Gabriel Ansah, Eden Ruffell, Delmiro Fernandez-Reyes
et al.
Automated blood morphology analysis can support hematological diagnostics in low- and middle-income countries (LMICs) but remains sensitive to dataset shifts from staining variability, imaging differences, and rare morphologies. Building centralized datasets to capture this diversity is often infeasible due to privacy regulations and data-sharing restrictions. We introduce a federated learning framework for white blood cell morphology analysis that enables collaborative training across institutions without exchanging training data. Using blood films from multiple clinical sites, our federated models learn robust, domain-invariant representations while preserving complete data privacy. Evaluations across convolutional and transformer-based architectures show that federated training achieves strong cross-site performance and improved generalization to unseen institutions compared to centralized training. These findings highlight federated learning as a practical and privacy-preserving approach for developing equitable, scalable, and generalizable medical imaging AI in resource-limited healthcare environments.
Mantle cell lymphoma (MCL) is a rare B-cell malignancy with significant clinical heterogeneity. We have recently shown that a deeper response of MCL cells to B-cell receptor (BCR) stimulation identified a subset of patients with a higher risk of progression. Although therapies targeting the BCR, such as the BTK inhibitor ibrutinib (Ibr), have a high initial response rate in MCL, relapse remains a challenge. To investigate the ability of Ibr to modulate the BCR signaling network, we analyzed the activating phosphorylation status of BCR-associated kinases (BAK), i.e., SYK, PLCγ2, STAT5, ERK1/2, NF-κB p65, AKT, BTK, STAT3, in cells from peripheral blood of 29 MCL patients at diagnosis following Ibr treatment in vitro, in the basal condition and after BCR stimulation with anti-IgM, using phospho-specific flow cytometry, a multiparametric assay allowing functional signaling analysis at a single-cell level, combined with fluorescent cell barcoding. In the basal condition, Ibr induced a significant average reduction of phosphorylation level for all BAKs but NF-κB p65 and STAT3. In the BCR-stimulated condition, we detected a significant average reduction of phosphorylation for all BAKs, including NF-κB p65, but STAT3. Comparison of the phosphorylation responses to Ibr for each BAK between the basal and BCR-stimulated conditions showed that the phosphorylation reduction in response to Ibr was significantly deeper in the BCR-stimulated condition compared with the basal one for SYK, ERK1/2, NF-κB p65, and AKT. In contrast, reduction in BTK phosphorylation resulted attenuated, although not significantly, in the BCR-stimulated condition compared with the basal one (Figure 1). In conclusion, our data show that in the basal condition most BAKs are highly connected in response to Ibr, thus suggesting that these proteins belong to the same signaling subnetwork. Importantly, the activation of the BCR quantitatively and qualitatively rewires the network of BAKs sensible to Ibr, thus suggesting that the microenvironment may influence the response of tumor cells to therapies. Moreover, identifying novel BAKs that are inhibited by Ibr may form the rationale to target multiple signaling nodes to overcome resistance in MCL. Thank to Excellence Project 2023‐2027 of the DNBM, University of Verona funded by MUR; European Union, MUR and MSAL (PNRR-PNC‐E3‐2022‐23683266 PNC‐HLS‐DA-INNOVA and PNRR-TR1-2023-12378287) for funding support.
Paula Feldman, Miguel Fainstein, Viviana Siless
et al.
Anatomical trees play an important role in clinical diagnosis and treatment planning. Yet, accurately representing these structures poses significant challenges owing to their intricate and varied topology and geometry. Most existing methods to synthesize vasculature are rule based, and despite providing some degree of control and variation in the structures produced, they fail to capture the diversity and complexity of actual anatomical data. We developed a Recursive variational Neural Network (RvNN) that fully exploits the hierarchical organization of the vessel and learns a low-dimensional manifold encoding branch connectivity along with geometry features describing the target surface. After training, the RvNN latent space can be sampled to generate new vessel geometries. By leveraging the power of generative neural networks, we generate 3D models of blood vessels that are both accurate and diverse, which is crucial for medical and surgical training, hemodynamic simulations, and many other purposes. These results closely resemble real data, achieving high similarity in vessel radii, length, and tortuosity across various datasets, including those with aneurysms. To the best of our knowledge, this work is the first to utilize this technique for synthesizing blood vessels.
