Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

General General

Comparison of Machine Learning Methods in the Study of Cancer Survivors' Return to Work: An Example of Breast Cancer Survivors with Work-Related Factors in the CONSTANCES Cohort.

In Journal of occupational rehabilitation

PURPOSE : Machine learning (ML) methods showed a higher accuracy in identifying individuals without cancer who were unable to return to work (RTW) compared to the classical methods (e.g. logistic regression models). We therefore aim to discuss the value of these methods in relation to RTW for cancer survivors.

METHODS : Breast cancer (BC) survivors who were working at diagnosis within the CONSTANCES cohort were included in the study. RTW was assessed five years after the BC diagnosis (early retirement was considered as non-RTW). Age and occupation at diagnosis, and physical occupational job exposures assessed using the Job Exposure Matrix, JEM-CONSTANCES, were evaluated as predictors of RTW five years after BC diagnosis. The following four ML methods were used: (i) k-nearest neighbors; (ii) random forest; (iii) neural network; and (iv) elastic net.

RESULTS : The training sample included 683 BC survivors (RTW: 85.7%), and the test sample 171 (RTW: 85.4%). The elastic net method had the best results despite low sensitivity (accuracy = 76.6%; sensitivity = 31.7%; specificity = 90.8%), and the random forest model was the most accurate (= 79.5%) but also the least sensitive (= 14.3%).

CONCLUSION : This study takes a first step towards opening up new possibilities for identifying the occupational determinants of cancer survivors' RTW. Further work, including a larger sample size, and more predictor variables, is now needed.

Badreau Marie, Fadel Marc, Roquelaure Yves, Bertin Mélanie, Rapicault Clémence, Gilbert Fabien, Porro Bertrand, Descatha Alexis

2023-Mar-20

Breast cancer, Machine learning, Methods, Prediction, Return to work, Survivors

Public Health Public Health

Mechanisms influencing the factors of urban built environments and coronavirus disease 2019 at macroscopic and microscopic scales: The role of cities.

In Frontiers in public health

In late 2019, the coronavirus disease 2019 (COVID-19) pandemic soundlessly slinked in and swept the world, exerting a tremendous impact on lifestyles. This study investigated changes in the infection rates of COVID-19 and the urban built environment in 45 areas in Manhattan, New York, and the relationship between the factors of the urban built environment and COVID-19. COVID-19 was used as the outcome variable, which represents the situation under normal conditions vs. non-pharmacological intervention (NPI), to analyze the macroscopic (macro) and microscopic (micro) factors of the urban built environment. Computer vision was introduced to quantify the material space of urban places from street-level panoramic images of the urban streetscape. The study then extracted the microscopic factors of the urban built environment. The micro factors were composed of two parts. The first was the urban level, which was composed of urban buildings, Panoramic View Green View Index, roads, the sky, and buildings (walls). The second was the streets' green structure, which consisted of macrophanerophyte, bush, and grass. The macro factors comprised population density, traffic, and points of interest. This study analyzed correlations from multiple levels using linear regression models. It also effectively explored the relationship between the urban built environment and COVID-19 transmission and the mechanism of its influence from multiple perspectives.

Zhang Longhao, Han Xin, Wu Jun, Wang Lei

2023

COVID-19, computer vision, deep learning, relevance, street view images, urban built environment

Radiology Radiology

LIMITR: Leveraging Local Information for Medical Image-Text Representation

ArXiv Preprint

Medical imaging analysis plays a critical role in the diagnosis and treatment of various medical conditions. This paper focuses on chest X-ray images and their corresponding radiological reports. It presents a new model that learns a joint X-ray image & report representation. The model is based on a novel alignment scheme between the visual data and the text, which takes into account both local and global information. Furthermore, the model integrates domain-specific information of two types -- lateral images and the consistent visual structure of chest images. Our representation is shown to benefit three types of retrieval tasks: text-image retrieval, class-based retrieval, and phrase-grounding.

Gefen Dawidowicz, Elad Hirsch, Ayellet Tal

2023-03-21

Public Health Public Health

A hybrid deep forest-based method for predicting synergistic drug combinations.

In Cell reports methods

Combination therapy is a promising approach in treating multiple complex diseases. However, the large search space of available drug combinations exacerbates challenge for experimental screening. To predict synergistic drug combinations in different cancer cell lines, we propose an improved deep forest-based method, ForSyn, and design two forest types embedded in ForSyn. ForSyn handles imbalanced and high-dimensional data in medium-/small-scale datasets, which are inherent characteristics of drug combination datasets. Compared with 12 state-of-the-art methods, ForSyn ranks first on four metrics for eight datasets with different feature combinations. We conduct a systematic analysis to identify the most appropriate configuration parameters. We validate the predictive value of ForSyn with cell-based experiments on several previously unexplored drug combinations. Finally, a systematic analysis of feature importance is performed on the top contributing features extracted by ForSyn. The resulting key genes may play key roles on corresponding cancers.

