scholarly journals Machine Learning–Based Screening of Healthy Meals From Image Analysis: System Development and Pilot Study (Preprint)

2020 ◽  
Author(s):  
Kyoko Sudo ◽  
Kazuhiko Murasaki ◽  
Tetsuya Kinebuchi ◽  
Shigeko Kimura ◽  
Kayo Waki
Keyword(s):  
Health Care ◽  
System Development ◽  
Background Information ◽  
Dietary Control ◽  
Image Analysis System ◽  
Registered Dietitian ◽  
Care Workers ◽  
Patients With Diabetes ◽  
Analysis System ◽  
The Relationship

BACKGROUND Recent research has led to the development of many information technology–supported systems for health care control, including systems estimating nutrition from images of meals. Systems that capture data about eating and exercise are useful for people with diabetes as well as for people who are simply on a diet. Continuous monitoring is key to effective dietary control, requiring systems that are simple to use and motivate users to pay attention to their meals. Unfortunately, most current systems are complex or fail to motivate. Such systems require some manual inputs such as selection of an icon or image, or by inputting the category of the user’s food. The nutrition information fed back to users is not especially helpful, as only the estimated detailed nutritional values contained in the meal are typically provided. OBJECTIVE In this paper, we introduce healthiness of meals as a more useful and meaningful general standard, and present a novel algorithm that can estimate healthiness from meal images without requiring manual inputs. METHODS We propose a system that estimates meal healthiness using a deep neural network that extracts features and a ranking network that learns the relationship between the degrees of healthiness of a meal using a dataset prepared by a human dietary expert. First, we examined whether a registered dietitian can judge the healthiness of meals solely by viewing meal images using a small dataset (100 meals). We then generated ranking data based on comparisons of sets of meal images (850 meals) by a registered dietitian’s viewing meal images and trained a ranking network. Finally, we estimated each meal’s healthiness score to detect unhealthy meals. RESULTS The ranking estimated by the proposed network and the ranking of healthiness based on the dietitian’s judgment were correlated (correlation coefficient 0.72). In addition, extracting network features through pretraining with a publicly available large meal dataset enabled overcoming the limited availability of specific healthiness data. CONCLUSIONS We have presented an image-based system that can rank meals in terms of the overall healthiness of the dishes constituting the meal. The ranking obtained by the proposed method showed a good correlation to nutritional value–based ranking by a dietitian. We then proposed a network that allows conditions that are important for judging the meal image, extracting features that eliminate background information and are independent of location. Under these conditions, the experimental results showed that our network achieves higher accuracy of healthiness ranking estimation than the conventional image ranking method. The results of this experiment in detecting unhealthy meals suggest that our system can be used to assist health care workers in establishing meal plans for patients with diabetes who need advice in choosing healthy meals.

scholarly journals Machine Learning–Based Screening of Healthy Meals From Image Analysis: System Development and Pilot Study

10.2196/18507 ◽  
2020 ◽  
Vol 4 (10) ◽  
pp. e18507
Author(s):  
Kyoko Sudo ◽  
Kazuhiko Murasaki ◽  
Tetsuya Kinebuchi ◽  
Shigeko Kimura ◽  
Kayo Waki
Keyword(s):  
Health Care ◽  
System Development ◽  
Background Information ◽  
Dietary Control ◽  
Image Analysis System ◽  
Registered Dietitian ◽  
Care Workers ◽  
Patients With Diabetes ◽  
Analysis System ◽  
The Relationship

Background Recent research has led to the development of many information technology–supported systems for health care control, including systems estimating nutrition from images of meals. Systems that capture data about eating and exercise are useful for people with diabetes as well as for people who are simply on a diet. Continuous monitoring is key to effective dietary control, requiring systems that are simple to use and motivate users to pay attention to their meals. Unfortunately, most current systems are complex or fail to motivate. Such systems require some manual inputs such as selection of an icon or image, or by inputting the category of the user’s food. The nutrition information fed back to users is not especially helpful, as only the estimated detailed nutritional values contained in the meal are typically provided. Objective In this paper, we introduce healthiness of meals as a more useful and meaningful general standard, and present a novel algorithm that can estimate healthiness from meal images without requiring manual inputs. Methods We propose a system that estimates meal healthiness using a deep neural network that extracts features and a ranking network that learns the relationship between the degrees of healthiness of a meal using a dataset prepared by a human dietary expert. First, we examined whether a registered dietitian can judge the healthiness of meals solely by viewing meal images using a small dataset (100 meals). We then generated ranking data based on comparisons of sets of meal images (850 meals) by a registered dietitian’s viewing meal images and trained a ranking network. Finally, we estimated each meal’s healthiness score to detect unhealthy meals. Results The ranking estimated by the proposed network and the ranking of healthiness based on the dietitian’s judgment were correlated (correlation coefficient 0.72). In addition, extracting network features through pretraining with a publicly available large meal dataset enabled overcoming the limited availability of specific healthiness data. Conclusions We have presented an image-based system that can rank meals in terms of the overall healthiness of the dishes constituting the meal. The ranking obtained by the proposed method showed a good correlation to nutritional value–based ranking by a dietitian. We then proposed a network that allows conditions that are important for judging the meal image, extracting features that eliminate background information and are independent of location. Under these conditions, the experimental results showed that our network achieves higher accuracy of healthiness ranking estimation than the conventional image ranking method. The results of this experiment in detecting unhealthy meals suggest that our system can be used to assist health care workers in establishing meal plans for patients with diabetes who need advice in choosing healthy meals.

