LSLudomir Stefańczyk

Ludomir Stefańczyk

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Ludomir Stefańczyk has a total of 152 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Frugal Learning Methods for Kidney Segmentation in Non-Contrast MRI

    by Jan Podlaszewski, Artur Klepaczko, Ludomir Stefańczyk, Marcin Majos

    Background/Objectives: Chronic kidney disease is a growing global health concern, necessitating effective tools for early detection and monitoring. While non-contrast T1-weighted magnetic resonance imaging offers a non-invasive means to assess kidney morphology, robust automated segmentation remains challenging due to limited annotated data, high inter-patient variability, and low signal-to-noise ratios. Methods: In this study, we address these obstacles by developing and evaluating a series of frugal learning methodologies for kidney segmentation in non-contrast MRI. Building upon the U-Net architecture, we aim to maximize segmentation accuracy despite scarce labeled data. Our experimental framework leverages three diverse datasets to evaluate performance-boosting strategies such as transfer learning: a clinically relevant local cohort (the Barlicki dataset) as the primary target domain and two auxiliary public datasets (AMOS22 and AbdomenCT). Utilizing these data streams, we systematically compare seven frugal learning strategies incorporating data augmentation, semi-supervised learning, and weak supervision against a fully supervised baseline. Results: The results demonstrate that frugal learning methods enable accurate and reliable kidney segmentation while substantially reducing the need for manual annotations. The best-performing semi-supervised and transfer learning approaches achieved a Dice similarity coefficient of 0.89, which was only moderately lower than that of the fully supervised model (Dice = 0.92). Conclusions: This work highlights the potential of data-efficient deep learning techniques to accelerate the adoption of automated kidney segmentation in clinical workflows, particularly in settings where annotated medical images are limited.

  • Open Access

    Texture Analysis of Multiparametric Kidney MRI–A Non-Invasive Approach to Chronic Kidney Disease State

    by Marcin Majos, Artur Klepaczko, Katarzyna Szychowska, Ludomir Stefanczyk, Ilona Kurnatowska

    Introduction: The most common indication of chronic kidney disease (CKD) is a sudden deterioration of renal function. At that point, the only reliable test determining disease progression is the histopathological evaluation of material obtained during kidney biopsy. Although biopsy is currently the gold standard of diagnosis, the procedure is associated with a number of complications that further burden kidney function. The aim of this work is to evaluate the potential of multiparametric magnetic resonance image (MRI) in the assessment of the renal parenchyma state. Material and Methods: The study is based on kidney MRI examinations of healthy volunteers (group 1, n = 11) and patients suffering from CKD in the active phase (group 2, n = 29) or the non-active phase (group 3, n = 14), as confirmed by kidney biopsy. The T1-weighted images, T2-weighted images and DWI images were extracted and used to create algorithms, based on Support Vector Machine and Sequential Floating Forward Search, aimed to differentiate defined groups. Results: The final algorithms obtained the following metrics: for T1-weighted images, analysis–precision 91.5%, sensitivity 89.6%, specificity 94.8%; for T2-weighted images, analysis–precision 80.2%, sensitivity 79.0%, specificity 89.5%; for DWI images, analysis–precision 82.3%, sensitivity 82.1%, specificity 91.0%; and for multiparametric MRIs, analysis–precision 90.9%, sensitivity 90.4%, specificity 95.2%. Conclusions: Texture analysis of multiparametric MRI seems to have diagnostic potential in CKD and its development may reduce the frequency of invasive procedures such as kidney biopsy.

  • Open Access

    Resting-State vs. Task-Based Functional Magnetic Resonance Imaging in Neurosurgical Planning: A Narrative Review of Clinical Applications

    by Maurycy Rakowski, Natalia Anna Koc, Anna DęBska, Bartosz Szmyd, Agata Zawadzka, Karol Zaczkowski, Małgorzata Podstawka, Dagmara Wilmańska, Adam Dobek, Ludomir Stefańczyk, et al.

