arXiv subjects

Arkadiusz Sitek

Publications and source records attributed to Arkadiusz Sitek.

At least 19 recordsLinked to original sources

arXiv2026

X-Splat: Gaussian Splatting for 3D CBCT Generation from Single Panoramic Radiograph

Generating a 3D dental volume from a single panoramic radiograph (PXR) could provide a low-radiation alternative to Cone-Beam Computed Tomography (CBCT), but the problem is highly underdetermined: panoramic acquisition integrates 3D attenuation along curved X-ray paths into a 2D image, leaving depth-resolved anatomy unobserved. Existing implicit and generative approaches often produce oversmoothed geometry or anatomically inconsistent hallucinations, lacking geometry-driven supervision and relying on smooth representations unable to precisely localize sharp anatomical boundaries. We propose X-Splat, the first Gaussian Splatting framework for generating CBCT-like 3D dental volumes from a single PXR. X-Splat uses the known panoramic acquisition geometry as a generation scaffold: learnable anisotropic Gaussian primitives are initialized along the X-ray paths that formed the input image and adjusted in a single feed-forward pass, constrained by Beer-Lambert reprojection and multi-view radiographic training supervision. A lightweight residual refiner adds dataset-level anatomical priors without overriding the geometry already resolved by the Gaussians. We train on synthetic PXR-CBCT pairs, enabling direct volumetric supervision without paired real scans. We further introduce segmentation-based geometry-aware metrics, providing the first evaluation of PXR-based generation over maxillofacial anatomy. X-Splat outperforms NeRF- and GAN-based baselines, recovering individual teeth, cortical boundaries, and alveolar structure, including the mandibular canal which prior methods fail to reconstruct. Code will be available at https://github.com/tomek1911/X-Splat

cs.CV↗

arXiv2026

Distilling Photon-Counting CT into Routine Chest CT through Clinically Validated Degradation Modeling

Photon-counting CT (PCCT) provides superior image quality with higher spatial resolution and lower noise compared to conventional energy-integrating CT (EICT), but its limited clinical availability restricts large-scale research and clinical deployment. To bridge this gap, we propose SUMI, a simulated degradation-to-enhancement method that learns to reverse realistic acquisition artifacts in low-quality EICT by leveraging high-quality PCCT as reference. Our central insight is to explicitly model realistic acquisition degradations, transforming PCCT into clinically plausible lower-quality counterparts and learning to invert this process. The simulated degradations were validated for clinical realism by board-certified radiologists, enabling faithful supervision without requiring paired acquisitions at scale. As outcomes of this technical contribution, we: (1) train a latent diffusion model on 1,046 PCCTs, using an autoencoder first pre-trained on both these PCCTs and 405,379 EICTs from 145 hospitals to extract general CT latent features that we release for reuse in other generative medical imaging tasks; (2) construct a large-scale dataset of over 17,316 publicly available EICTs enhanced to PCCT-like quality, with radiologist-validated voxel-wise annotations of airway trees, arteries, veins, lungs, and lobes; and (3) demonstrate substantial improvements: across external data, SUMI outperforms state-of-the-art image translation methods by 15% in SSIM and 20% in PSNR, improves radiologist-rated clinical utility in reader studies, and enhances downstream top-ranking lesion detection performance, increasing sensitivity by up to 15% and F1 score by up to 10%. Our results suggest that emerging imaging advances can be systematically distilled into routine EICT using limited high-quality scans as reference.

cs.CV↗

arXiv2026

Novel Asymmetrical High-Resolution and High-Sensitivity Brain Dedicated PET system: Design optimization and performance evaluation