One-dimensional (1D) blood flow simulations are extensively used in cardiovascular research due to their computational efficiency and effectiveness in analyzing pulse wave dynamics. Despite their versatility and simplicity, there is a lack of a unified, step-by-step guide integrating theoretical derivations with practical implementation details. In this work, we summarize key components for comprehensive 1D blood flow simulations, including the derivation of reduced-order governing equations, the method of characteristics (Riemann invariants), a finite volume-based numerical scheme, boundary conditions (application of Riemann invariants for reflective/non-reflective and 3-element Windkessel outlet boundaries), junction treatments, verification of presented methodologies, and relevant practical applications. Additionally, we provide detailed step-by-step instructions for implementing the numerical scheme, applying boundary conditions, and treatment of junctions. By integrating rigorous theory with practical guidance for implementation, we seek to improve accessibility of 1D blood flow simulations. We anticipate that this guide will serve as a valuable resource and foundational reference for both novice and experienced researchers in cardiovascular modeling.
This study investigates the effects of geometric model reduction on blood flow simulations in the patient-specific descending aorta, followed by speed-accuracy trade-off analysis using 3D simulations. We demonstrate how wall shear stresses (WSS) can be reliably estimated for such realistic arteries using significantly faster simulations of highly idealized equivalent geometries, for any blood rheology model. CFD simulations (3D) are performed at two levels of geometry reduction employing realistic pulsatile inflow and pressure outlet boundary conditions and utilizing both Newtonian and non-Newtonian blood rheology models, including the one developed recently by Apostolidis and Beris. The first level of reduction does not retain effects due to local asymmetry but can approximate various flow parameters and patterns, while showing a significant computational speedup. However, further simplification to an idealized smooth geometry loses all information about the vortex structures and flow circulation. The non-Newtonian models retain more accuracy than the Newtonian models in geometry reductions, as quantified by correlations defined in this study. The idealized smooth geometry, combined with area correction, yields WSS estimates that closely approximate those of the actual artery. This study is expected to be applicable in geometric reductions (and speed enhancements) for more complex patient-specific 3D simulations while maintaining accuracy.
Recent discoveries have suggested that the promising avenue of using circulating tumor DNA (ctDNA) levels in blood samples provides reasonable accuracy for cancer monitoring, with extremely low burden on the patient's side. It is known that the presence of ctDNA can result from various mechanisms leading to DNA release from cells, such as apoptosis, necrosis or active secretion. One key idea in recent cancer monitoring studies is that monitoring the dynamics of ctDNA levels might be sufficient for early multi-cancer detection. This interesting idea has been turned into commercial products, e.g. in the company named GRAIL. In the present work, we propose to explore the use of Signature theory for detecting aggressive cancer tumors based on the analysis of blood samples. Our approach combines tools from continuous time Markov modelling for the dynamics of ctDNA levels in the blood, with Signature theory for building efficient testing procedures. Signature theory is a topic of growing interest in the Machine Learning community (see Chevyrev2016 and Fermanian2021), which is now recognised as a powerful feature extraction tool for irregularly sampled signals. The method proposed in the present paper is shown to correctly address the challenging problem of overcoming the inherent data scarsity due to the extremely small number of blood samples per patient. The relevance of our approach is illustrated with extensive numerical experiments that confirm the efficiency of the proposed pipeline.
T cell receptor (TCR) repertoires encode critical immunological signatures for autoimmune diseases, yet their clinical application remains limited by sequence sparsity and low witness rates. We developed EAMil, a multi-instance deep learning framework that leverages TCR sequencing data to diagnose systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) with exceptional accuracy. By integrating PrimeSeq feature extraction with ESMonehot encoding and enhanced gate attention mechanisms, our model achieved state-of-the-art performance with AUCs of 98.95% for SLE and 97.76% for RA. EAMil successfully identified disease-associated genes with over 90% concordance with established differential analyses and effectively distinguished disease-specific TCR genes. The model demonstrated robustness in classifying multiple disease categories, utilizing the SLEDAI score to stratify SLE patients by disease severity as well as to diagnose the site of damage in SLE patients, and effectively controlling for confounding factors such as age and gender. This interpretable framework for immune receptor analysis provides new insights for autoimmune disease detection and classification with broad potential clinical applications across immune-mediated conditions.
