<br><b>Objective:</b> Positron emission tomography (PET) scanners with plastic scintillators offermore cost-effective instrumentation to image the distribution of radiopharmaceuticals.However, inter-detector scatters among plastic scintillators can lead to more falsecoincidences than in conventional PET scanners, since annihilation photons in plasticscintillators dominantly interact via Compton scattering, which deposits only a portionof the photon energy. A scatter test (ST), combined with a lower energy depositionthreshold of 200 keV, has been used to preselect the coincidence events.</br><br><b>Methods:</b> In this work, we investigate the impact of temporal and spatial resolutionlimitations as well as a variation of the energy threshold on the preselection anddifferent subsequent coincidence event selection policies via Monte Carlo simulations.We simulate the total-body Jagiellonian-PET (TB-J-PET), combined with a brain PETinsert imaging a human-sized water phantom.</br><br><b>Results:</b> We find that coincidence time resolution (CTR) worse than 200 ps poseslimitations on the ST for scanners close to the patient, such as the brain PET. Also,coincidence event selection requiring energy loss higher than 200 keV performs suboptimally,whereas a lower energy threshold (50 keV), combined with a time-basedselection policy, can capture a higher percentage of true events, even under realistictime resolution.</br><br><b>Conclusions:</b> We recommend the adaptation of a time-based event selection policytogether with a lowered energy threshold, which can also significantly increasesensitivity, as the latter rises faster than the fraction of true and non-phantom--scattered events decreases. Dedicated analyses in the scatter-corrected imagedomain are necessary to further investigate this potential.</br>

<br><b>Objective:</b> The use of neural networks for disease classification based on medical imaging is susceptible to variations in results caused by even a single-pixel change, a phenomenon known as a one-pixel attack, which should be examined qualitatively and quantitatively.</br> <br><b>Methods:</b> For an extended dataset of brain MRI images representing four diagnoses, the networks VGG-16, ResNet-50, DenseNet-121, MobileNetV2, EfficientNet-B0, NASNetMobile, and ViT Base were implemented. Each model was trained three times on 96 × 96 inputs, with the best-performing trial selected for adversarial testing (Phase 1). The three most robust models from Phase 1 (VGG-16, MobileNetV2, EfficientNet-B0) were then retrained on 224 × 224 inputs to assess the effect of higher resolution on susceptibility (Phase 2). The susceptibility of a diagnosis change to a single bright pixel alteration in the input image was assessed, and an average number of vulnerable pixels (ANVP) per image was carried out.</br> <br><b>Results:</b> At 96 × 96 resolution, the least vulnerable model was MobileNetV2 (ANVP: 20.45, susceptibility: 0.22%). This was followed by ViT Base (22.20, 0.24%), EfficientNet-B0 (38.55, 0.42%), DenseNet-121 (43.52, 0.47%), ResNet-50 (69.11, 0.75%), and VGG-16 (78.66, 0.85%). The most vulnerable was NASNetMobile (119.52, 1.30%). At 224 × 224 resolution, robustness further improved for EfficientNet-B0 (37.53, 0.07%) and MobileNetV2 (49.51, 0.10%), while VGG-16 remained less stable (99.44, 0.20%).</br> <br><b>Conclusions:</b> Implementing disease classification based on medical imaging using neural networks may pose a potential risk of misinterpretation due to changes in data irrelevant to the study, which are clearly noticeable to a human.</br>

<br><b>Objective:</b> The integration of omics technologies has opened new opportunities in toxicological research. This article aims to explore how toxico-proteomics and toxico-metabolomics contribute to the understanding of xenobiotic mechanisms, biomarker discovery, and modern risk assessment frameworks.</br> <br><b>Methods:</b> Relevant literature was analysed to highlight recent advances in proteomics and metabolomics applied to toxicology. Particular attention was given to mass spectrometry-based approaches, spatial omics, in silico modelling, and combined omics strategies. Case examples from drug- and environment-related toxicology were used to illustrate practical applications.</br> <br><b>Results:</b> High-resolution mass-spectrometry-based proteomics enables the sensitive detection of changes in protein levels, post-translational modifications, and proteinprotein interactions. Toxico-proteomic studies have clarified mechanisms of cardio-, hepato-, and atd-neurotoxic effects. Metabolomics supports the profiling of low molecular weight compounds and early responses to toxicants. Toxico-metabolomic analyses identified changes related to energy metabolism and amino acid metabolism. In vitro models and zebrafish embryos provided organ-specific insights. Integrating omics data has led to the identification of candidate biomarkers of exposure and toxic effects.</br> <br><b>Conclusions:</b> Toxico-proteomics and toxico-metabolomics represent powerful tools for toxicology. Their application enhances the sensitivity of toxicity detection, reduces reliance on animal models, and supports the development of predictive strategies. As analytical platforms and computational tools continue to evolve, these disciplines are expected to play an increasingly central role in environmental and biomedical toxicology, with implications for diagnostics, therapeutics, and regulatory demands.</br>