Objective. This study investigates the best achievable performance of a brain-dedicated PET system with high resolution and sensitivity by evaluating different detector configurations, while maintaining a practical system design suitable for dynamic brain Imaging. Approach. Monte Carlo simulations were performed to evaluate system sensitivity and image quality under various timing resolutions (200 ps, 100 ps and 50 ps). The PET scanner geometry was optimized for human head imaging, featuring an elliptical cylindrical configuration with a neck cut-out, and front/back panels to enhance sensitivity and line of response (LOR) sampling. Detector configurations using LYSO:Ce crystals of varying thicknesses (15 mm and 20 mm) and depth of interaction (DOI) levels were simulated. Sensitivity was calculated using a point-like 511 keV back-to-back gamma source simulated at multiple locations within the field of view (FOV). Image reconstruction was conducted using list mode Maximum Likelihood Expectation Maximization (MLEM), assessing both a Derenzo-type phantom and a voxelated digital brain phantom. Main results. A location-dependent sensitivity ranging from 35.04% to 13.59% was achieved using a 20 mm thick LYSO:Ce crystal. Spatial resolution ranged from 0.8 mm to 1.5 mm within the FOV without applying resolution recovery techniques, measured using the FWHM of reconstructed hot rods. Previous results using 15 mm crystals with three DOI levels showed sensitivity between 23.42% and 15.99%, confirming the performance benefits of increased detector thickness and DOI capability. Significance. This study demonstrates the potential of a practical, brain-optimized PET system to achieve superior resolution and sensitivity for brain imaging. The findings offer valuable insights into optimal detector configurations, supporting the development of next-generation high-performance brain PET technologies.

physics.med-ph↗

arXiv2026

Maximum-Likelihood--Based Position Decoding of Laser Processed Converging Pixel CsI: Tl Detectors for High-Resolution SPECT

This study demonstrates the feasibility of a novel fabrication technique for high spatial resolution CsI: Tl scintillation detectors tailored for single photon emission computed tomography (SPECT) systems. Building upon our previously developed laser induced optical barrier (LIOB) method, which achieved high spatial resolution, excellent sensitivity, and 100% fabrication yield in CsI: Tl detectors, we extend this approach to a converging-pixel architecture. A CsI: Tl crystal array with converging pixels was designed and fabricated, featuring entrance-face pixels of 1.6x1.6 mm2 and photodetector side pixels of 2x2 mm2. To localize gamma-ray interactions, both the center of gravity (CoG) algorithm and a maximum-likelihood (ML) based decoding method were implemented. A custom built four axis motion platform was developed to deliver a finely collimated pencil beam at precisely controlled positions and angles across the array, enabling generation of a comprehensive dataset for prior knowledge and validation. The results demonstrate an energy resolution of 11.79+/-0.53% (collimated experiment) and a position localization accuracy of 1.00+/-0.42 mm (nearest neighbor interpolation), confirming that the proposed converging-pixel architecture, combined with statistical decoding algorithms, provides a promising path toward the development of high-performance SPECT detectors.

physics.med-ph↗

arXiv2026

Early and Prediagnostic Detection of Pancreatic Cancer from Computed Tomography

Pancreatic ductal adenocarcinoma (PDAC), one of the deadliest solid malignancies, is often detected at a late and inoperable stage. Retrospective reviews of prediagnostic CT scans, when conducted by expert radiologists aware that the patient later developed PDAC, frequently reveal lesions that were previously overlooked. To help detecting these lesions earlier, we developed an automated system named ePAI (early Pancreatic cancer detection with Artificial Intelligence). It was trained on data from 1,598 patients from a single medical center. In the internal test involving 1,009 patients, ePAI achieved an area under the receiver operating characteristic curve (AUC) of 0.939-0.999, a sensitivity of 95.3%, and a specificity of 98.7% for detecting small PDAC less than 2 cm in diameter, precisely localizing PDAC as small as 2 mm. In an external test involving 7,158 patients across 6 centers, ePAI achieved an AUC of 0.918-0.945, a sensitivity of 91.5%, and a specificity of 88.0%, precisely localizing PDAC as small as 5 mm. Importantly, ePAI detected PDACs on prediagnostic CT scans obtained 3 to 36 months before clinical diagnosis that had originally been overlooked by radiologists. It successfully detected and localized PDACs in 75 of 159 patients, with a median lead time of 347 days before clinical diagnosis. Our multi-reader study showed that ePAI significantly outperformed 30 board-certified radiologists by 50.3% (P < 0.05) in sensitivity while maintaining a comparable specificity of 95.4% in detecting PDACs early and prediagnostic. These findings suggest its potential of ePAI as an assistive tool to improve early detection of pancreatic cancer.