Objective To systematically evaluate the safety and efficacy of intraoperative cell salvage (IOCS) for placenta previa women undergoing cesarean section. Methods The PubMed, Web of Science, Embase, the Cochrane Library of clinical trials, China National Knowledge Infrastructure (CNKI) and Wan Fang databases were searched from inception to September 2022. The standardized mean differences (SMDs) or relative risk (RR) and 95% confidence intervals (CIs) and prediction intervals (PIs) regarding the comparison between the IOCS and allogeneic blood transfusion (ABT) groups were analyzed using R software Version 4.1.2 and Stata Version 12.0. Results Five RCTs and ten retrospective cohort studies were included in this meta-analysis. Analysis of cohort studies showed that compared with the ABT group, women with placenta previa who underwent IOCS had generally higher postoperative hemoglobin (Hb) (SMD, 0.626; 95% CI: 0.103 to 1.149; 95% PI: -1.320 to 2.572) and hematocrit (Hct) (SMD, 0.617; 95% CI: 0.130 to 1.104; 95% PI: -1.084 to 2.317) levels. In RCTs, we observed that placenta previa women undergoing IOCS were almost 72.7% less likely to suffer from adverse events AEs than the ABT group (RR, 0.273; 95% CI, 0.082 to 0.904). The difference in postoperative prothrombin time (PT), activated prothrombin time (APTT), fibrinogen (Fib) concentration, blood urea nitrogen (BUN) and creatinine (Cr) between the IOCS and ABT group did not reach statistical significance. Conclusion Women with placenta previa who undergo IOCS have higher postoperative Hb and Hct concentrations compared to those treated with ABT. IOCS has no major effects on postoperative coagulation parameters and renal function parameters. IOCS is associated with a significantly lower risk of transfusion-related AEs among women undergoing cesarean section.
Diseases of the blood and blood-forming organs, Medicine
Maria Jesús Serrano-Ripoll, Rocío Zamanillo-Campos, Sofía Mira Martínez
et al.
Objectives: This study aims to explore perspectives and opinions from healthcare professionals and people at risk of type 2 diabetes mellitus (T2DM) to inform the design of PREDIABETEXT, a new digital multifaceted intervention to prevent T2DM. Methods: in this qualitative study, we purposefully recruited 15 healthcare professionals (doctors and nurses) working in primary healthcare centers in Mallorca (Spain), and 15 of their patients at risk of T2DM (HbA1c 6–6.4%, and/or fasting plasma glucose 110–125 mg/dL). We collected the data through semi-structured phone interviews, using an interview guide aimed at gathering participants’ views about the two PREDIABETEXT proposed co-interventions (educational intervention targeted at professionals and delivered as an online training course, and behavioral intervention targeted at individuals at risk of T2DM and delivered using SMS short text messages). The interviews were audio-recorded, verbatim transcribed, and analyzed using a thematic analysis approach. Results: Primary healthcare professionals valued a prediabetes training course for standardizing care and supporting diabetes prevention. They preferred a blended format with content on early detection, intervention, and monitoring. They perceived SMS reminders to their patients as potentially useful for reinforcing guidelines and improving care. Individuals at risk of T2DM, who faced challenges maintaining a healthy diet and exercise routine, viewed SMS as potentially motivational and informative, offering suggestions on content, format, and frequency to enhance its intended benefits. Conclusions: This qualitative study provided valuable insights from primary care professionals and people at risk of T2DM that will inform the user-centered design of the PREDIABETEXT intervention.
Revaz D. Chachanidze, Othmane Aouane, Jens Harting
et al.