    Background: Accurate presurgical localization of eloquent cortex and subcortical pathways is essential in neurosurgery, guiding the balance between maximal safe resection and preservation of neurological function. This narrative review compares the clinical utility of task-based functional magnetic resonance imaging (tb-fMRI) and resting-state functional magnetic resonance imaging (rs-fMRI) in neurosurgical populations, with emphasis on brain tumors and epilepsy. Methods: This narrative review was based on a non-systematic literature search of PubMed/MEDLINE, Scopus, Web of Science, and Google Scholar from database inception to March 2026. The review focused on tb-fMRI and rs-fMRI for presurgical functional mapping in neurosurgical populations, including clinical utility, feasibility, validation, limitations, and workflow integration. Results: Tb-fMRI remains the most established noninvasive modality for motor and language mapping and language lateralization because of its task-specific activation maps and established role in clinical workflows. However, its use is limited by dependence on patient cooperation, task performance, and intact neurovascular coupling; thus, aphasia, cognitive impairment, fatigue, paresis, pediatric age, sedation, and tumor-related neurovascular uncoupling may render tb-fMRI inconclusive or misleading. Rs-fMRI offers a task-free alternative based on intrinsic functional connectivity, enabling simultaneous mapping of multiple resting-state networks from a single acquisition and providing particular value in non-cooperative, cognitively impaired, aphasic, pediatric, or sedated patients. Evidence indicates that rs-fMRI is most robust for sensorimotor mapping, with reported agreement with tb-fMRI and intraoperative direct electrical stimulation, whereas language mapping remains less consistent and more dependent on analytical methodology. Neither modality replaces intraoperative stimulation, which remains the reference standard. Conclusions: Current evidence supports a multimodal presurgical strategy in which tb-fMRI is used first-line in cooperative patients; rs-fMRI is added when task-based mapping is limited or infeasible, and both are interpreted alongside tractography, neuronavigation, and intraoperative mapping.

  • Open Access

    Kidney MRI Texture Analysis—A Universal Assessment of Kidney State and Function?

    by Marcin Majos, Artur Klepaczko, Katarzyna Szychowska, Weronika Banasik, Ludomir Stefanczyk, Ilona Kurnatowska

    Introduction: Currently, chronic kidney disease (CKD) is detected based on glomerular filtration rate (GFR), proteinuria levels or kidney biopsy. However, the development of MRI techniques and AI algorithms gives hope to the assessment of CKD activity and kidney function with profound MRI image analysis. Methods: MRI images from healthy volunteers with no history of CKD were compared with those from CKD patients who had undergone both kidney MRI and kidney biopsy; the latter group was also divided into two subgroups based on CKD histopathological activity. Patients from both groups were scanned using either a 1.5 T or 3 T MRI scanner following sequential allocation (nine healthy controls and 28 CKD patients and 11 healthy volunteers and 43 CKD patients respectively for each scanner). Results: The final algorithm based on T1-weighted, T2-weighted and DWI images was able to distinguish patients with sensitivity ranging 77.78–87.50%, specificity 86.67–94.12% and precision 77.78–87.50%. Features of T1-weighted images and of T2-weighted images were found to correlate strongly with GFR with coefficients ranging from −0.5922 to −0.7090 and from 0.6126 to 0.6380, respectively. Conclusions: MRI image texture analysis may be suitable for assessing CKD activity, irrespective of the type of MRI scanner used. Furthermore, MRI image texture features correlate with eGFR values.

  • Reply to “Methodological pitfalls in MRI‑based quantification of post‑biopsy perirenal hematomas”

    by Piotr Białek, Weronika Banasik, Katarzyna Szychowska, Adam Dobek, Michał Żuberek, Krzysztof Falenta, Ilona Kurnatowska, Ludomir Stefańczyk

  • Open Access

    Endoscopic Ultrasound-Guided Versus Percutaneous Transhepatic Biliary Drainage After Failed Endoscopic Retrograde Cholangiopancreatography in Malignant Biliary Obstruction: A Single-Center Retrospective Cohort

    by Wojciech Ciesielski, Łukasz Durko, Ludomir Stefańczyk, Adam Dobek, Anna Bulicz, Amelia Wojnicka, Zuzanna Sosnowska, Agata Grochowska, Janusz Strzelczyk, Piotr Hogendorf, et al.