cs.CV↗

arXiv2025

Towards Integrated Clinical-Computational Nuclear Medicine

The field of Clinical-Computational Nuclear Medicine is rapidly advancing, fueled by AI, tracer kinetic modeling, radiomics, and integrated informatics. These technologies improve imaging quality, automate lesion detection, and enable personalized radiopharmaceutical therapy through physiologically based pharmacokinetic (PBPK) modeling and voxel-level dosimetry. Workflow automation and Natural Language Processing (NLP) further enhance operational efficiency. However, successful implementation and adoption of these tools require clinical oversight to ensure accuracy, interpretability, and patient safety. This paper highlights key computational innovations and emphasizes the critical role of clinician-guided evaluation in shaping the future of precision imaging and therapy.

physics.med-ph↗

arXiv2025

GEPAR3D: Geometry Prior-Assisted Learning for 3D Tooth Segmentation

Tooth segmentation in Cone-Beam Computed Tomography (CBCT) remains challenging, especially for fine structures like root apices, which is critical for assessing root resorption in orthodontics. We introduce GEPAR3D, a novel approach that unifies instance detection and multi-class segmentation into a single step tailored to improve root segmentation. Our method integrates a Statistical Shape Model of dentition as a geometric prior, capturing anatomical context and morphological consistency without enforcing restrictive adjacency constraints. We leverage a deep watershed method, modeling each tooth as a continuous 3D energy basin encoding voxel distances to boundaries. This instance-aware representation ensures accurate segmentation of narrow, complex root apices. Trained on publicly available CBCT scans from a single center, our method is evaluated on external test sets from two in-house and two public medical centers. GEPAR3D achieves the highest overall segmentation performance, averaging a Dice Similarity Coefficient (DSC) of 95.0% (+2.8% over the second-best method) and increasing recall to 95.2% (+9.5%) across all test sets. Qualitative analyses demonstrated substantial improvements in root segmentation quality, indicating significant potential for more accurate root resorption assessment and enhanced clinical decision-making in orthodontics. We provide the implementation and dataset at https://github.com/tomek1911/GEPAR3D.

eess.IV↗

arXiv2025

RegScore: Scoring Systems for Regression Tasks

Scoring systems are widely adopted in medical applications for their inherent simplicity and transparency, particularly for classification tasks involving tabular data. In this work, we introduce RegScore, a novel, sparse, and interpretable scoring system specifically designed for regression tasks. Unlike conventional scoring systems constrained to integer-valued coefficients, RegScore leverages beam search and k-sparse ridge regression to relax these restrictions, thus enhancing predictive performance. We extend RegScore to bimodal deep learning by integrating tabular data with medical images. We utilize the classification token from the TIP (Tabular Image Pretraining) transformer to generate Personalized Linear Regression parameters and a Personalized RegScore, enabling individualized scoring. We demonstrate the effectiveness of RegScore by estimating mean Pulmonary Artery Pressure using tabular data and further refine these estimates by incorporating cardiac MRI images. Experimental results show that RegScore and its personalized bimodal extensions achieve performance comparable to, or better than, state-of-the-art black-box models. Our method provides a transparent and interpretable approach for regression tasks in clinical settings, promoting more informed and trustworthy decision-making. We provide our code at https://github.com/SanoScience/RegScore.

eess.IV↗

arXiv2025

Mamba Goes HoME: Hierarchical Soft Mixture-of-Experts for 3D Medical Image Segmentation

