Platelets are often described in simple terms as small anucleate cells that mediate hemostasis, but studies over more than half a century have shown that circulating platelets are heterogeneous in size, density, age, and functional responses. These subtypes not only contribute to normal hemostasis but also play pivotal roles in the pathogenesis of diverse diseases, including cardiovascular, inflammatory, and malignant disorders. Accumulating evidence indicates that alterations in specific platelet subtypes are closely linked to disease onset, progression, and severity, underscoring their importance as both biomarkers and potential therapeutic targets. Current clinical assessments of platelet status rely primarily on platelet count and mean platelet volume (MPV) as part of routine complete blood count analysis. However, these global measures often fail to capture differences in platelet subtypes, which may remain undetected despite their significant contribution to disease pathology. This gap highlights the necessity of moving beyond conventional metrics toward a more nuanced understanding of platelet heterogeneity and its clinical implications. In this review, we discuss the diversity of platelet subpopulations and their roles in health and disease, emphasizing how specific subsets contribute to divergent pathological mechanisms. We also highlight emerging strategies that target defined platelet subpopulations, illustrating how this knowledge could pave the way for more precise diagnostic and therapeutic approaches.
Publications by Year: 2025
2025
Background/Objectives: Retinitis pigmentosa (RP) is a progressive hereditary retinal disorder that frequently leads to vision loss, with cystoid macular edema (CME) occurring in approximately 10-50% of affected patients. Early detection of CME is crucial for timely intervention, yet most existing studies lack longitudinal data capable of capturing subtle disease progression. Methods: We propose a deep learning-based framework utilizing longitudinal optical coherence tomography (OCT) imaging for early detection of CME in patients with RP. A total of 2280 longitudinal OCT images were preprocessed using denoising and data augmentation techniques. Multiple pre-trained deep learning architectures were evaluated using a patient-wise data split to ensure robust performance assessment. Results: Among the evaluated models, ResNet-34 achieved the best performance, with an accuracy of 98.68%, specificity of 99.45%, and an F1-score of 98.36%. Conclusions: These results demonstrate the potential of longitudinal OCT-based artificial intelligence as a reliable, non-invasive screening tool for early CME detection in RP. To the best of our knowledge, this study is the first to leverage longitudinal OCT data for AI-driven CME prediction in this patient population.
BACKGROUND/OBJECTIVES: Uveal melanoma is a rare but aggressive intraocular malignancy that metastasizes in up to half of patients, most commonly to the liver, despite effective local treatment. In the absence of robust evidence, there are no standardized guidelines for post-treatment surveillance, resulting in wide variation in imaging modalities, frequency, and duration across physicians and institutions. This study aimed to develop expert consensus recommendations for surveillance strategies in patients with uveal melanoma.
METHODS: A modified Delphi method was conducted across three iterative survey rounds between September 2024 and February 2025 using an online platform. Panelists included medical oncologists, ocular oncologists, radiologists, and surgical oncologists from North America. A multidisciplinary steering committee developed statements addressing risk-based surveillance using both molecular and clinical prognostic factors, including gene expression profiling (GEP) and PRAME status. Consensus was defined a priori as ≥70% of panelists rating a statement 7-9 on a 9-point Likert scale.
RESULTS: Forty-nine experts were invited, and 41 completed at least one survey round. The panel represented 17 U.S. states, Washington, D.C., and two Canadian provinces. Twelve statements reached stable consensus, including recommendations for imaging modality, frequency, and duration in intermediate- and high-risk patients. Although there was agreement that low-risk patients warrant surveillance, no consensus was reached on the optimal approach for this group.
CONCLUSIONS: This is the first study to provide consensus-based guidance incorporating GEP and PRAME status into surveillance recommendations for uveal melanoma, offering a standardized framework to guide clinical practice and future research.
OBJECTIVES: Anterior cruciate ligament (ACL) injuries can go undetected in the initial encounter, and delayed diagnosis can cause instability and an elevated risk of osteoarthritis. We hypothesized that dynamic testing under handheld ultrasound assessments of ACL insufficiency offers high diagnostic accuracy and reliability.
