College of Staten Island

Staten Island is less developed than the other boroughs of New York city, however outcrops of rock and surface sediment are limited, making interpretation of its geologic history challenging. When small areas of sediment are exposed, they can be used to improve our understanding of changes in sediment erosion and deposition over time. In this study of two small temporary outcrops, the beds of sediment were logged in the field and samples were collected for textural and compositional analyses. The results were interpreted in the context of previous work on similar exposures nearby. The sediments were found to be sands and gravels of fluvioglacial origin, containing reworked sediments of both the Pliocene Pensauken Formation, and older Triassic rocks of the Newark Basin. It is likely that they were deposited on an outwash plain during the Illinoian glaciation. They were deposited in a topographic low, directly overlying Cretaceous sedimentary rocks, but adjacent to sediments of the Pensauken Formation which had in turn been deposited as an earlier valley fill. This interpretation solves an apparent disagreement between previous studies, by illustrating how both the Pensauken Formation and later fluvioglacial sediments can be exposed over a small area.

Computer vision and deep learning are essential in diverse fields such as autonomous driving, medical imaging, face recognition, and object detection. However, enhancing low-light remote sensing images remains challenging for both research and real-world applications. Low illumination degrades image quality due to sensor limitations and environmental factors, weakening visual fidelity and reducing performance in vision tasks. Common issues such as insufficient lighting, backlighting, and limited exposure create low contrast, heavy shadows, and poor visibility, particularly at night. We propose QWR-Dec-Net, a quaternion-based Retinex decomposition network tailored for low-light image enhancement. QWR-Dec-Net consists of two key modules: a decomposition module that separates illumination and reflectance, and a denoising module that fuses a quaternion holistic color representation with wavelet multi-frequency information. This structure jointly improves color constancy and noise suppression. Experiments on low-light remote sensing datasets (LSCIDMR and UCMerced) show that QWR-Dec-Net outperforms current methods in PSNR, SSIM, LPIPS, and classification accuracy. The model’s accurate illumination estimation and stable reflectance make it well-suited for remote sensing tasks such as object detection, video surveillance, precision agriculture, and autonomous navigation.

Gaze-based object manipulation intention inference is pivotal to natural and intuitive human-robot interaction. Existing methods confine spatial regularity to independent object selection and treat temporal regularity only as sequential order, thus spatio-temporal gaze regularities remain insufficiently exploited and lack a unified treatment. The objective of this study is to statistically analyze, model, and integrate gaze regularities within a unified probabilistic framework for inference. Accordingly, we propose a spatio-temporal gaze regularities guided dynamic fuzzy Bayesian network (DFBN) for intention inference. We statistically analyze the spatial gaze regularity as mutual exclusion and co-occurrence in joint object selection patterns, and the temporal regularity as duration-dependent attention with sequential dependencies. The regularities are modeled into probabilistic form, with spatio regularity modeled by autoregressive logistic regression and temporal regularity modeled by a Fuzzy Gaze-LSTM that fuses gaze duration with sequential order. These probabilistic models are fused into a likelihood modifier to generate interpretable posterior probabilities of intention, integrating a Bayesian network and sequential inference. Cross-dataset evaluations indicate stable and high performance. DFBN attains 96.47 1.75% accuracy and 96.41 1.85% F1 score, maintains accuracy on error sequences, and generalizes across younger and older groups, supporting robust intention inference. This study has the potential to inform other human-robot interaction assistance strategies by serving as intuitive gaze regularity cues.

Autonomous driving systems usually depend on RGB images but face difficulties in low-visibility conditions. Thermal infrared (TIR) imaging offers an alternative that performs well in such environments. However, TIR is hindered by low contrast, indistinct boundaries, and the absence of color, making it less compatible with standard algorithms. Recent advances in thermal infrared colorization are promising, but they face challenges such as semantic distortions, frame instability, and issues with small objects. We introduce the frame-based Weber-law-driven Generative Adversarial Network (FWGAN), a new model that converts thermal images into visible-spectrum images. FWGAN improves object recognition by maintaining textures, increasing temporal consistency, and enabling reliable detection in different scenarios. Tests on benchmark datasets like FLIR and KAIST show that FWGAN's enhancements in semantic accuracy, object detection, and visual quality significantly push forward autonomous driving technology.

Respiratory disease presents major global health challenges, and its early detection could save numerous lives, prompting the demand for accurate and non-intrusive diagnostic methods. This study presents a multi-model machine learning framework for classifying respiratory diseases (COPD, Bronchiolitis, Pneumonia, URTI, and Healthy) using audio data from the Respiratory Sound Database. We extract Mel frequency cepstral coefficients (MFCCs) and apply data augmentation to address class imbalance. We propose a novel hybrid model, ConvMambda, and compare its performance against other popular classification models, including 1D CNN, random forest, SVM, and LSTM. The ConvMamba model achieves the highest test accuracy (90.77%) and macro F1-score (0.8932), outperforming 1D-CNN (90% accuracy) and random forest (82% accuracy), SVM (68% accuracy), and LSTM (68% accuracy). We analyze the performance across classes, highlighting challenges with minority classes (Bronchiolitis, URTI) and propose strategies to mitigate imbalance. Our findings demonstrate the potential of audio-based diagnostics, with the ConvMamba model offering the superior performance for respiratory diseases classification.

