Using deep learning and Google Street View to estimate the demographic makeup of neighborhoods across the United States
- Proceedings of the National Academy of Sciences of the United States of America
- Published over 2 years ago
The United States spends more than $250 million each year on the American Community Survey (ACS), a labor-intensive door-to-door study that measures statistics relating to race, gender, education, occupation, unemployment, and other demographic factors. Although a comprehensive source of data, the lag between demographic changes and their appearance in the ACS can exceed several years. As digital imagery becomes ubiquitous and machine vision techniques improve, automated data analysis may become an increasingly practical supplement to the ACS. Here, we present a method that estimates socioeconomic characteristics of regions spanning 200 US cities by using 50 million images of street scenes gathered with Google Street View cars. Using deep learning-based computer vision techniques, we determined the make, model, and year of all motor vehicles encountered in particular neighborhoods. Data from this census of motor vehicles, which enumerated 22 million automobiles in total (8% of all automobiles in the United States), were used to accurately estimate income, race, education, and voting patterns at the zip code and precinct level. (The average US precinct contains [Formula: see text]1,000 people.) The resulting associations are surprisingly simple and powerful. For instance, if the number of sedans encountered during a drive through a city is higher than the number of pickup trucks, the city is likely to vote for a Democrat during the next presidential election (88% chance); otherwise, it is likely to vote Republican (82%). Our results suggest that automated systems for monitoring demographics may effectively complement labor-intensive approaches, with the potential to measure demographics with fine spatial resolution, in close to real time.
Touch-screen mobile phones/devices (TMPs/Ds) are increasingly used in hospitals. They may act as a mobile reservoir for microbial pathogens. The rates of microbial contamination of TMPs/Ds and keypad mobile phones (KMPs) with respect to different variables including use by healthcare workers (HCWs)/non-HCWs and the demographic characteristics of users were investigated.
In March 2015, Apple Inc announced ResearchKit, a novel open-source framework intended to help medical researchers to easily create apps for medical studies. With the announcement of this framework, Apple presented 5 apps built in a beta phase based on this framework.
The rapid growth in the use of mobile phone applications (apps) provides the opportunity to increase access to evidence-based mental health care.
With continued increases in smartphone ownership, researchers and clinicians are investigating the use of this technology to enhance the management of chronic illnesses such as bipolar disorder (BD). Smartphones can be used to deliver interventions and psychoeducation, supplement treatment, and enhance therapeutic reach in BD, as apps are cost-effective, accessible, anonymous, and convenient. While the evidence-based development of BD apps is in its infancy, there has been an explosion of publicly available apps. However, the opportunity for mHealth to assist in the self-management of BD is only feasible if apps are of appropriate quality.
Current measures of health and disease are often insensitive, episodic, and subjective. Further, these measures generally are not designed to provide meaningful feedback to individuals. The impact of high-resolution activity data collected from mobile phones is only beginning to be explored. Here we present data from mPower, a clinical observational study about Parkinson disease conducted purely through an iPhone app interface. The study interrogated aspects of this movement disorder through surveys and frequent sensor-based recordings from participants with and without Parkinson disease. Benefitting from large enrollment and repeated measurements on many individuals, these data may help establish baseline variability of real-world activity measurement collected via mobile phones, and ultimately may lead to quantification of the ebbs-and-flows of Parkinson symptoms. App source code for these data collection modules are available through an open source license for use in studies of other conditions. We hope that releasing data contributed by engaged research participants will seed a new community of analysts working collaboratively on understanding mobile health data to advance human health.
Mobile phones have become nearly ubiquitous, offering a promising means to deliver health interventions. However, little is known about smartphone applications (apps) for cancer.
BACKGROUND: Smartphone usage has spread to many settings including that of healthcare with numerous potential and realised benefits. The ability to download custom-built software applications (apps) has created a new wealth of clinical resources available to healthcare staff, providing evidence-based decisional tools to reduce medical errors.Previous literature has examined how smartphones can be utilised by both medical student and doctor populations, to enhance educational and workplace activities, with the potential to improve overall patient care. However, this literature has not examined smartphone acceptance and patterns of medical app usage within the student and junior doctor populations. METHODS: An online survey of medical student and foundation level junior doctor cohorts was undertaken within one United Kingdom healthcare region. Participants were asked whether they owned a Smartphone and if they used apps on their Smartphones to support their education and practice activities. Frequency of use and type of app used was also investigated. Open response questions explored participants' views on apps that were desired or recommended and the characteristics of apps that were useful. RESULTS: 257 medical students and 131 junior doctors responded, equating to a response rate of 15.0% and 21.8% respectively. 79.0% (n=203/257) of medical students and 74.8% (n=98/131) of junior doctors owned a smartphone, with 56.6% (n=115/203) of students and 68.4% (n=67/98) of doctors owning an iPhone.The majority of students and doctors owned 1–5 medical related applications, with very few owning more than 10, and iPhone owners significantly more likely to own apps (Chi sq, p<0.001). Both populations showed similar trends of app usage of several times a day. Over 24hours apps were used for between 1--30 minutes for students and 1--20 minutes for doctors, students used disease diagnosis/management and drug reference apps, with doctors favouring clinical score/calculator apps. CONCLUSIONS: This study found a high level of smartphone ownership and usage among medical students and junior doctors. Both groups endorse the development of more apps to support their education and clinical practice.
New (mobile phones, smartphones, tablets, and social media) and traditional media (television) have come to dominate the lives of many children and adolescents. Despite all of this media time and new technology, many parents seem to have few rules regarding the use of media by their children and adolescents.
As the rise of tablets and smartphones move the dominant interface for digital content from mouse or trackpad to direct touchscreen interaction, work is needed to explore the role of interfaces in shaping psychological reactions to online content. This research explores the role of direct-touch interfaces in product search and choice, and isolates the touch element from other form factor changes such as screen size. Results from an experimental study using a travel recommendation Web site show that a direct-touch interface (vs. a more traditional mouse interface) increases the number of alternatives searched, and biases evaluations toward tangible attributes such as décor and furniture over intangible attributes such as WiFi and employee demeanor. Direct-touch interfaces also elevate the importance of internal and subjective satisfaction metrics such as instinct over external and objective metrics such as reviews, which in turn increases anticipated satisfaction metrics. Findings suggest that interfaces can strongly affect how online content is explored, perceived, remembered, and acted on, and further work in interface psychology could be as fruitful as research exploring the content itself.