METHODS: Ten fresh-frozen knee specimens were evaluated in three conditions: (1) intact ACL, (2) complete ACL deficiency, and (3) ACL and Anterolateral ligament (ALL) deficiency. Dynamic testing under ultrasound was performed while simulating the Lachman test (tibial anterior translation with 0 and 100N, at 20 degrees knee flexion) and pivot shift test (internal rotation torque with 0 and 10Nm). The probe was placed at the anterior medial (AM) and lateral (AL) joint line, and anterior translation of the tibia (ΔD) relative to the femur was calculated. Intra-observer and inter-observer reliability were calculated, and receiver operating characteristic (ROC) curve analysis was performed for an ideal cutoff point for detecting knee instability.
RESULTS: From the AM view, ACL-deficient and ACL+ALL-deficient conditions significantly increased ΔD of the tibia during anterior loading compared to the intact ACL condition. The median ΔD values of the tibia in AM view with anterior drawer load were 0.64 (-0.10, 2.67) mm (intact ACL), 4.76 (2.46, 7.21) mm (ACL deficiency), and 3.88 (2.08, 7.23) mm (ACL+ALL deficiencies). The area under the ROC curve identifying ACL deficiency was 0.89 (95%CI, 0.66-0.97) in the AM view with anterior loading simulating the Lachman test. The optimal cut-off value to distinguish between the intact and ACL-deficient condition for anterior tibial translation with loading was 2.6 mm (sensitivity=80%, specificity=90%).
CONCLUSION: Dynamic examination of the knee using portable handheld ultrasound from an AM view has high sensitivity and specificity in diagnosing ACL injury.
Visual information reaches the cortex via multiple pathways, most notably the magnocellular (M) and parvocellular (P) streams. How their inputs are integrated remains unclear. We previously documented the case of Davida, a young woman who perceives high-contrast, sharply bounded 2D shapes as rotated or mirror-reversed, while he perception normalizes when stimuli are blurred, low in contrast, or brief. Here, we first review previous findings and propose that her disorder may reflect abnormally strong P-on-M inhibition. We then present three studies supporting this account: two demonstrate abnormally strong P-on-M inhibition in Davida, and a third shows that the degree of P-M imbalance predicts mirror-image discrimination difficulty in neurotypical individuals. Finally, re-analysis of published data suggests that distinct aspects of shape orientation are differentially sensitive to P-M imbalance. Together, these findings highlight the crucial role of dynamic P-M interactions in shaping visual perception.
INTRODUCTION: Simulation training is an essential component of transesophageal echocardiography (TEE) education, particularly for novice learners. However, a critical limitation of current mannequin-based TEE simulators is their inability to emphasize the spatial relationship between the probe's position and the resulting ultrasound images. This limitation, referred to as the "black box" effect, can make it difficult to understand how probe manipulation affects image acquisition, including probe orientation, beam trajectory, and the intersection of the ultrasound beam with cardiac structures. As a result, educators may struggle to convey these critical spatial concepts using existing simulation tools.
METHODS: A set of 3D-printed TEE training models was first developed to simulate the steps involved in TEE scanning procedures. They also provide beginners with an intuitive visual representation and a mental rotation process, enhancing their spatial awareness. Then, a pre-post pilot study was conducted among third-year clinical anesthesia (CA-3) residents. Three aspects of teaching quality were evaluated before and after the simulation training with 3D-printed models: TEE view-related knowledge, image quality on a mannequin-based simulator, and learners' confidence.
RESULTS: A total of 10 residents were included in this study. The knowledge assessment score significantly improved after the session (56.0 ± 22.60 vs 93.5 ± 6.26, P < .001). Image quality scores of 8 out of 12 required TEE views were also improved. Using the models positively influenced trainees' confidence in understanding and memorizing basic TEE principles: omniplane angles and corresponding TEE views [1.00 (0.75, 1.25) vs 0.00 (-1.00, 1.00), P < .01]; orientation interpretation [1.00 (1.00, 2.00) vs 0.00 (-1.00, 1.00), P < .01].
CONCLUSIONS: A simulation training session with a set of 3D-printed TEE teaching models significantly improved TEE view-related knowledge, image quality, and learners' confidence among CA-3 residents.