<b>Introduction:</b> Medical education faces increasing content demands, and digital anatomy atlases have become valuable adjuncts to traditional anatomy courses. However, most available atlases are limited to two-dimensional displays, restricting the interactive, spatial learning that is essential for deep anatomical understanding. In response, we developed and implemented a year-long anatomy course for first-year medical students at Jagiellonian University Medical College that integrated a mixed reality-based holographic anatomy software suite with mixed reality (MR) technology. This curriculum aimed to complement traditional anatomy education by offering interactive 3D holographic representations of anatomical structures, allowing realtime exploration and manipulation in a spatial context. <br><b>Methods:</b> A series of MR-enhanced anatomy lessons was created in alignment with the existing first-year anatomy curriculum. Sessions were conducted in a dedicated mixed reality laboratory, each led by an anatomy instructor trained in MR equipment and accommodating up to 9 students (with the instructor present as the 10<sup>th</sup> person). A total of 98 first-year medical students participated in the course. After each session, students were asked to complete a structured survey evaluating their experiences and perceptions of the MR learning environment. Ninety-four students (96% of participants) responded to at least one survey, and complete data from 85 students were included in the final analysis. <br><b>Results:</b> The vast majority of participants reported positive experiences with the MR-based curriculum. Students indicated that the MR sessions enhanced their understanding of anatomical structures and spatial relationships. No significant differences in overall satisfaction were observed between student subgroups. For example, when grouped by prior anatomy coursework, 100% of students without prior anatomy experience and 95% of those with prior experience reported that they could identify anatomical structures after the MR sessions. Similarly, 90% vs. 93% of these groups, respectively, noted improved recognition of anatomical spatial relationships. When grouped by prior use of 3D visualization tools, some differences emerged in self-assessed proficiency: students with previous 3D experience reported greater ease in identifying structures (95% vs. 81%, <i>p</i> = 0.03) and understanding anatomical relationships (97% vs. 81%, <i>p</i> = 0.03), compared to those without such experience. In contrast, students without prior 3D experience found certain MR features more useful than did experienced students – for instance, 88% vs. 70% rated the layer toggle function as helpful (<i>p</i> = 0.048). Despite these subgroup variations, there was broad agreement on the value of MR: 71.8% of all respondents preferred a hybrid learning model combining MR with traditional methods, unanimously emphasizing that MR should supplement rather than replace cadaveric dissection. <br><b>Discussion:</b> Our findings suggest that MR technology is a valuable tool for enhancing anatomy education, particularly by enabling visualization of spatial relationships that are difficult to achieve with textbooks or cadaveric dissection alone. Students appreciated the interactive 3D features of the MR software, which fostered engagement and helped them explore complex anatomical details more intuitively. At the same time, participants recognized the continued importance of hands-on cadaveric labs for tactile learning experiences, indicating that an optimal approach is a hybrid, model integrating MR with traditional anatomy instruction. Notably, this study’s conclusions are drawn from self-reported student data, so any assumptions about long-term learning outcomes must be made cautiously. Future research should evaluate the impact of MR on objective learning measures (such as exam performance and knowledge retention over time) and explore best practices for integrating MR technology into anatomy curricula in diverse educational settings. <br><b>Conclusions:</b> Mixed reality technology was well-received by first-year medical students and effectively enhanced their spatial understanding of anatomical structures. MR sessions were found to be engaging, intuitive and supportive of traditional cadaveric dissection. Students strongly favored a hybrid learning model, suggesting that MR should supplement – not replace – classical methods in anatomy education.

<b>Objective:</b> This study aimed to evaluate the information content of individual electrocardiogram (ECG) leads and their inter-lead correlations using a deep learning approach. Specifically, we investigated the capability of a neural network to reconstruct missing ECG leads from reduced-lead configurations, thereby revealing each lead's unique and shared informational value.<br><b>Methods:</b> We developed a U-net convolutional neural network to reconstruct missingleads in 12-lead ECG recordings. The model was trained using the PTB-XL dataset andtested on the PTB dataset. We trained the model with varying combinations of inputleads, including single limb leads, combinations of two limb leads and configurations,including one or two precordial leads. We evaluated the model's performance usingmean squared error (MSE) between the reconstructed and actual signals. <br><b>Results:</b> The models demonstrated varying reconstruction accuracy depending on the input lead configuration. Precordial leads V1 and V6 showed the highest reconstruction fidelity from limb leads alone, while V3 consistently exhibited the lowest, indicating its unique informational content. <br><b>Conclusions:</b> The proposed method effectively quantifies the informational value of ECG leads. This has significant implications for optimizing lead selection in diagnostic scenarios, particularly in settings where complete 12-lead ECGs are impractical. In addition, the study provides insights into the physiological underpinnings of ECG signals and their propagation. The findings pave the way for advances in telemedicine, portable ECG devices and personalized cardiac diagnostics by reducing redundancy and improving signal interpretation.