Margination, a fundamental process in which leukocytes migrate from the flowing blood to the vessel wall, is well-documented in physiology. However, it is still an open question on how the differences in cell size and stiffness of white and red cells contribute to this phenomenon. To investigate the specific influence of cell stiffness, we conduct experimental and numerical studies on the segregation of a binary mixture of artificially stiffened red blood cells within a suspension of healthy cells. The resulting distribution of stiffened cells within the channel is found to depend on the channel geometry, as demonstrated with slit, rectangular, and cylindrical cross-sections. Notably, an unexpected central peak in the distribution of stiffened RBCs, accompanied by fourfold peaks at the corners, emerges in agreement with simulations. Our results unveil a non-monotonic variation in segregation/margination concerning hematocrit and flow rate, challenging the prevailing belief that higher flow rates lead to enhanced margination.
Thien-Tam Nguyen, Davina Kasperski, Phat Kim Huynh
et al.
Currently, it is challenging to investigate aneurismal hemodynamics based on current in-vivo data such as Magnetic Resonance Imaging or Computed Tomography due to the limitations in both spatial and temporal resolutions. In this work, we investigate the use of modal analysis at various resolutions to examine its usefulness for analyzing blood flows in brain aneurysms. Two variants of Dynamic Mode Decomposition (DMD): (i) Hankel-DMD; and (ii) Optimized-DMD, are used to extract the time-dependent dynamics of blood flows during one cardiac cycle. First, high-resolution hemodynamic data in patient-specific aneurysms are obtained using Computational Fluid Dynamics. Second, the dynamics modes, along with their spatial amplitudes and temporal magnitudes are calculated using the DMD analysis. Third, an examination of DMD analyses using a range of spatial and temporal resolutions of hemodynamic data to validate the applicability of DMD for low-resolution data, similar to ones in clinical practices. Our results show that DMD is able to characterize the inflow jet dynamics by separating large-scale structures and flow instabilities even at low spatial and temporal resolutions. Its robustness in quantifying the flow dynamics using the energy spectrum is demonstrated across different resolutions in all aneurysms in our study population. Our work indicates that DMD can be used for analyzing blood flow patterns of brain aneurysms and is a promising tool to be explored in in-vivo.
Bastian Wittmann, Lukas Glandorf, Johannes C. Paetzold
et al.
Segmentation of blood vessels in murine cerebral 3D OCTA images is foundational for in vivo quantitative analysis of the effects of neurovascular disorders, such as stroke or Alzheimer's, on the vascular network. However, to accurately segment blood vessels with state-of-the-art deep learning methods, a vast amount of voxel-level annotations is required. Since cerebral 3D OCTA images are typically plagued by artifacts and generally have a low signal-to-noise ratio, acquiring manual annotations poses an especially cumbersome and time-consuming task. To alleviate the need for manual annotations, we propose utilizing synthetic data to supervise segmentation algorithms. To this end, we extract patches from vessel graphs and transform them into synthetic cerebral 3D OCTA images paired with their matching ground truth labels by simulating the most dominant 3D OCTA artifacts. In extensive experiments, we demonstrate that our approach achieves competitive results, enabling annotation-free blood vessel segmentation in cerebral 3D OCTA images.