[This corrects the article DOI: 10.1016/j.jlb.2025.100442.].
The O1 serogroup of Vibrio cholerae has caused all cholera pandemics and for over a century V. cholerae O1 outbreak strains have been characterized by their serotype. The two V. cholerae serotypes differ by the presence (Ogawa) or absence (Inaba) of methylation of the terminal sugar on the lipopolysaccharide O1-antigen. Serotype switching often occurs during epidemics and has historically been attributed to the pathogen adapting to immune pressures. Here we address the impact of serotype on V. cholerae pathogenicity using otherwise isogenic Ogawa and Inaba versions of several clinical V. cholerae O1 isolates. Our findings indicate that O1 antigen methylation in Ogawa strains promotes V. cholerae colonization, infectivity and resistance to antimicrobial peptides. We propose that methylation of the O1 antigen elevates colonization by shielding the bacterium from cationic antimicrobial peptides at the pH of the small intestine. These observations provide insights into the biological significance of the V. cholerae O1 serotypes.
Neuroimaging is an essential part of epilepsy diagnosis and management. In this review, we examine the shift in our understanding of epilepsy from a focal, lesional-centric process to a broader, network-based disorder and how it has affected the acquisition and interpretation of neuroimaging for epilepsy diagnosis and management. We discuss the current clinical practice of neuroimaging in epilepsy, both in terms of acquisitions and interpretation, through computed tomography, magnetic resonance imaging (including structural, diffusion, functional, and high-field), magnetoencephalography, and nuclear and metabolic imaging. Additionally, we look ahead to some of the latest research advances in both the acquisition and analysis of these different neuroimaging modalities to help promote the eventual goal of a consolidated, multi-modality imaging approach to understanding, diagnosing, and treating epilepsy.
BACKGROUND: Postoperative delirium is a serious yet underrecognized complication affecting diverse surgical populations, with profound implications for morbidity, mortality, and long-term cognitive function. Its prediction remains imprecise, and screening practices vary widely.
METHODS: We conducted a retrospective analysis of the 2021-2023 data from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). The study included 217,783 adult surgical patients with documented delirium assessment. Patients were categorized as delirium, non-delirium, or unscreened, with an additional 2.7 million unscreened patients analyzed to assess selection bias in screening. The primary outcome was the incidence of delirium. Other outcomes included surgical and medical complications, mortality, length of stay, functional decline, discharge destination and conditions, and perioperative lab values. Multivariable regression models were used to identify independent predictors of delirium and its associated outcomes.
FINDINGS: Delirium screening was performed exclusively in patients aged ≥75 years, accounting for 7.3% of all surgical patients. Screening rates declined with advancing age (44% of patients aged 90+ vs 56% aged 75-79), while the proportion screening positive increased sharply (3.1% at 75-79 years vs 12.8% at ≥90 years). Delirium occurred in 10.6% (n = 23,100) of screened patients. Compared with non-delirium patients, those with delirium were older (mean 81.3 vs 79.9 years, p < 0.0001), more functionally dependent, and had higher ASA class and comorbidity burden. Dementia (37% vs 7.9%, p < 0.0001), recent falls (40% vs 18%, p < 0.0001), and urgent/emergency surgery (55% vs 26%, p < 0.0001) were strongly associated. Each 10-min increase in operative time was seen to independently be associated with 2.3% raised odds of delirium (p < 0.0001). Delirium was independently associated with higher 30-day mortality (OR 3.2, 95% CI 2.9-3.5), reoperation (OR 2.3, 95% CI 2.1-2.5), surgical complications (OR 1.8, 95% CI 1.6-1.9), loss of independence (OR 1.6, 95% CI 1.5-1.7), and reduced odds of home discharge (OR 0.5, 95% CI 0.4-0.5).
INTERPRETATION: Postoperative delirium is an independent predictor of adverse surgical outcomes yet remains substantially under-screened, with disparities across patient groups. These findings underscore the need for standardized, routine screening and targeted prevention strategies to improve perioperative care.
FUNDING: No funding was received for this study.