China
Africa
Explore chapters and articles related to this topic
Human biomedical research, medical innovations and information technologies in healthcare †
Published in Gary Chan Kok Yew, Health Law and Medical Ethics in Singapore, 2020
Electronic data may be aggregated for the purpose of detecting trends and making predictions. Big data allows for predictive analysis of trends relating to diseases and symptoms globally or in certain regions or countries. During the COVID-19 pandemic, big data was used to predict and to reduce the spread of the coronavirus. For example, a heat map was developed to detect the users who were experiencing symptoms of COVID-19 and ascertain if they were staying home in order to contain the spread.121 An app known as TraceTogether was developed in Singapore to trace infected persons by allowing users who have downloaded the app on their smart phones to detect other users who may be within close proximity during a specified period. Big data can generate information for clinical practice for the benefit of patients as well as in medical research to enhance the store of generalisable knowledge on health and well-being. However, obtaining consent from patients to use their personal data for specific purposes would be impractical where the aggregated data may be used in a wide range of contexts and in ways that are quite removed from the original source of information.
Statistical Graphics
Published in Albert Vexler, Alan D. Hutson, Xiwei Chen, Statistical Testing Strategies in the Health Sciences, 2017
Albert Vexler, Alan D. Hutson, Xiwei Chen
The heat map is a graphical technique to represent data where colors and darkness are used to represent the individual values contained in a data matrix. The heat maps are often used to look for similarities between variables, for example, genes, and between samples. Heat maps can also be made from dissimilarity matrices, which are particularly useful when clustering patterns might not be easily visible in the data matrix, as with absolute correlation distance (van der Laan and Pollard 2003). Various color-coding schemes can be used to illustrate the heat map, with perceptual advantages and disadvantages for each. For example, rainbow color maps are often used, as humans can perceive more shades of color than they can of gray, and this would purportedly increase the amount of detail perceivable in the image. However, in many cases, rainbow color maps may make actual gradients less prominent and sometimes actually obscure detail rather than enhancing it, while grayscale or blackbody spectrum color maps can keep the natural perceptual ordering. In the microarray data analysis, the red–green color map is the most commonly used, ranging from pure green at the low end, through black in the middle, to pure red at the high end.
A type I interferon footprint in pre-operative biopsies is an independent biomarker that in combination with CD8+ T cell quantification can improve the prediction of response to neoadjuvant treatment of rectal adenocarcinoma
Published in OncoImmunology, 2023
Azar Rezapour, Daniel Rydbeck, Fabian Byvald, Viktor Tasselius, Gustaf Danielsson, Eva Angenete, Ulf Yrlid
In this study, we explored the correlations between infiltration of T cell subsets, type I IFN response, and tumor regression following neoadjuvant treatment. We stratified the patients according to the density of CD8+ and MxA+ cells in the entire tumor tissue and tumor stroma, respectively, with a heat-map approach. We also added clinical data previously suggested to predict tumor response, and still our model provided better correlation with treatment response. The lack of a validation of the heat-map approach in an independent cohort is a limitation of the study. Ongoing studies will hopefully validate these data in the future. Following a successful verification, this novel approach could aid in identifying patients with a good chance of achieving a pCR. The stratification could potentially also be used to identify patients who would benefit from intratumoral immune cell reinvigoration preceding, or in combination with, neoadjuvant treatment to replace surgery.
Using deep learning and explainable artificial intelligence to assess the severity of gastroesophageal reflux disease according to the Los Angeles Classification System
Published in Scandinavian Journal of Gastroenterology, 2023
Zhenyang Ge, Bowen Wang, Jiuyang Chang, Zequn Yu, Zhenyuan Zhou, Jing Zhang, Zhijun Duan
There are mainly three XAI paradigms, including post-hoc, intrinsic, and distillation, of which the post-hoc paradigm has been extensively studied. This method typically provides a heat-map highlighting important regions for making decision. The heat-map is computed, as well as the forward path of the model. The most popular post-hoc methods are based on channel activation or back-propagation, including class activation mapping (CAM) [31] and gradient-weighted CAM (Grad-CAM) [32]. Another type of post-hoc method is perturbation-based, including Occlusion [33] and RISE [34]. They randomly mask some pixel regions of the input image and calculate the importance of the masked part for the prediction by the corresponding output. Then, a merged heat-map is generated by running this operation for several times. Different from the post-hoc method that provides explanations after the model training, the intrinsic paradigm [35,36] explores the important piece of information within the forward path of the model. This allows the explanations to be generated in a more model-nature manner.
Bystander CD4+ T cells infiltrate human tumors and are phenotypically distinct
Published in OncoImmunology, 2022
Shamin Li, Summer Zhuang, Antja Heit, Si-Lin Koo, Aaron C. Tan, I-Ting Chow, William W. Kwok, Iain Beehuat Tan, Daniel S.W. Tan, Yannick Simoni, Evan W. Newell
After mass cytometry (CyTOF) acquisition, any zero values were randomized using a uniform distribution of values between 0 and −1 using R. The signal of each parameter was normalized based on EQ beads (Fluidigm) as described previously.69 Samples were then used for UMAP analysis similar to that previously described using customized R scripts based on the ‘flowCore’ and ‘uwot’ R packages.37 In R, all data were transformed using the logicleTransform function (flowCore package) using parameters: w = 0.25, t = 16409, m = 4.5, a = 0 to roughly match scaling historically used in FlowJo. For heatmaps, median intensity corresponds to a logical data scale using formula previously described.70 The colors in the heat map represent the measured means intensity value of a given marker in a given sample. A seven-color scale is used with black–blue indicating low expression values, green–yellow indicating intermediately expressed markers, and orange-red representing highly expressed markers.