Capturing children food exposure using wearable cameras and deep learning.

Children's dietary habits are influenced by complex factors within their home, school and neighborhood environments. Identifying such influencers and assessing their effects is traditionally based on self-reported data which can be prone to recall bias. We developed a culturally acceptable machine-learning-based data-collection system to objectively capture school-children's exposure to food (including food items, food advertisements, and food outlets) in two urban Arab centers: Greater Beirut, in Lebanon, and Greater Tunis, in Tunisia. Our machine-learning-based system consists of 1) a wearable camera that captures continuous footage of children's environment during a typical school day, 2) a machine learning model that automatically identifies images related to food from the collected data and discards any other footage, 3) a second machine learning model that classifies food-related images into images that contain actual food items, images that contain food advertisements, and images that contain food outlets, and 4) a third machine learning model that classifies images that contain food items into two classes, corresponding to whether the food items are being consumed by the child wearing the camera or whether they are consumed by others. This manuscript reports on a user-centered design study to assess the acceptability of using wearable cameras to capture food exposure among school children in Greater Beirut and Greater Tunis. We then describe how we trained our first machine learning model to detect food exposure images using data collected from the Web and utilizing the latest trends in deep learning for computer vision. Next, we describe how we trained our other machine learning models to classify food-related images into their respective categories using a combination of public data and data acquired via crowdsourcing. Finally, we describe how the different components of our system were packed together and deployed in a real-world case study and we report on its performance.

Maternal and infant outcomes of Syrian and Palestinian refugees, Lebanese and migrant women giving birth in a tertiary public hospital in Lebanon: a secondary analysis of an obstetric database.

OBJECTIVES: This study aims to assess whether the characteristics, management and outcomes of women varied between Syrian and Palestinian refugees, migrant women of other nationalities and Lebanese women giving birth at a public tertiary centre in Beirut, Lebanon. METHODS: This was a secondary data analysis of routinely collected data from the public Rafik Hariri University Hospital (RHUH) between January 2011 and July 2018. Data were extracted from medical notes using text mining machine learning methods. Nationality was categorised into Lebanese, Syrian, Palestinian and migrant women of other nationalities. The main outcomes were diabetes, pre-eclampsia, placenta accreta spectrum, hysterectomy, uterine rupture, blood transfusion, preterm birth and intrauterine fetal death. Logistic regression models estimated the association between nationality and maternal and infant outcomes, and these were presented using ORs and 95% CIs. RESULTS: 17 624 women gave birth at RHUH of whom 54.3% were Syrian, 39% Lebanese, 2.5% Palestinian and 4.2% migrant women of other nationalities. The majority of women had a caesarean section (73%) and 11% had a serious obstetric complication. Between 2011 and 2018, there was a decline in the use of primary caesarean section (caesarean section performed for the first time) from 7% to 4% of births (p<0.001). The odds of preeclampsia, placenta abruption and serious complications were significantly higher for Palestinian and migrant women of other nationalities compared to Lebanese women, but not for Syrian women. Very preterm birth was higher for Syrians (OR: 1.23, 95% CI: 1.08 to 1.40) and migrant women of other nationalities (OR: 1.51, 95% CI: 1.13 to 2.03) compared to Lebanese women. CONCLUSION: Syrian refugees in Lebanon had similar obstetric outcomes compared to the host population, except for very preterm birth. However, Palestinian women and migrant women of other nationalities appeared to have worse pregnancy complications than the Lebanese women. There should be better healthcare access and support for migrant populations to avoid severe complications of pregnancy.

School and community drivers of child diets in two Arab cities: The SCALE protocol and innovative tools to assess children's food environments.

BACKGROUND: In the context of the rapid nutrition transition experienced by middle-income countries of the Arab region, children and adolescent's food choices and dietary behaviors are early risk factors for the development of non-communicable diseases. Assessment of factors influencing food choices among this age group is challenging and is usually based on self-reported data, which are prone to information and recall bias. As the popularity of technologies and video gaming platforms increases, opportunities arise to use these tools to collect data on variables that affect food choice, dietary intake, and associated outcomes. This protocol paper describes the SCALE study (School and community drivers of child diets in Arab cities; identifying levers for intervention) which aims to explore the environments at the level of households, schools and communities in which children's food choices are made and consequently identify barriers and enablers to healthy food choices within these environments. METHODS: Field studies are being conducted in primary schools, among children aged 9-12 years, in Greater Beirut, Lebanon and Greater Tunis, Tunisia. A stratified random sample of 50 primary schools (public and private) are selected and 50 children are randomly selected from grades 4-5-6 in each school. The study includes surveys with children, parents/caregivers, school directors, teachers, and nutrition/health educators to assess individual diets and the contextual factors that influence children's food choices. Innovative locally adapted tools and methods such as game-based choice experiments, wearable cameras and neighborhood mapping are used to describe the environments in which children's food choices are made. DISCUSSION: The SCALE study will generate contextual knowledge on factors in school and neighborhood environments that influence child dietary behaviors and will inform multi-level interventions and policies to address childhood malnutrition (under-and over-nutrition). By integrating methods from various disciplines, including economics, data science, nutrition, and public health and by considering factors at various levels (home, school, and neighborhood), the study will identify levers for intervention with the potential to improve children's dietary behaviors. This will help fill existing gaps in research on food systems and consequently guide positive change in Lebanon and Tunisia, with the potential for replicability in other contexts.

Meta-learning for fake news detection surrounding the Syrian war.

In this article, we pursue the automatic detection of fake news reporting on the Syrian war using machine learning and meta-learning. The proposed approach is based on a suite of features that include a given article's linguistic style; its level of subjectivity, sensationalism, and sectarianism; the strength of its attribution; and its consistency with other news articles from the same "media camp". To train our models, we use FA-KES, a fake news dataset about the Syrian war. A suite of basic machine learning models is explored, as well as the model-agnostic meta-learning algorithm (MAML) suitable for few-shot learning, using datasets of a modest size. Feature-importance analysis confirms that the collected features specific to the Syrian war are indeed very important predictors for the output label. The meta-learning model achieves the best performance, improving upon the baseline approaches that are trained exclusively on text features in FA-KES.

The impact of digital technology on health of populations affected by humanitarian crises: Recent innovations and current gaps.

Digital technology is increasingly used in humanitarian action and promises to improve the health and social well-being of populations affected by both acute and protracted crises. We set out to (1) review the current landscape of digital technologies used by humanitarian actors and affected populations, (2) examine their impact on health and well-being of affected populations, and (3) consider the opportunities for and challenges faced by users of these technologies. Through a systematic search of academic databases and reports, we identified 50 digital technologies used by humanitarian actors, and/or populations affected by crises. We organized them according to the stage of the humanitarian cycle that they were used in, and the health outcomes or determinants of health they affected. Digital technologies were found to facilitate communication, coordination, and collection and analysis of data, enabling timely responses in humanitarian contexts. A lack of evaluation of these technologies, a paternalistic approach to their development, and issues of privacy and equity constituted major challenges. We highlight the need to create a space for dialogue between technology designers and populations affected by humanitarian crises.