    Background: After a failed endoscopic retrograde cholangiopancreatography (ERCP) for malignant biliary obstruction (MBO), second-line drainage is performed with endoscopic ultrasound-guided biliary drainage (EUS-BD) or percutaneous transhepatic biliary drainage (PTBD). We compared their effectiveness, safety, and short-term survival. Methods: We conducted a single-center retrospective cohort of 101 adults with MBO after they had experienced a failed ERCP (EUS-BD n = 37; PTBD n = 64). Allocation was non-randomized and driven by operational availability. Baseline laboratory tests (complete blood count, platelets, and C-reactive protein) and derived indices (neutrophil-to-lymphocyte ratio [NLR], platelet-to-lymphocyte ratio [PLR], lymphocyte-to-monocyte ratio [LMR], systemic immune-inflammation index [SII], systemic inflammation response index [SIRI], neutrophil-to-platelet score [NPS], and lymphocyte-to-CRP ratio [LCR]) were compared. Outcomes that were a technical success include: an early biochemical response (bilirubin reduction), complications (Clavien–Dindo), length of stay (LOS), and overall survival (OS). Between-group comparisons used the two-sided Mann–Whitney U test (continuous) and Fisher’s exact (binary) test. Survival was assessed by the Kaplan–Meier estimator using log-rank testing. To address later adoption of EUS-BD, we also estimated a restricted mean survival time of 180 days (RMST_0–180) with 95% confidence intervals (CIs). Results: Baseline inflammatory markers and composite indices were similar; baseline total bilirubin was higher in PTBD. The technical success was 100% in both groups. Early biochemical response was 86.5% after EUS-BD vs. 78.1% after PTBD (p = 0.43). Any complication occurred in 29.7% vs. 12.5% (p = 0.04); major complications (Clavien–Dindo ≥ III) occurred in 10.8% vs. 0% (p = 0.02), respectively; and the LOS did not differ (p = 0.21). OS favored EUS-BD (median 143 vs. 54 days and log-rank p = 0.012). RMST_0–180 was 111.1 days for EUS-BD vs. 71.4 days for PTBD (difference + 39.6 days; 95% CI 11.3–65.9). Conclusions: After a failed ERCP for MBO, EUS-BD and PTBD achieved universal technical success and similar early biochemical responses, but EUS-BD was associated with higher complication rates and a significantly longer six-month survival. These findings support the individualized selection balancing procedural risk with the anticipated survival benefit and highlight the need for prospective comparative studies.


  • Open Access

    A Proof-of-Concept Pilot Study of Contrast-Enhanced Ultrasound as a Potential Alternative to Contrast-Enhanced Magnetic Resonance Imaging in the Surveillance of Hepatocellular Adenoma and Focal Nodular Hyperplasia

    by Adam Dobek, Mateusz Kobierecki, Adam Fabisiak, Wojciech Ciesielski, Marta Lenk-Jędrzejczak, Filip Franciszek Karuga, Filip Andrzej Dąbrowski, Ewa Małecka-Wojciesko, Ludomir Stefańczyk

    Background: Focal nodular hyperplasia (FNH) and hepatocellular adenoma (HA) are benign hepatic tumors that predominantly affect women of reproductive age and are associated with hormonal and metabolic factors. While FNH is a non-progressive lesion without malignant potential, HA carries a relevant risk of hemorrhage and malignant transformation. Differentiation between these entities remains challenging due to overlapping imaging features. Although contrast-enhanced magnetic resonance imaging (MRI) is considered the diagnostic reference standard, its cost, limited availability, and contraindications restrict routine long-term use. Therefore, contrast-enhanced ultrasound (CEUS) has emerged as an alternative modality for follow-up. This study evaluated the effectiveness of CEUS in long-term monitoring of FNH and HA compared with MRI. Methods: Patients with imaging-confirmed FNH or HA underwent paired CEUS and MRI examinations within 48 h at baseline and follow-up. Lesion size was assessed using maximal and minimal diameters, and longitudinal changes were classified according to RECIST-like criteria. Paired non-parametric statistical tests were applied. Results: 41 benign liver lesions (28 FNH and 13 HA) were analyzed across 92 paired examinations. Baseline lesion measurements were comparable between CEUS and MRI. A statistically significant difference was observed in the assessment of the largest lesion diameter, while no significant differences were detected for the shortest diameter. Longitudinal evaluation showed no significant differences between modalities in detecting lesion size changes. Response classification was concordant in 42 of 51 follow-up assessments, with stable disease as the most frequent outcome. Conclusions: After definitive diagnosis, CEUS may serve as a reliable standalone modality for routine long-term surveillance of FNH and HA in clinically stable patients. Its performance in lesion measurement and response assessment is comparable to MRI, while offering advantages in cost, accessibility, and patient tolerability. MRI may be reserved for cases with suspicious changes on CEUS.

  • Open Access

    Theoretical, Technical, and Analytical Foundations of Task-Based and Resting-State Functional Magnetic Resonance Imaging (fMRI)—A Narrative Review

    by Natalia Anna Koc, Maurycy Rakowski, Anna Dębska, Bartosz Szmyd, Agata Zawadzka, Karol Zaczkowski, Małgorzata Podstawka, Dagmara Wilmańska, Adam Dobek, Ludomir Stefańczyk, et al.