In recent years, artificial intelligence has significantly advanced medical image segmentation. Nonetheless, challenges remain, including efficient 3D medical image processing across diverse modalities and handling data variability. In this work, we introduce Hierarchical Soft Mixture-of-Experts (HoME), a two-level token-routing layer for efficient long-context modeling, specifically designed for 3D medical image segmentation. Built on the Mamba Selective State Space Model (SSM) backbone, HoME enhances sequential modeling through adaptive expert routing. In the first level, a Soft Mixture-of-Experts (SMoE) layer partitions input sequences into local groups, routing tokens to specialized per-group experts for localized feature extraction. The second level aggregates these outputs through a global SMoE layer, enabling cross-group information fusion and global context refinement. This hierarchical design, combining local expert routing with global expert refinement, enhances generalizability and segmentation performance, surpassing state-of-the-art results across datasets from the three most widely used 3D medical imaging modalities and varying data qualities. The code is publicly available at https://github.com/gmum/MambaHoME.

eess.IV↗

arXiv2025

Federated Timeline Synthesis: Scalable and Private Methodology For Model Training and Deployment

We present Federated Timeline Synthesis (FTS), a novel framework for training generative foundation models across distributed timeseries data applied to electronic health records (EHR). At its core, FTS represents patient history as tokenized Patient Health Timelines (PHTs), language-agnostic sequences encoding temporal, categorical, and continuous clinical information. Each institution trains an autoregressive transformer on its local PHTs and transmits only model weights to a central server. The server uses the generators to synthesize a large corpus of trajectories and train a Global Generator (GG), enabling zero-shot inference via Monte Carlo simulation of future PHTs. We evaluate FTS on five clinically meaningful prediction tasks using MIMIC-IV data, showing that models trained on synthetic data generated by GG perform comparably to those trained on real data. FTS offers strong privacy guarantees, scalability across institutions, and extensibility to diverse prediction and simulation tasks especially in healthcare, including counterfactual inference, early warning detection, and synthetic trial design.

cs.LG↗

arXiv2025

Foundation Model of Electronic Medical Records for Adaptive Risk Estimation

Hospitals struggle to predict critical outcomes. Traditional early warning systems, like NEWS and MEWS, rely on static variables and fixed thresholds, limiting their adaptability, accuracy, and personalization. We previously developed the Enhanced Transformer for Health Outcome Simulation (ETHOS), an AI model that tokenizes patient health timelines (PHTs) from EHRs and uses transformer-based architectures to predict future PHTs. ETHOS is a versatile framework for developing a wide range of applications. In this work, we develop the Adaptive Risk Estimation System (ARES) that leverages ETHOS to compute dynamic, personalized risk probabilities for clinician-defined critical events. ARES also features a personalized explainability module that highlights key clinical factors influencing risk estimates. We evaluated ARES using the MIMIC-IV v2.2 dataset together with its Emergency Department (ED) extension and benchmarked performance against both classical early warning systems and contemporary machine learning models. The entire dataset was tokenized resulting in 285,622 PHTs, comprising over 360 million tokens. ETHOS outperformed benchmark models in predicting hospital admissions, ICU admissions, and prolonged stays, achieving superior AUC scores. Its risk estimates were robust across demographic subgroups, with calibration curves confirming model reliability. The explainability module provided valuable insights into patient-specific risk factors. ARES, powered by ETHOS, advances predictive healthcare AI by delivering dynamic, real-time, personalized risk estimation with patient-specific explainability. Although our results are promising, the clinical impact remains uncertain. Demonstrating ARES's true utility in real-world settings will be the focus of our future work. We release the source code to facilitate future research.

cs.LG↗

arXiv2024

TabMixer: Noninvasive Estimation of the Mean Pulmonary Artery Pressure via Imaging and Tabular Data Mixing