Objetivo: Relatar uma apresentação rara de recaída de linfoma de Hodgkin (LH) em sítio extranodal atípico. Material e métodos: As informações deste trabalho foram obtidas da revisão do prontuário e revisão da literatura. Relato de caso: Homem, 50 anos, portador do vírus HIV em uso de TARV, com carga viral indetectável, e diagnóstico recente de tuberculose pulmonar em tratamento, foi diagnosticado em novembro de 2021 com LH clássico subtipo esclerose nodular estadio IVB, com infiltração neoplásica de fígado, baço, pulmão e medula óssea, confirmados por biópsia. Iniciou tratamento quimioterápico em janeiro/2022 com ABVD, sendo omitida bleomicina após dois ciclos, devido a PET-CT interim evidenciando resposta completa. Finalizou o tratamento de seis ciclos em julho/2022 com PET-CT final sem lesões hipermetabólicas. Dois meses após o fim do tratamento, apresentou crise convulsiva inédita. Ressonância magnética de encéfalo evidenciou lesão heterogênea subcortical parieto-occipital à esquerda, sendo aventada como principal hipótese causas infecciosas/granulomatosas. Realizada internação e tratamento empírico para infecções oportunistas devido aos achados de meningite linfomononuclear e hiperproteinorraquia. Apesar do tratamento, o paciente recorreu com novas crises convulsivas e manutenção da lesão encefálica, sendo procedida a realização de biópsia cerebral com achado de parênquima com denso infiltrado constituído, na sua maioria, por células linfoides maduras. Em meio às células linfoides, de aspecto reacional, haviam várias células volumosas, algumas binucleadas, de citoplasma basófilo, núcleos ovalados, cromatina granular e nucléolo único e proeminente, levemente eosinófilo (células de Reed-Sternberg) com expressão forte de CD30 e, em menor quantidade, CD15 e EBV-LMP, como também fraca expressão para PAX-5, compatível com infiltração cerebral por LH. O paciente evoluiu com rebaixamento de nível de consciência e foi a óbito antes do início da terapia de resgate para a neoplasia hematológica. Discussão e conclusão: O LH é uma neoplasia linfoide caracterizada por células de origem B neoplásicas, por vezes apresentando uma morfologia de célula binucleada com nucléolos proeminentes, denominadas de células de Reed-Sternberg, entremeadas por infiltrado inflamatório reacional. Corresponde a aproximadamente 10% dos casos de linfoma e é mais comum em adultos jovens. Pode estar associada ao HIV, condição na qual praticamente todos os casos também estão associados à infecção pelo EBV. A apresentação típica envolve linfonodomegalias, com predomínio em região cervical, de crescimento progressivo lento, o envolvimento de linfonodos mediastinais também é bastante comum, ocorrendo em mais da metade dos pacientes, não sendo incomum a presença de grandes massas mediastinais encontradas ao exame de imagem. Também é comum a ocorrência de sintomas constitucionais, como febre, perda de peso e sudorese noturna, além de prurido generalizado, que pode preceder o diagnóstico do linfoma em meses. O envolvimento direto do sistema nervoso central é raro, ocorrendo em menos de 0,5% dos casos. O relato objetiva enfatizar que, apesar de apresentação rara, a infiltração cerebral pode ocorrer e deve ser considerada no diagnóstico diferencial de lesões encefálicas, em especial em paciente com histórico pessoal da doença, com fins de não se atrasar a propedêutica adequada para o diagnóstico e tratamento.
Background and Objectives: The presence of irregular red cell antibodies in the serum of pregnant women is named maternal alloimmunization. Sensitizing events such as traumatic delivery and miscarriage can cause fetomaternal hemorrhage, increasing the risk of alloimmunization. Our objective was to estimate the prevalence of red cell alloantibodies among multigravidae and to study its association with those cases with bad obstetric history (BOH). Methods: This was a 1-year cross-sectional study done on 900 multigravidae in the Department of Transfusion Medicine, Medical College, Thiruvananthapuram. We excluded patients with a history of blood transfusion. Results: Out of 900 subjects, 49 (5.4%) were alloimmunized. The RhD-positive cases were predominant (77.4%), while 203 (22.6%) were RhD negative. Among the D-positive women, the prevalence of alloimmunization was 1.3%. The most frequent alloantibody identified was anti-D, followed by anti-G (18.4%). The alloantibodies identified among RhD-positive were anti-E, anti-c, anti-C, anti-Fya, anti-M, anti-Jka, and anti-Leb. We identified a total of 58 clinically significant alloantibodies. Thus, Rh blood group alloantibodies constitute 93.1% and the remaining 6.9% by other blood groups. Out of 159 cases with BOH, 17 were alloimmunized, and there was a statistically significant association between them (P = 0.001). Conclusion: Alloantibodies potentially causing hemolytic disease of the fetus and newborn were found in D-negative (19.7%) and D-positive (1.3%) mothers. The clinicians should change the current practice of screening only D-negative pregnant women, and antibody screening should be offered to all females or at least those with BOH. We may thus ensure early detection and periodic monitoring of all alloimmunized women.
