAG pipelines.

cs.CL↗

arXiv2023

Measurement of Individual Alteration in Perioperative ECGs During Elective Percutaneous Coronary Intervention

The increasing availability of wearable electrocardiography (ECG) devices enables the continuous monitoring of individual ECG alterations. This could be beneficial for patients suffering from acute ischemia but with non-standard ECG findings that do not fit to the subject-independent and absolute thresholds defined in clinical guidelines. In this work, we evaluate the inter-patient magnitude of individual ECG alterations during ischemia. The freely available STAFF III database provides 12-lead ECG recordings of patients before, during, and after elective percutaneous coronary intervention(PCI), where a coronary vessel is widened with a balloon inflation. We compute individual alterations of ST-interval and T-wave amplitudes w.r.t. QRS amplitude over time for each patient and lead. We demonstrate that determining relative ST-interval/T-wave amplitudes and deriving individual alterations over time is feasible in standard and non-standard ECG recordings. To demonstrate clinical relevance, we use the features for differentiating N=54 STAFF III patients with atherosclerotic plaque in either the right coronary artery (RCA) or left ascending artery (LAD). Results show significant differences in 5 leads for ST-interval alterations and 3 leads for T-wave alterations, which are also suggested by clinical guidelines for ischemia detection. Clinical relevance- Assessing individual alterations over time could eventually close the gap in ECG evaluation of patients presenting with pre-existing heart conditions and non-standard ECGs.

q-bio.TO↗

arXiv2021

The FeatureCloud AI Store for Federated Learning in Biomedicine and Beyond

Machine Learning (ML) and Artificial Intelligence (AI) have shown promising results in many areas and are driven by the increasing amount of available data. However, this data is often distributed across different institutions and cannot be shared due to privacy concerns. Privacy-preserving methods, such as Federated Learning (FL), allow for training ML models without sharing sensitive data, but their implementation is time-consuming and requires advanced programming skills. Here, we present the FeatureCloud AI Store for FL as an all-in-one platform for biomedical research and other applications. It removes large parts of this complexity for developers and end-users by providing an extensible AI Store with a collection of ready-to-use apps. We show that the federated apps produce similar results to centralized ML, scale well for a typical number of collaborators and can be combined with Secure Multiparty Computation (SMPC), thereby making FL algorithms safely and easily applicable in biomedical and clinical environments.

cs.LG↗

arXiv2020

Privacy-preserving Artificial Intelligence Techniques in Biomedicine

Artificial intelligence (AI) has been successfully applied in numerous scientific domains. In biomedicine, AI has already shown tremendous potential, e.g. in the interpretation of next-generation sequencing data and in the design of clinical decision support systems. However, training an AI model on sensitive data raises concerns about the privacy of individual participants. For example, summary statistics of a genome-wide association study can be used to determine the presence or absence of an individual in a given dataset. This considerable privacy risk has led to restrictions in accessing genomic and other biomedical data, which is detrimental for collaborative research and impedes scientific progress. Hence, there has been a substantial effort to develop AI methods that can learn from sensitive data while protecting individuals' privacy. This paper provides a structured overview of recent advances in privacy-preserving AI techniques in biomedicine. It places the most important state-of-the-art approaches within a unified taxonomy and discusses their strengths, limitations, and open problems. As the most promising direction, we suggest combining federated machine learning as a more scalable approach with other additional privacy preserving techniques. This would allow to merge the advantages to provide privacy guarantees in a distributed way for biomedical applications. Nonetheless, more research is necessary as hybrid approaches pose new challenges such as additional network or computation overhead.

cs.CR↗

arXiv2020

Exploring the SARS-CoV-2 virus-host-drug interactome for drug repurposing

Coronavirus Disease-2019 (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. It was first identified in Wuhan, China, and has since spread causing a global pandemic. Various studies have been performed to understand the molecular mechanisms of viral infection for predicting drug repurposing candidates. However, such information is spread across many publications and it is very time-consuming to access, integrate, explore, and exploit. We developed CoVex, the first interactive online platform for SARS-CoV-2 and SARS-CoV-1 host interactome exploration and drug (target) identification. CoVex integrates 1) experimentally validated virus-human protein interactions, 2) human protein-protein interactions and 3) drug-target interactions. The web interface allows user-friendly visual exploration of the virus-host interactome and implements systems medicine algorithms for network-based prediction of drugs. Thus, CoVex is an important resource, not only to understand the molecular mechanisms involved in SARS-CoV-2 and SARS-CoV-1 pathogenicity, but also in clinical research for the identification and prioritization of candidate therapeutics. We apply CoVex to investigate recent hypotheses on a systems biology level and to systematically explore the molecular mechanisms driving the virus life cycle. Furthermore, we extract and discuss drug repurposing candidates involved in these mechanisms. CoVex renders COVID-19 drug research systems-medicine-ready by giving the scientific community direct access to network medicine algorithms integrating virus-host-drug interactions. It is available at https://exbio.wzw.tum.de/covex/.

q-bio.MN↗