Wu Lianlian, Gao Jie, Zhang Yixin, Sui Binsheng, Wen Yuqi, Wu Qingqiang, Liu Kunhong, He Song, Bo Xiaochen

2023-Feb-27

class imbalance, deep forest, deep learning, drug combination prediction, synergy

General General

BOMA, a machine-learning framework for comparative gene expression analysis across brains and organoids.

In Cell reports methods

Our machine-learning framework, brain and organoid manifold alignment (BOMA), first performs a global alignment of developmental gene expression data between brains and organoids. It then applies manifold learning to locally refine the alignment, revealing conserved and specific developmental trajectories across brains and organoids. Using BOMA, we found that human cortical organoids better align with certain brain cortical regions than with other non-cortical regions, implying organoid-preserved developmental gene expression programs specific to brain regions. Additionally, our alignment of non-human primate and human brains reveals highly conserved gene expression around birth. Also, we integrated and analyzed developmental single-cell RNA sequencing (scRNA-seq) data of human brains and organoids, showing conserved and specific cell trajectories and clusters. Further identification of expressed genes of such clusters and enrichment analyses reveal brain- or organoid-specific developmental functions and pathways. Finally, we experimentally validated important specific expressed genes through the use of immunofluorescence. BOMA is open-source available as a web tool for community use.

He Chenfeng, Kalafut Noah Cohen, Sandoval Soraya O, Risgaard Ryan, Sirois Carissa L, Yang Chen, Khullar Saniya, Suzuki Marin, Huang Xiang, Chang Qiang, Zhao Xinyu, Sousa Andre M M, Wang Daifeng

2023-Feb-27

cell-type functional genomics, comparative analysis of brains and organoids, developmental gene expression patterns, manifold alignment tool and webapp

General General

Exploring obesity, physical activity, and digital game addiction levels among adolescents: A study on machine learning-based prediction of digital game addiction.

In Frontiers in psychology ; h5-index 92.0

Primary study aim was defining prevalence of obesity, physical activity levels, digital game addiction level in adolescents, to investigate gender differences, relationships between outcomes. Second aim was predicting game addiction based on anthropometric measurements, physical activity levels. Cross-sectional study design was implemented. Participants aged 9-14 living in Kirikkale were part of the study. The sample of the study consists of 405 adolescents, 231 girls (57%) and 174 boys (43%). Self-reported data were collected by questionnaire method from a random sample of 405 adolescent participants. To determine the physical activity levels of children, the Physical Activity Questionnaire for Older Children (PAQ-C). Digital Game addiction was evaluated with the digital game addiction (DGA) scale. Additionally, body mass index (BMI) status was calculated by measuring the height and body mass of the participants. Data analysis were performed using Python 3.9 software and SPSS 28.0 (IBM Corp., Armonk, NY, United States) package program. According to our findings, it was determined that digital game addiction has a negative relationship with physical activity level. It was determined that physical activity level had a negative relationship with BMI. In addition, increased physical activity level was found to reduce obesity and DGA. Game addiction levels of girl participants were significantly higher than boy participants, and game addiction was higher in those with obesity. With the prediction model obtained, it was determined that age, being girls, BMI and total physical activity (TPA) scores were predictors of game addiction. The results revealed that the increase in age and BMI increased the risk of DGA, and we found that women had a 2.59 times greater risk of DGA compared to men. More importantly, the findings of this study showed that physical activity was an important factor reducing DGA 1.51-fold. Our prediction model Logit (P) = 1/(1 + exp(-(-3.384 + Age*0.124 + Gender-boys*(-0.953) + BMI*0.145 + TPA*(-0.410)))). Regular physical activity should be encouraged, digital gaming hours can be limited to maintain ideal weight. Furthermore, adolescents should be encouraged to engage in physical activity to reduce digital game addiction level. As a contribution to the field, the findings of this study presented important results that may help in the prevention of adolescent game addiction.

Gülü Mehmet, Yagin Fatma Hilal, Gocer Ishak, Yapici Hakan, Ayyildiz Erdem, Clemente Filipe Manuel, Ardigò Luca Paolo, Zadeh Ali Khosravi, Prieto-González Pablo, Nobari Hadi

2023

addiction, body mass index, children, obesity, physical inactivity, sedentary behaviors