<b>Objective:</b> Scintigraphy with technetium-99m (Tc-99m) is essential for sentinel lymph node (SLN) detection in gynecological cancers, aiding preoperative planning and reducing unnecessary lymphadenectomy. However, traditional image interpretation is subjective. This study applies numerical tools: signal-to-noise ratio (SNR) and C factor (SLN SNR/ background SNR) to objectively assess scintigraphic image quality and protocol efficacy. <br><b>Materials and methods:</b> A total of 172 patients with endometrial cancer underwent SLN detection at the Maria Sklodowska-Curie National Research Institute of Oncology, Kraków branch (from 2016–2024). Two protocols were compared: 1. short protocol: 40 MBq Tc-99m, planar imaging – 30 minutes post-injection; 2. long protocol: 120 MBq Tc-99m, planar imaging – 18 hours post-injection. SLN detection was confirmed intraoperatively using a gamma probe and subsequently verified by histopathology. <br><b>Results:</b> The long protocol achieved a detection sensitivity of 94%, compared to 70% in the short protocol. Based on the analysis, the minimum SNR required for effective SLN detection was 1.46, and the threshold C factor was ≥ 3.61. Although the mean SNR in the long protocol was lower (3.51), the prolonged biological retention of Tc-99m in SLNs contributed to improved detection accuracy. <br><b>Conclusions:</b> Numerical analysis of SLN scintigraphy in 172 patients provides an objective framework for evaluating image quality and improves detection accuracy. The long protocol offers superior visualization of sentinel lymph nodes. Standardized quantitative imaging may support the development of automated and AI-assisted diagnostic tools in nuclear medicine.

<b>Objective:</b> To explore readout architectures for the simultaneous high-resolution timing and bidimensional tracking of charged particles with Resistive Plate Chambers (TOF-tracker) and for the accurate detection of gamma rays for Positron Emission Tomography (PET).<b>Materials and methods:</b> Resistive plate chambers and their corresponding readout systems under evaluation were exposed to cosmic rays and β<sup>+</sup> sources.<b>Results:</b> Over an active area of 625 cm<sup>2</sup>, we obtained a time resolution of 61 ps ơ and bidimensional position resolution below 150 μm ơ for the tracking and timing of charged particles from cosmic rays. The intrinsic precision for localising a small β<sup>+</sup> source via the detection of its annihilation radiation was determined to be 0.49 mm FWHM.<b>Conclusions:</b> The proposed device exhibits excellent timing and position resolution for the tracking and timing of charged particles, with potential applications in nuclear and high-energy particle physics, as well as gamma imaging with applications in PET.

