Explore the vast range of titles available.

We present a novel x-ray lithography based micromanufacturing methodology that offers scalable manufacturing of high precision optical components. It is accomplished through simultaneous usage of multiple stencil masks made moveable with respect to one another through custom made micromotion stages. The range of spectral flux reaching the sample surface at the LiMiNT micro/nanomanufacturing facility of Singapore Synchrotron Light Source (SSLS) is about 2 keV to 10 keV, offering substantial photon energy to carry out deep x-ray lithography. In this energy range, x-rays penetrate through resist materials with only little scattering. The highly collimated rectangular beam architecture of the x-ray source enables a full 4″ wafer scale fabrication. Precise control of dose deposited offers determined chain scission in the polymer to required depth enabling 1800 discrete gray levels in a chip of area 20 mm 2 and with more than 2000 within our reach. Due to its parallel processing capability, our methodology serves as a promising candidate to fabricate micro/nano components of optical quality on a large scale to cater for industrial requirements. Usage of these fine components in analytical devices such as spectrometers and multispectral imagers transforms their architecture and shrinks their size to pocket dimension. It also reduces their complexity and increases affordability while also expanding their application areas. Consequently, equipment based on these devices is made available and affordable for consumers and businesses expanding the horizon of analytical applications. Mass manufacturing is especially vital when these devices are to be sold in large quantities especially as components for original equipment manufacturers (OEM), which has also been demonstrated through our work. Furthermore, we also substantially improve the quality of the micro-components fabricated, 3D architecture generated, throughput, capability and availability for industrial application. Manufacturing 1800 Gray levels or more through other competing techniques is either limited due to multiple process steps involved or due to unacceptably long time required owing to their pencil beam architecture. Our manufacturing technique presented here overcomes both these shortcomings in terms of the maximum number of gray levels that can be generated, and the time required to generate the same.

A 3D silicon micromachining method based on proton beam writing combined with electrochemical anodization of p-type silicon enables fabrication of mid-infrared photonic crystals made of silicon and porous silicon. Here, example structures of silicon 1D and 2D photonic crystals are demonstrated. Progress and problems of fabricating 3D photonic crystals made of silicon are discussed. The strategy of fabricating photonic crystals purely made of porous silicon, and the characterization method of all these mid-infrared structures, are discussed. Due to the flexibility of this fabrication method, photonic devices and integrated photonic circuits may be built on a single chip, for which two 2D silicon photonic crystals with one on top of the other are demonstrated.

Digital health data are multimodal and high-dimensional. A patient’s health state can be characterized by a multitude of signals including medical imaging, clinical variables, genome sequencing, conversations between clinicians and patients, and continuous signals from wearables, among others. This high volume, personalized data stream aggregated over patients’ lives has spurred interest in developing new artificial intelligence (AI) models for higher-precision diagnosis, prognosis, and tracking. While the promise of these algorithms is undeniable, their dissemination and adoption have been slow, owing partially to unpredictable AI model performance once deployed in the real world. We posit that one of the rate-limiting factors in developing algorithms that generalize to real-world scenarios is the very attribute that makes the data exciting—their high-dimensional nature. This paper considers how the large number of features in vast digital health data can challenge the development of robust AI models—a phenomenon known as “the curse of dimensionality” in statistical learning theory. We provide an overview of the curse of dimensionality in the context of digital health, demonstrate how it can negatively impact out-of-sample performance, and highlight important considerations for researchers and algorithm designers.

With the increased demand for wearable sensors, image representations—such as persistence images and Gramian angular fields—transformed from time-series data have been investigated to address challenges in wearables arising from physiological variations, sensor noise, and limitations in capturing contextual information. To preserve the lightweight structural design of models, knowledge distillation (KD) has also been employed alongside image representations during training to distill smaller and more efficient models. Although image representations play a key role in providing richer and more informative features in training a model, their effectiveness within the KD framework has not been thoroughly explored. In this paper, we focus on image representation-driven KD to investigate whether these representations can provide useful knowledge leading to improved time-series interpretation in activity classification tasks. We explore the benefits of integrating image representations into KD, and we analyze the interplay between representation richness and model compactness with different combinations of teacher and student networks. We also introduce diverse KD strategies to utilize image representations, and we demonstrate the strategies with various perspectives, such as analysis of noises, generalizability, and compatibility, across datasets of varying scales to obtain comprehensive and insightful observations. These offer valuable insights for designing efficient and high-performance wearable sensor-based systems.