    Functional magnetic resonance imaging (fMRI) is a valuable tool for presurgical brain mapping, traditionally implemented with task-based paradigms (tb-fMRI) that measure blood oxygenation level-dependent (BOLD) signal changes during controlled motor or cognitive tasks. Tb-fMRI is a well-established tool for non-invasive localization of cortical eloquent areas, yet its dependence on patient cooperation and intact cognition limits use in individuals with aphasia, cognitive impairment, or in pediatric and other vulnerable populations. Resting-state fMRI (rs-fMRI) provides a task-free alternative by leveraging spontaneous low-frequency BOLD fluctuations to delineate intrinsic functional networks, including motor and language systems that show good spatial concordance with tb-fMRI and with direct cortical stimulation. This narrative review outlines the methodological foundations of tb-fMRI and rs-fMRI, comparing acquisition protocols, preprocessing and denoising pipelines, analytic approaches, and validation strategies relevant to presurgical planning. Particular emphasis is given to the technical and physiological foundations of BOLD imaging, statistical modeling, and the influence of motion, noise, and standardization on data reliability. Emerging evidence indicates that rs-fMRI can reliably expand mapping to patients with limited task compliance and may serve as a robust complementary modality in complex clinical contexts, though its methodological heterogeneity and absence of unified practice guidelines currently constrain widespread adoption. Future advances in harmonized preprocessing, multicenter validation, and integration with connectomics and machine learning frameworks are likely to be critical for translating rs-fMRI into routine, reliable presurgical workflows.


  • Open Access

    Magnetic resonance-based quantification of post-renal biopsy perirenal hematomas volumes

    by Piotr Białek, Adam Dobek, Katarzyna Szychowska, Weronika Banasik, Krzysztof Falenta, Mateusz Kobierecki, Ilona Kurnatowska, Ludomir Stefańczyk

    Percutaneous renal biopsy (PRB) is the gold standard for diagnosing nephropathies but, despite being generally safe, it carries the risk of hemorrhagic complications, particularly perirenal hematomas (PHs). Ultrasound, although commonly used, tends to underestimate hematoma volumes, whereas computed tomography (CT) accurately measures volumes but poses radiation concerns and often requires contrast media. Magnetic resonance imaging (MRI), free of these risks, offers high tissue resolution but remains underutilized for PH evaluation post-PRB. To evaluate the utility of MRI-based segmentation techniques for accurately quantifying PH volumes after PRB, as a complementary imaging modality to ultrasound and CT. We retrospectively analyzed MRI data from 85 patients who underwent PRB between July 2020 and May 2024. MRI-derived PH volumes were measured using manual segmentation. Clinical data were extracted from patient records, and the results were compared with data from a previous CT-based study. Perirenal hematoma was detected in 63 patients (74.1%) with a median volume of 26.2 mL (interquartile range (IQR): 7.2-59.3 mL), slightly smaller than CT-derived volumes (median: 38 mL, IQR: 18-85 mL). Using Spearman's rank correlation coefficient (rs), we found that serum creatinine (Cr; rs = 0.299, p = 0.039) and systolic blood pressure (SBP; rs = 0.333, p = 0.017) correlated positively with PH volume, while hemoglobin levels showed a negative correlation (rs = -0.322, p = 0.021). Hemodialysis was associated with larger PHs (odds ratio (OR) = 4.59, 95% confidence interval (95% CI): 1.20-17.58, p = 0.026); however, this finding is based on a model with modest predictive performance and requires further validation. Although its routine use may be limited, MRI could serve as a complementary tool for the detailed evaluation of PHs, offering a radiation-free and contrast media-free alternative to CT in clinical scenarios where immediate decision-making is not critical.


  • Open Access

    The Role of Preprocedural Computed Tomography Angiography in Enhancing Arterial Embolisation for Life-Threatening Haemoptysis: A Case Series

    by Anna Ziętarska, Adam Dobek, Piotr Białek, Wojciech Szubert, Sebastian Majewski, Ludomir Stefańczyk

    Haemoptysis arises from diverse respiratory diseases and may involve a broad spectrum of thoracic vessels. Arterial embolisation (AE) is an effective, repeatable, minimally invasive treatment option for life-threatening haemoptysis. This case series included 10 patients (mean age 34 years; six males; five with cystic fibrosis) who underwent 17 AE procedures for life-threatening haemoptysis between January 2018 and September 2025. The study assessed the role of wide-field computed tomography angiography (CTA), extending from the thoracic inlet to L2, in preprocedural planning, bleeding localisation and detection of systemic collaterals. CTA accurately predicted the culprit region in 16 out of 17 procedures. Non-bronchial systemic arteries were identified in 6 out of 10 patients, consistent with previous reports. CTA showed strong concordance with angiography and enabled the detection of uncommon collaterals, including subclavian and phrenic branches. Recurrence of hemoptysis occurred in one patient during follow-up; however, three patients were lost to follow-up. Wide-field CTA enhances the identification of systemic feeders and supports procedural planning, potentially reducing recurrence associated with missed culprit vessels. AE remains a valuable option for haemoptysis control in cystic fibrosis, with outcomes further improved following initiation of CFTR modulators. The small sample size and incomplete follow-up limit generalisability, but findings highlight the importance of CTA in guiding AE and improving clinical outcomes.