arXiv subjects

Sandip Patel

Publications and source records attributed to Sandip Patel.

4 recordsLinked to original sources

arXiv2023

Exploring The Contribution of Innate Immune Cells to Breast Cancer Immunotherapy

Breast cancer is the leading type of cancer in women. About 10-15% of breast cancers are triple-negative breast cancer (TNBC), a subtype with the worst prognosis. Due to the lack of estrogen, progesterone and HER2 receptor expression, chemotherapies have been the standard of care for decades. Immunotherapy has emerged as promising for TNBC treatment. In 2020, the Food and Drug Administration (FDA) granted approval to pembrolizumab in combination with chemotherapy for patients with advanced triple-negative breast cancer. However, only a subgroup of advanced TNBC patients live longer whose tumors have a PD-L1 Combined Positive Score of at least 10 (CPS>=10). There is still an unmet medical need to provide alternative treatment for the rest of patients. Interestingly, a few of patients at UCSD Moores Cancer Center were found to have had excellent responses to pembrolizumab despite low CPS scores (termed Elite Responders). The hypothesis of this project is that there may be an alternative immune response mechanism and/or crosstalk happening between the innate and adaptive immune systems, especially in Natural Killer Cells and Macrophages, that contributed to this unexpected excellent response. Our procedure used ACDBio RNAscope Multiplex Fluorescence v2 method to spatially analyze innate immune cells (Natural Killer cells and macrophages) and adaptive immune cells (T-cells) in the Tumor Micro Environment. Our data demonstrated increased tumor infiltration of innate immune cells (macrophage and Natural Killer cells) in the Elite Responders. This conclusion indicated the joint effort of two immune systems (innate and adaptive) which eventually led to increased survival.

q-bio.CB↗

arXiv2023

Decision Making for Human-in-the-loop Robotic Agents via Uncertainty-Aware Reinforcement Learning

In a Human-in-the-Loop paradigm, a robotic agent is able to act mostly autonomously in solving a task, but can request help from an external expert when needed. However, knowing when to request such assistance is critical: too few requests can lead to the robot making mistakes, but too many requests can overload the expert. In this paper, we present a Reinforcement Learning based approach to this problem, where a semi-autonomous agent asks for external assistance when it has low confidence in the eventual success of the task. The confidence level is computed by estimating the variance of the return from the current state. We show that this estimate can be iteratively improved during training using a Bellman-like recursion. On discrete navigation problems with both fully- and partially-observable state information, we show that our method makes effective use of a limited budget of expert calls at run-time, despite having no access to the expert at training time.

cs.RO↗

arXiv2022

Magnetic field measurements on the mini-ICAL detector using Hall probes

The magnetised 51 kton Iron Calorimeter (ICAL) detector proposed to be built at INO is designed with a focus on detecting 1-20 GeV muons. The magnetic field will enable the measurement of the momentum of the $\mu^-$ and $\mu^+$ generated from the charge current interactions of $\nu_\mu$ and $\bar\nu_\mu$ separately within iron in the detector, thus permitting the determination of the neutrino mass ordering/hierarchy, among other important goals of ICAL. Hence it is important to determine the magnetic field as accurately as possible. The mini-ICAL detector is an 85-ton prototype of ICAL, which is operational at Madurai in South India. We describe here the first measurement of the magnetic field in mini-ICAL using Hall sensor PCBs. A set-up developed to calibrate the Hall probe sensors using an electromagnet. The readout system has been designed using an Arduino Nano board for selection of channels of Hall probes mounted on the PCB and to convert the analog voltage to a digital output. The magnetic field has been measured in the small gaps (provided for the purpose) between iron plates in the top layer of mini-ICAL as well as in the air just outside the detector. A precision of better than 3% was obtained, with a sensitivity down to about 0.03 kGauss when measuring the small fringe fields outside the detector.

physics.ins-det↗

arXiv2019

ICU Delirium Prediction Models: A Systematic Review

Purpose: Summarize ICU delirium prediction models published within the past five years. Methods: Electronic searches were conducted in April 2019 using PubMed, Embase, Cochrane Central, Web of Science, and CINAHL to identify peer reviewed studies published in English during the past five years that specifically addressed the development, validation, or recalibration of delirium prediction models in adult ICU populations. Data were extracted using CHARMS checklist elements for systematic reviews of prediction studies, including the following characteristics: study design, participant descriptions and recruitment methods, predicted outcomes, a priori candidate predictors, sample size, model development, model performance, study results, interpretation of those results, and whether the study included missing data. Results: Twenty studies featuring 26 distinct prediction models were included. Model performance varied greatly, as assessed by AUROC (0.68-0.94), specificity (56.5%-92.5%), and sensitivity (59%-90.9%). Most models used data collected from a single time point or window to predict the occurrence of delirium at any point during hospital or ICU admission, and lacked mechanisms for providing pragmatic, actionable predictions to clinicians. Conclusions: Although most ICU delirium prediction models have relatively good performance, they have limited applicability to clinical practice. Most models were static, making predictions based on data collected at a single time-point, failing to account for fluctuating conditions during ICU admission. Further research is needed to create clinically relevant dynamic delirium prediction models that can adapt to changes in individual patient physiology over time and deliver actionable predictions to clinicians.

stat.AP↗