Patients with unresectable Colorectal Liver Metastases (CRLM) are increasingly being managed using Hepatic Artery Based Therapies (HAT), including Hepatic Arterial Infusion (HAI), Radioembolization (RE), and Transcatheter Arterial Chemoembolization (TACE). Limited data is available on the comparative effectiveness of these options. We hypothesized that outcomes in terms of survival and toxicity were equivalent across the three strategies. A meta-analysis was performed using a prospectively registered search strategy at PROSPERO (CRD42013003861) that utilized studies from PubMed (2003-2013). Primary outcome was median overall survival (OS). Secondary outcomes were treatment toxicity, tumor response, and conversion of the tumor to resectable. Additional covariates included prior or concurrent systemic therapy. Of 491 studies screened, 90 were selected for analyses-52 (n = 3,000 patients) HAI, 24 (n = 1,268) RE, 14 (n = 1,038) TACE. The median OS (95% CI) for patients receiving HAT in the first-line were RE 29.4 vs. HAI 21.4 vs. TACE 15.2 months (p = 0.97, 0.69 respectively). For patients failing at least one line of prior systemic therapy, the survival outcomes were TACE 21.3 (20.6-22.4) months vs. HAI 13.2 (12.2-14.2) months vs. RE 10.7 (9.5-12.0). Grade 3-4 toxicity for HAT alone was 40% in the HAI group, 19% in the RE group, and 18% in the TACE groups, which was increased with the addition of systemic chemotherapy. Level 1 evidence was available in 5 studies for HAI, 2 studies for RE and 1 for TACE. HAI, RE, and TACE are equally effective in patients with unresectable CRLM with marginal differences in survival.

To stimulate progress in automating the reconstruction of neural circuits, we organized the first international challenge on 2D segmentation of electron microscopic (EM) images of the brain. Participants submitted boundary maps predicted for a test set of images, and were scored based on their agreement with a consensus of human expert annotations. The winning team had no prior experience with EM images, and employed a convolutional network. This \"deep learning\" approach has since become accepted as a standard for segmentation of EM images. The challenge has continued to accept submissions, and the best so far has resulted from cooperation between two teams. The challenge has probably saturated, as algorithms cannot progress beyond limits set by ambiguities inherent in 2D scoring and the size of the test dataset. Retrospective evaluation of the challenge scoring system reveals that it was not sufficiently robust to variations in the widths of neurite borders. We propose a solution to this problem, which should be useful for a future 3D segmentation challenge.

Malignant peritoneal mesothelioma historically carried a grim prognosis, but outcomes have improved substantially in recent decades. The prognostic significance of clinical, morphologic, and immunophenotypic features remains ill-defined. This multi-institutional cohort comprises 225 malignant peritoneal mesotheliomas, which were assessed for 21 clinical, morphologic, and immunohistochemical parameters. For epithelioid mesotheliomas, combining nuclear pleomorphism and mitotic index yielded a composite nuclear grade, using a previously standardized grading system. Correlation of clinical, morphologic, and immunohistochemical parameters with overall and disease-free survival was examined by univariate and multivariate analyses. On univariate analysis, longer overall survival was significantly associated with diagnosis after 2000 (P = 0.0001), age <60 years (P = 0.0001), ECOG performance status 0 or 1 (P = 0.01), absence of radiographic lymph-node metastasis (P = 0.04), cytoreduction surgery (P < 0.0001), hyperthermic intraperitoneal chemotherapy (P = 0.0001), peritoneal carcinomatosis index <27 (P = 0.01), absence of necrosis (P = 0.007), and epithelioid histotype (P < 0.0001). Among epithelioid malignant mesotheliomas only, longer overall survival was further associated with female sex (P = 0.03), tubulopapillary architecture (P = 0.005), low nuclear pleomorphism (P < 0.0001), low mitotic index (P = 0.0007), and low composite nuclear grade (P < 0.0001). On multivariate analyses, the low composite nuclear grade was independently associated with longer overall and disease-free survival (P < 0.0001). Our data further clarify the interactions of clinical and pathologic features in peritoneal mesothelioma prognosis and validate the prognostic significance of a standardized nuclear-grading system in epithelioid malignant mesothelioma of the peritoneum.