When the article was first published in 2022, there were a few incorrect details and authors would like to submit the following corrections for the article:(1) In the original version of this article the affiliation of Monika Szczepanek, who is the corresponding author, is incorrect.On the page 144, the following sentence: <br><b>*Corresponding authors: Monika Szczepanek,</b> Center for Theranostics, Jagiellonian University, Kopernika 40 St., 31-034 Krakw, Poland, e-mail: monika.szczepanek@doctoral.uj.edu.plshould have read: <br><b>*Corresponding authors: Monika Szczepanek,</b> Department of Medical Physics, M. Smoluchowski Institute of Physics, Jagiellonian University, Łojasiewicza 11 St., 30-348 Kraków, Poland; Center for Theranostics, Jagiellonian University, Kopernika 40 St., 31-034 Kraków, Poland(2) In the original version of this article there is an error in the name of the 4F2hc gene in the Abstract and Results/Discussion sections. In these sections, the incorrect name „4Fhc” appears three times instead of the correct name.On the page 144 (Abstract), the following sentence: <br>Results: Transcriptomic data show that in melanocytes, LFC for SLC7A5 (LAT1) and SLC3A2 (4Fhc) was higher than in melanoma cell lines, which corresponded with their melanin content.should have read: <br>Results: Transcriptomic data show that in melanocytes, LFC for SLC7A5 (LAT1) and SLC3A2 (4F2hc) was higher than in melanoma cell lines, which corresponded with their melanin content.On the pages 147-148 (Results/Discussion), the following sentences: <br>This is because this gene encodes a heavy chain (4Fhc) subunit necessary for the proper function of LAT1 [10]. This relationship was also evident for WM115 and WM266- 4 where we observed the highest expression levels of the SLC7A5 (LAT1) and SLC3A2 (4Fhc) genes.should have read: <br>This is because this gene encodes a heavy chain (4F2hc) subunit necessary for the proper function of LAT1 [10]. This relationship was also evident for FM55p and WM266-4 where we observed the highest expression levels of the SLC7A5 (LAT1) and SLC3A2 (4F2hc) genes.(3) In the original version of this article an error occurred while typing of the name of the HEMa-LP cell line. The incorrect name “HEMa-Lp” appears instead of the correct name in the article. This error has since been corrected throughout the text and in Figures.(4) In the original version of this article high melanotic melanoma cell line have been subject to misnomer. The incorrect name “WM115” appears instead of the correct name “FM55p” in the article. This error has now been corrected throughout the text and in Figures, and this correction has not changed the description, interpretation or the original conclusions of the manuscript.(5) In the corrected version of this article a new reference number 26 has been added with information on the FM55p melanoma cell line.On the pages 147 (Results/Discussion), the following sentence: <br>HEMa-Lp is a cell line derived from donor normal skin melanocytes, while WM115 and WM266-4 are melanoma cell lines from one patient [26].should have read: <br>HEMa-LP is a cell line derived from donor normal skin melanocytes, while FM55p is a primary melanoma cell line [26] and WM266-4 is a metastatic melanoma cell line [27].(6) Due to the addition of a new reference item (number 26) in the corrected version of this article, the numbering of references 26-27 in the original version of this article has been changed. Now they have numbers from 27-30.The original article has been corrected. The authors apologize for any inconvenience caused by these errors.

Objective: Breast cancer remains a leading cause of mortality among women worldwide. Early detection and accurate prognosis are crucial for improving patient outcomes. This study presents a novel approach that integrates feature elimination techniques with machine learning to enhance the accuracy of breast cancer prognosis. The approach addresses class imbalance in the dataset to improve sensitivity, particularly in minimizing false negatives. Additionally, it emphasizes the use of machine learning algorithms, which are considered more transparent and computationally efficient compared to deep learning methods. Method: The Wisconsin Breast Cancer (WBC) dataset was used to develop an interpretable machine learning model. Recursive Feature Elimination (RFE) identified key features, while Principal Component Analysis (PCA) reduced dimensionality. The optimized feature set was trained using XGBoost. To address class imbalance, class weighting and decision threshold adjustments were applied to improve sensitivity and minimize false negatives.Results: The model achieved high performance: accuracy of 99.12%, precision of 100%, recall of 97.69%, and an F1 score of 98.9%. Feature selection and class imbalance handling enhanced sensitivity and computational efficiency. The model's interpretable results highlight its suitability for clinical applications.Conclusions: This study presents an interpretable machine learning model integrating RFE, PCA, and XGBoost to enhance breast cancer prognosis. High accuracy and sensitivity, coupled with explainability, make it a promising tool for clinical decision-making in early detection and treatment planning.

<b>Objective:</b> Pancreatic cancer is an asymptomatic disease and, based on statistical studies, it is the fourth leading cause of cancer-related death. Pancreatic ductal adenocarcinoma (PDAC), which accounts for over 95% of pancreatic cancers, is typically detectable at advanced stages. Standard diagnostic methods include bloodbased tests and imaging. Standard diagnostic methods include blood-based tests and imaging. Biomarkers play a key role as indicators in blood tests, offering valuable insights into disease detection and monitoring. Mesothelin, a cell-surface glycoprotein, and vimentin, an intermediate filament protein, are promising biomarker candidates. <br><b>Methods:</b> In this study, these biomarkers were conjugated with magnetic nanoparticles (MNPs) and utilized for cellular tracking through magnetic particle spectroscopy (MPS). Capan-1 (a pancreatic cancer cell line) and bone marrow stem cells (BMSC) were treated with the targeted MNPs. Subsequently, MNP-labelled cells were evaluated with imaging modalities such as MPS and confocal microscopy. <br><b>Results:</b> In the case of the MPS modality, a home-made MPS device with a detection limit of 1 ng of MNPs was used. The results showed that MPS can quantitatively trace MNPs signals and differentiate between various treatments. <br><b>Conclusions:</b> Detection of labelled cells via MPS is a novel method with features such as sensitivity, non-invasiveness, and no background noise. This new technology can pave the way for imaging quantification of pancreatic cancer in its primary stages and for tracking cancer cell populations. populations.