HighlightsWe have created a novel Taxol derivative that induces anti-tumor immunity in glioblastoma (GBM) tumors without triggering autoimmune reactions, leading to long-term remission in a syngeneic mouse model of GBM.➢Creating anti-tumor immunity to treat cancers has been gaining ground but this approach also involves the risk of triggering systemic inflammation and autoimmunity.➢Using a calculated strategy, we have created a novel Taxol derivative, STO-1, that induces anti-tumor immunity within glioblastoma (GBM) tumors.➢Most notably, STO-1 does not trigger systemic inflammation and autoimmune reactions.➢As an immunotherapeutic, STO-1 accomplishes sustained remission in a syngeneic mouse model of GBM.Reprogramming of macrophages into the inflammatory state (also known as M1) is currently considered as an effective way of eliminating cancer cells, but systemic deployment of this strategy is likely to induce dangerous autoimmune reactions. Consequently, converting immunosuppressive M2-type macrophages into M1 systemically is not a safe and effective therapeutic approach against cancer. Through cleavable covalent linking of curcumin to the chemotherapeutic agent Paclitaxel (Taxol), we have created a novel prodrug (STO-1) that, upon intravenous delivery, selectively reprograms tumor-associated microglia and macrophages (TAMs) and eliminates glioblastoma (GBM) without triggering autoimmunity. Demonstrating its therapeutic efficacy, prolonged treatment of six orthotopic GBM-bearing mice with STO-1 resulted in 67% long-term survival, with three surviving mice exhibiting complete tumor clearance and one displaying minimal residual disease, as confirmed by high-resolution ex vivo T2-weighted MRI 85 days after tumor inoculation. In contrast, the vehicle-treated mice displayed extensive intracranial tumors with edema and hemorrhage. Mechanistically, scRNA-seq analysis indicated induction of multiple M1-associated transcripts (ccrl2, cxcl9, ccr2, ccl5) consistent with robust TAMs reprogramming. In striking contrast to the M2⟶M1 reprogramming of TAMs, M1-type macrophages were suppressed in the spleens of STO-1-treated cancer-free mice. Therefore, STO-1 induces selective anti-tumor immunity and GBM elimination without triggering systemic autoimmune reactions.

Transspinal stimulation with direct current or at intensities and frequencies that produces intermittent depolarization of motoneurons can be an adjunct treatment strategy for spasticity and recovery of movement in persons with spinal cord injury (SCI). The main objective of this study was to assess changes in neurophysiological biomarkers after multiple sessions of transspinal direct current stimulation (tsDCS) in people with and without SCI. Nine SCI and 10 healthy control subjects received daily cathodal tsDCS over the Thoracic 10 vertebra while supine with an average stimulation intensity of 2.28 ± 0.02mA. SCI and healthy control subjects received an average of 15 and 10 stimulation sessions, respectively. Before and 1-2 days post intervention, we assessed changes in soleus H-reflex recruitment input-output curves, homosynaptic depression and postactivation depression. tsDCS did not alter the strength of homosynaptic depression in any subject group but reversed postactivation depression to facilitation in AIS D subjects. tsDCS resulted in depression of reflex excitability in both subject groups. No significant changes in clinically assessed hyperreflexia were observed in SCI subjects. The results suggest decreased reflex hyperexcitability without recovery of spinal inhibitory control in the injured human spinal cord after multiple sessions of tsDCS. More systematic investigations on reorganization of spinal and cortical interneuronal networks are needed to delineate the tsDCS-induced neuroplasticity in people with SCI to develop effective treatments.

Changes in climate and land use land cover are a widely recognized threat to the stability of natural species assemblages’ composition and biodiversity. Species-specific responses to these changes can result in a rearrangement of the species composition of assemblages, altering the stability, resilience, and functioning of the ecosystems of which these assemblages are a part. We assessed the relationship between the rate of change in avian species richness and assemblage dissimilarity and the rate of change in climatic and land use land cover variables across 30 years in five ecoregions of the United States. Areas where effects of changing land use land cover and/or climate were most strongly felt were high elevations and latitudes. Rates of species replacement and loss were associated with changing environmental factors in opposite directions. Rates of change in biodiversity were more strongly predicted by rates of change in land use land cover than by rates of change in climate. For the species assemblages studied here, rapidly changing climate and/ or land use land cover was more strongly affecting total assemblage dissimilarity patterns than species richness even though species richness has received much more research attention. Trends in multiple biodiversity indices capture multiple levels of action (richness vs. assemblage dissimilarity). A study that integrates these allows us to observe the complex and changing interrelationships between biodiversity and the environment (climate and land cover), and thus, plan effectively for preservation of processes that generate patterns of biodiversity.

Classical and quantum technologies have traditionally been viewed as orthogonal, with classical systems being deterministic and quantum systems inherently probabilistic. This distinction hinders the development of a scalable quantum internet even as the global internet continues expanding. We report a classical-decisive quantum internet architecture in which the integration of quantum information into advanced photonic technologies enables efficient entanglement distribution over a commercially deployed fiber network. On-chip precise synchronization between classical headers and quantum payloads enables dynamic routing and networking of high-fidelity entanglement guided by classical light. The quantum states are preserved through real-time error mitigation, relying solely on classical signal readout without disturbing quantum information. These classical-decisive features demonstrate a practical path to a scalable quantum internet using existing network infrastructure and operating systems.