Background: Stroke is a common neurological condition, and post-stroke seizures are known to occur. Early seizures may suggest the severity of insult and may have an effect on the outcome. There are conflicting results on the frequency of early seizures, and studies from India are scarce. Aim: To study the frequency and predictors of early seizures following the first acute stroke, both arterial and venous stroke, as well as to assess their effect on clinical outcome. Patients and Methods: This is a hospital-based, prospective, observational study conducted among 279 eligible consecutive patients admitted in the Neurology department with first acute stroke, including venous stroke. The demographic data, clinical history, risk factors, examination, and all other relevant investigations are done. Early seizures occurring within 7 days of the acute stroke are identified and correlated to various risk factors. Results: Out of the 279 patients enrolled in the study, ischemic stroke (IS) (62.4%) was the most common stroke subtype, followed by hemorrhagic stroke (HS) (20.4%), cerebral sinus venous thrombosis (CSVT) (15.8%), and IS with hemorrhagic transformation (ISH) (1.8%). Thirty-three patients (11.8%) had early seizures, among them CSVT 18 (40.9%) had the highest frequency followed by ISH 1 (20%), HS 5 (8.7%), and IS 9 (5.2%). Conclusions: The frequency of early onset post-stroke seizures is 11.8%, with most of them occurring within 24 hours. Venous stroke, large lesion, cortical location, supratentorial location, hypercoaguable states, and hyperhomocysteinemia are independent predictors. Duration of hospital stay is increased in patients with early seizures, however, they did not influence the in-hospital mortality.

Background: Adoptive cell therapies are emerging as a promising therapeutic option against hematological and solid malignancies. Memory-like natural killer (mlNK) cells are a specific subtype of NK cells generated after cytokine preactivation that have shown enhanced in vivo persistence after infusion into patients, an issue that has hindered traditional NK cell immunotherapy. However, the quality and variability of mlNK cell products remains poorly defined. Methods: In this study, we evaluated heterogeneity across critical functional and molecular aspects of mlNK cells generated from independent donors, including mlNK cytotoxicity, cluster formation, motility, mitochondria morphology, and gene expression. Results: We observed a correlation between changes in gene expression associated with glycolysis and key NK cell functions such as cytotoxicity and motility. For further characterization, we blocked glycolysis and oxidative phosphorylation (OXPHOS) and observed an impaired mlNK functional response, suggesting the importance of metabolism. Conclusions: Our findings provide insights into discriminating between mlNK cell products and how the predictive markers can identify optimal mlNK cell products for adoptive cell therapy of cancer.

Unlike that of its main counterpart, the functional organization of the accessory olfactory bulb, important for detecting socially relevant odors, remains to be detailed. Here the authors map out Ca 2+ signals from vomeronasal inputs to the accessory olfactory bulb in response to socially relevant compounds and find a non-chemotopic spatial organization. The mammalian accessory olfactory system extracts information about species, sex and individual identity from social odors, but its functional organization remains unclear. We imaged presynaptic Ca 2+ signals in vomeronasal inputs to the accessory olfactory bulb (AOB) during peripheral stimulation using light sheet microscopy. Urine- and steroid-responsive glomeruli densely innervated the anterior AOB. Glomerular activity maps for sexually mature female mouse urine overlapped maps for juvenile and/or gonadectomized urine of both sexes, whereas maps for sexually mature male urine were highly distinct. Further spatial analysis revealed a complicated organization involving selective juxtaposition and dispersal of functionally grouped glomerular classes. Glomeruli that were similarly tuned to urines were often closely associated, whereas more disparately tuned glomeruli were selectively dispersed. Maps to a panel of sulfated steroid odorants identified tightly juxtaposed groups that were disparately tuned and dispersed groups that were similarly tuned. These results reveal a modular, nonchemotopic spatial organization in the AOB.