A5041401789- Cancer Genomics and Diagnostics
- Data Management and Algorithms
- Data Mining Algorithms and Applications
- Gene expression and cancer classification
- Advanced Database Systems and Queries
- Algorithms and Data Compression
- Radiomics and Machine Learning in Medical Imaging
- Genomic variations and chromosomal abnormalities
- Business Process Modeling and Analysis
- HER2/EGFR in Cancer Research
- Metabolomics and Mass Spectrometry Studies
- Bioinformatics and Genomic Networks
- Time Series Analysis and Forecasting
- Data Stream Mining Techniques
- Electronic Health Records Systems
- Multiple Sclerosis Research Studies
- Complex Network Analysis Techniques
- Image Retrieval and Classification Techniques
- Imbalanced Data Classification Techniques
- Opinion Dynamics and Social Influence
- Advanced Clustering Algorithms Research
- Machine Learning and Data Classification
- Peer-to-Peer Network Technologies
- Atherosclerosis and Cardiovascular Diseases
- Network Security and Intrusion Detection
Poznań University of Technology
2011-2025
Science Museum of Minnesota
2022
Abstract There is a certain belief among data science researchers and enthusiasts alike that clustering can be used to improve classification quality. Insofar as this fairly uncontroversial, it also very general therefore produces lot of confusion around the subject. are many ways using in obviously cannot always quality predictions, so question arises, which scenarios exactly does help? Since we were unable find rigorous study addressing question, paper, try shed some light on concept for...
Abstract Background Ulcerative colitis (UC) is characterized by persistent colonic mucosal inflammation. Despite available treatments, most patients experience only temporary remission. The lack of universally effective treatments results from our incomplete understanding UC pathophysiology. While many gene expression studies have provided valuable insights into mechanisms, the growing availability transcriptomics datasets offers a unique opportunity for large-scale, cross-validating...
Background The mechanism of cognitive dysfunction in systemic lupus erythematosus (SLE) is still not fully understood. Even though many SLE patients present some neurological dysfunction, including various deficits, neither a specific pattern nor structural changes associated with impairment have been established. Moreover, although prevalent and bothersome, deficits included the most recent diagnostic criteria. Purpose aim this study was to determine relationship between presence white...
Now that the use of XML is prevalent, methods for mining semi-structured documents have become even more important. In particular, one areas could greatly benefit from in-depth analysis XML's nature cluster analysis. Most clustering approaches developed so far employ pairwise similarity measures. this paper, we study algorithms, which patterns to without need comparisons. We investigate shortcomings existing and establish a new pattern-based framework called XPattern, tries address these...
In this paper we tackle the issue of exchanging and integrating medical information originating from different health care systems. We propose a solution - Healthcare Integration Platform (HIP) which utilises some concepts contained in IHE profiles combined with existing EHR standards order to maintain high level interoperability. confirm value our by presenting working prototype based on concepts. The relies Service Oriented Architecture paradigm using RESTful web services.
Abstract Feature extraction is the key to a successfully trained classifier. Although many automatic methods exist for traditional data, other data types (e.g., sequences, graphs) usually require dedicated approaches. In this paper, we study universal feature method based on distance from reference points. First, formalize process and provide an instantiation network centrality. To reliably select best points, introduce notion of θ -neighborhood which allows us navigate topography fully...
<title>Abstract</title> In this paper, we analyze the edit-distance-based approach to classification of sequences sets. Our goal is push edit distance measure its limits see just how weak a signal it can detect when applied It thorough experimental study exploring various aspects in isolation. To achieve this, needed precise control over characteristics data. That why also propose flexible dataset generator capable controlling main properties sets model, which make publicly available as an...
e21077 Background: The appearance of immune checkpoint inhibitors (ICIs) has revolutionized treatment strategies in advanced NSCLC. clinical efficacy and indications for ICIs are currently estimated using the immunohistochemistry (IHC) profile programmed death-ligand 1 (PD-L1) protein. Recent findings indicate that tumor mutational burden (TMB) is a unique feature may improve response prediction to across multiple cancer types [1]. Nevertheless, only minority NSCLC patients will benefit from...
Whole-genome sequencing has revolutionized biosciences by providing tools for constructing complete DNA sequences of individuals. With entire genomes at hand, scientists can pinpoint fragments responsible oncogenesis and predict patient responses to cancer treatments. Machine learning plays a paramount role in this process. However, the sheer volume whole-genome data makes it difficult encode characteristics genomic variants as features algorithms.In article, we propose three feature...
Human epidermal growth factor receptor 2 (HER2) protein overexpression is one of the most significant biomarkers for breast cancer diagnostics, treatment prediction, and prognostics. The high accessibility HER2 inhibitors in routine clinical practice directly translates into diagnostic need precise robust marker identification. Even though multigene next-generation sequencing methodologies have slowly taken over field single-biomarker molecular tests, copy number alterations such as...
Abstract Motivation Whole-genome sequencing has revolutionized biosciences by providing tools for constructing complete DNA sequences of individuals. With entire genomes at hand, scientists can pinpoint fragments responsible different cancers and predict patient responses to cancer treatments. However, the sheer volume whole-genome data makes it difficult encode characteristics genomic variants as features machine learning algorithms. Results We present three feature extraction methods that...
The wealth of data being gathered about humans and their surroundings drives new machine learning applications in various fields. Consequently, more often, classifiers are trained using not only numerical but also complex objects. For example, multi-omics analyses attempt to combine descriptions with distributions, time series data, discrete sequences, graphs. Such integration from different domains requires either omitting some the creating separate models for formats, or simplifying adhere...
Abstract The HER2 protein overexpression is one of the most significant biomarkers for breast cancer diagnostics, prediction, and prognostics. availability HER2-inhibitors in routine clinical practice directly translates into diagnostic need precise robust marker identification. At brink genomic era, multigene next-generation sequencing methodologies slowly take over field single-biomarker molecular cytogenetic tests. However, copy number alterations such as amplification HER2-coding ERBB2...