Right Heart Catheterization is a gold standard procedure for diagnosing Pulmonary Hypertension by measuring mean Pulmonary Artery Pressure (mPAP). It is invasive, costly, time-consuming and carries risks. In this paper, for the first time, we explore the estimation of mPAP from videos of noninvasive Cardiac Magnetic Resonance Imaging. To enhance the predictive capabilities of Deep Learning models used for this task, we introduce an additional modality in the form of demographic features and clinical measurements. Inspired by all-Multilayer Perceptron architectures, we present TabMixer, a novel module enabling the integration of imaging and tabular data through spatial, temporal and channel mixing. Specifically, we present the first approach that utilizes Multilayer Perceptrons to interchange tabular information with imaging features in vision models. We test TabMixer for mPAP estimation and show that it enhances the performance of Convolutional Neural Networks, 3D-MLP and Vision Transformers while being competitive with previous modules for imaging and tabular data. Our approach has the potential to improve clinical processes involving both modalities, particularly in noninvasive mPAP estimation, thus, significantly enhancing the quality of life for individuals affected by Pulmonary Hypertension. We provide a source code for using TabMixer at https://github.com/SanoScience/TabMixer.

eess.IV↗

arXiv2024

Zero Shot Health Trajectory Prediction Using Transformer

Integrating modern machine learning and clinical decision-making has great promise for mitigating healthcare's increasing cost and complexity. We introduce the Enhanced Transformer for Health Outcome Simulation (ETHOS), a novel application of the transformer deep-learning architecture for analyzing high-dimensional, heterogeneous, and episodic health data. ETHOS is trained using Patient Health Timelines (PHTs)-detailed, tokenized records of health events-to predict future health trajectories, leveraging a zero-shot learning approach. ETHOS represents a significant advancement in foundation model development for healthcare analytics, eliminating the need for labeled data and model fine-tuning. Its ability to simulate various treatment pathways and consider patient-specific factors positions ETHOS as a tool for care optimization and addressing biases in healthcare delivery. Future developments will expand ETHOS' capabilities to incorporate a wider range of data types and data sources. Our work demonstrates a pathway toward accelerated AI development and deployment in healthcare.

cs.LG↗

arXiv2024

Let Me DeCode You: Decoder Conditioning with Tabular Data

Training deep neural networks for 3D segmentation tasks can be challenging, often requiring efficient and effective strategies to improve model performance. In this study, we introduce a novel approach, DeCode, that utilizes label-derived features for model conditioning to support the decoder in the reconstruction process dynamically, aiming to enhance the efficiency of the training process. DeCode focuses on improving 3D segmentation performance through the incorporation of conditioning embedding with learned numerical representation of 3D-label shape features. Specifically, we develop an approach, where conditioning is applied during the training phase to guide the network toward robust segmentation. When labels are not available during inference, our model infers the necessary conditioning embedding directly from the input data, thanks to a feed-forward network learned during the training phase. This approach is tested using synthetic data and cone-beam computed tomography (CBCT) images of teeth. For CBCT, three datasets are used: one publicly available and two in-house. Our results show that DeCode significantly outperforms traditional, unconditioned models in terms of generalization to unseen data, achieving higher accuracy at a reduced computational cost. This work represents the first of its kind to explore conditioning strategies in 3D data segmentation, offering a novel and more efficient method for leveraging annotated data. Our code, pre-trained models are publicly available at https://github.com/SanoScience/DeCode .

eess.IV↗

arXiv2024

Collimator-less SPECT System Design for Dynamic Whole-body Imaging

In this study, we introduce a Compton SPECT system for whole-body imaging of Actinium-225 (225Ac), one of the trending radionuclides for targeted alpha therapy (TAT). The Compton SPECT system enables multi-energy gamma photon detection with higher efficiency compared to mechanically collimated SPECT. The system consists of two detectors, providing a field of view (FOV) adequate for whole-body imaging, while achieving high sensitivity and clinically usable imaging resolution within a reasonable scanning time. This work focuses on the system design and evaluation using the Monte Carlo simulation toolkit Gate. The imaging performance is evaluated at two energy peaks (218 keV, 440 keV), representing the major detectable gamma energies generated from 225Ac. We explore the possibility of using the Compton SPECT system for treatment response monitoring in TAT. Results demonstrate an image resolution of 1.0 cm using a NEMA IQ phantom with 5.7 MBq of 225Ac simulated in a cold background. An image resolution of 1.3 cm can be achieved with a hot to background ratio of 30:1, and a resolution of 3.7 cm can be achieved with an activity ratio of 12:1. The best achievable sensitivity at 10 cm distance to the detector is 0.5% with the two energy windows (211-225 keV, 430-450 keV) selected. The proposed system may serve as an alternative imaging tool for TAT scanning in clinical settings.