Objetivo: Descrever a experiência de organizar a III Jornada de Hematologia e Hemoterapia em Belém do Pará. Relato da experiência: Promover um evento de cunho cientifico sempre é necessário para atualizações na especialidade, sendo assim surgiu a necessidade de organizar um evento que abordasse a Hematologia e Hemoterapia, e abrangesse profissionais da área da saúde, incluindo médicos, enfermeiros, técnicos de enfermagem e acadêmicos das áreas da saúde. A III Jornada de Hematologia e Hemoterapia contou com submissão e apresentação de trabalhos científicos, sendo esses selecionados se relevantes para a comunidade cientifica. Além do mais, a programação incluiu diversos assuntos importantes, como: manejo da anemia na jornada do paciente, Patient Blood Management (PBM), tromboelastometria, tromboelastograma na cirurgia, tipagem ABO, orientações para demandas imunohematológicas das agências transfusionais e captação de doadores.O intuito principal era difundir conhecimentos e novas atualizações sobre a área da Hematologia, com isso o Hospital das Clínicas Gaspar Viana (HCGV) promoveu o evento em conjunto com a Liga Acadêmica de Hematologia do Pará (LAHEPA). O evento ocorreu no mês de Junho, com 2 dias seguidos de evento, e contando com a participação de palestrantes renomados de outros estados. Destarte, a jornada proporcionou uma boa interação entre acadêmicos e profissionais da área da saúde. Discussão: Entende-se que atualizações são mais que necessárias em qualquer área, principalmente naquelas relacionadas a saúde, pois protocolos mudam com frequência, novas drogas surgem, além de novos estudos sobre a especialidade. Sendo assim, analisou-se a necessidade de promover e organizar um evento que atualizasse profissionais e acadêmicos sobre a Hematologia e Hemoterapia, incluindo os temas mais diversos da especialidade. Com isso, recrutou-se a liga acadêmica para auxiliar no processo de organização, pois um dos seus pilares principais é aproximar os acadêmicos de suas temáticas de interesse e promover maior visualização dos cenários reais. Dessa maneira, o evento foi promovido e elogiado pelos participantes pela amplitude dos assuntos debatidos. Conclusão: Diante do exposto, é essencial a promoção de eventos que atualizem frequentemente médicos, enfermeiros e técnicos de enfermagem à respeito de mudanças nas mais diversas áreas. A partir disso, a III Jornada de Hematolgia e Hemoterapia ampliou debates sobre anemias, tromboelastometria e mais temáticas, com o apoio da liga acadêmica e das médicas hematologistas do hospital que proporcionou o evento. surgem e tem por objetivo promover um contato maior dos participantes com a temática. Com isso, conclui-se a importância de aulas para profissionais da área da saúde como forma de incentivo para a busca de atualizações, e o evento promovido atuou justamente nesse objetivo de atualização e integração entre médicos, enfermeiros e acadêmicos da área da saúde.
Many methods to computationally predict red blood cell damage have been introduced, and among these are Lagrangian methods which track the cells along their pathlines. Such methods typically do not explicitly include cell-cell interactions. Due to the high volume fraction of red blood cells in blood, these interactions could impact cell mechanics and thus, the amount of damage caused by the flow. To investigate this question, cell-resolved simulations of red blood cells in shear flow were performed for multiple interacting cells, as well as for single cells in unbounded flow at an effective viscosity. Simulations run without adjusting the bulk viscosity produced larger errors unilaterally and were not considered further for comparison. We show that a periodic box containing at least 8 cells and a spherical harmonic of degree larger than 10 are necessary to produce converged higher-order statistics. The maximum difference between the single-cell and multiple-cell cases in terms of peak strain was 3.7%. To achieve this agreement, one must use the whole blood viscosity and average over multiple cell orientations when adopting a single-cell simulation approach. There were some differences between the two models in terms of average strain (maximum difference of 13%). However, given the accuracy of the single-cell approach in predicting the maximum strain, which is useful in hemolysis prediction, and its computational cost that is orders of magnitude less than the multiple-cell approach, one may use it as an affordable cell-resolved approach for hemolysis prediction.