physics.med-ph↗

arXiv2024

Multi-Center Fetal Brain Tissue Annotation (FeTA) Challenge 2022 Results

Segmentation is a critical step in analyzing the developing human fetal brain. There have been vast improvements in automatic segmentation methods in the past several years, and the Fetal Brain Tissue Annotation (FeTA) Challenge 2021 helped to establish an excellent standard of fetal brain segmentation. However, FeTA 2021 was a single center study, and the generalizability of algorithms across different imaging centers remains unsolved, limiting real-world clinical applicability. The multi-center FeTA Challenge 2022 focuses on advancing the generalizability of fetal brain segmentation algorithms for magnetic resonance imaging (MRI). In FeTA 2022, the training dataset contained images and corresponding manually annotated multi-class labels from two imaging centers, and the testing data contained images from these two imaging centers as well as two additional unseen centers. The data from different centers varied in many aspects, including scanners used, imaging parameters, and fetal brain super-resolution algorithms applied. 16 teams participated in the challenge, and 17 algorithms were evaluated. Here, a detailed overview and analysis of the challenge results are provided, focusing on the generalizability of the submissions. Both in- and out of domain, the white matter and ventricles were segmented with the highest accuracy, while the most challenging structure remains the cerebral cortex due to anatomical complexity. The FeTA Challenge 2022 was able to successfully evaluate and advance generalizability of multi-class fetal brain tissue segmentation algorithms for MRI and it continues to benchmark new algorithms. The resulting new methods contribute to improving the analysis of brain development in utero.

eess.IV↗

arXiv2024

MISS: Multiclass Interpretable Scoring Systems

In this work, we present a novel, machine-learning approach for constructing Multiclass Interpretable Scoring Systems (MISS) - a fully data-driven methodology for generating single, sparse, and user-friendly scoring systems for multiclass classification problems. Scoring systems are commonly utilized as decision support models in healthcare, criminal justice, and other domains where interpretability of predictions and ease of use are crucial. Prior methods for data-driven scoring, such as SLIM (Supersparse Linear Integer Model), were limited to binary classification tasks and extensions to multiclass domains were primarily accomplished via one-versus-all-type techniques. The scores produced by our method can be easily transformed into class probabilities via the softmax function. We demonstrate techniques for dimensionality reduction and heuristics that enhance the training efficiency and decrease the optimality gap, a measure that can certify the optimality of the model. Our approach has been extensively evaluated on datasets from various domains, and the results indicate that it is competitive with other machine learning models in terms of classification performance metrics and provides well-calibrated class probabilities.

cs.LG↗

arXiv2023

Can gamification reduce the burden of self-reporting in mHealth applications? A feasibility study using machine learning from smartwatch data to estimate cognitive load

The effectiveness of digital treatments can be measured by requiring patients to self-report their state through applications, however, it can be overwhelming and causes disengagement. We conduct a study to explore the impact of gamification on self-reporting. Our approach involves the creation of a system to assess cognitive load (CL) through the analysis of photoplethysmography (PPG) signals. The data from 11 participants is utilized to train a machine learning model to detect CL. Subsequently, we create two versions of surveys: a gamified and a traditional one. We estimate the CL experienced by other participants (13) while completing surveys. We find that CL detector performance can be enhanced via pre-training on stress detection tasks. For 10 out of 13 participants, a personalized CL detector can achieve an F1 score above 0.7. We find no difference between the gamified and non-gamified surveys in terms of CL but participants prefer the gamified version.

cs.LG↗