At&T Image Analysis System Development for Manual Lifts

1993 ◽  
Vol 37 (16) ◽  
pp. 1056-1059
Author(s):  
Bahador Ghahramani ◽  
Lawrence B. Fleischer ◽  
Jerome J. Congleton ◽  
Joseph W. Foster
Keyword(s):  
Material Handling ◽  
System Development ◽  
Background Information ◽  
Video Tape ◽  
Image Analysis System ◽  
Manual Material Handling ◽  
System User ◽  
Analysis System ◽  
User Friendly ◽  
Niosh Lifting Equation

AT&T Video Ergonomics Evaluator System (AT&T-VEE) is a state-of-the-art computer analysis system used to analyze the hazards of a Manual Material Handling (MMH) task. This project was funded and managed by the American Telephone & Telegraph Company (AT & T) and designed and developed at Texas A&M University. The system is a lifting analyses tool for the 1992 NIOSH Lifting Equation to determine minimal user effort in lifting an object and to standardize the outputs. This system is now capable of providing the AT & T safety specialists with clear access to: complete NIOSH analyses, pictures of the lift environment, and background information on the MMH process. It was determined that a system, such as AT & T-VEE, was needed to reduce AT & T's MMH injuries and ensuing problems. The scope of the project was to develop a menu driven, user friendly system that would provide employees with the ability to determine the maximum weight of objects utilized at MMH tasks. In order to perform a lifting analysis, a video tape of a manual lift is first produced. The system user, in conjunction with a computer, enters employee's measurement landmarks and points of analyses, e.g., ankles, load center, weight, etc. AT & T-VEE produces a NIOSH Lifting Equation analyses, pictures and records of the MMH task.

scholarly journals Mind your words. Language and war metaphors in the COVID - 19 pandemic

2020 ◽  
Vol 9 (2) ◽  
pp. 58-73
Author(s):  
Francesca Brencio
Keyword(s):  
Health Care ◽  
Health Care Workers ◽  
Direct Contact ◽  
War Narratives ◽  
Care Workers ◽  
Philosophical Account ◽  
The World ◽  
The Relationship ◽  
The Way

Through this contribution I aim to show how the role of language and metaphors is fundamental to our understanding of reality, affecting the way we ordinarily act and live, and particularly important in facing fears and anguish. This is more evident in these times of the COVID-19 pandemic, where our experiences of language and of the world seem to be characterised mainly by war terminology. Politicians declare themselves at war fighting an invisible enemy and health care workers, who are in direct contact with COVID-19 positive patients, are said to be “fighting” on the “frontlines”. Starting from a philosophical account of the relationship between language, fear and anguish, I aim to show how this narrative is unhelpful, both for society at large and especially for patients and health care workers. While war narratives instil fear, it seems to me that new forms of solidarity and new models of coexistence are required. Since language shapes the way in which we think, live and act, it is important to choose words that encourage people to act responsibly, to cooperate and to overcome the hardships of the COVID19 pandemic together.

scholarly journals The Characteristics of Health Care Workers with COVID-19 and Relationship Between COVID-19 Mortality and BCG/Tuberculosis History: a multi-center study

2020 ◽  
Author(s):  
Ş Torun ◽  
Ş Özkaya ◽  
N Şen ◽  
F Kanat ◽  
I Karaman ◽  
...  
Keyword(s):  
Health Care ◽  
Mortality Rate ◽  
Health Care Workers ◽  
Healthcare Workers ◽  
Normal Population ◽  
Bcg Vaccine ◽  
Direct Care ◽  
Care Workers ◽  
History Of ◽  
The Relationship

AbstractBackgroundToday, COVID-19 pandemic has brought countries’ health services into sharp focus. Despite the low incidence of cases(1.2%) and high mortality rate(2.4%) among Turkish population, the low mortality rate(0.3%) despite the high incidence(11.5%) declared in healthcare workers drew our group’s attention. Therefore, we aimed to report the characteristics of infected health-care workers and investigate the relationship between BCG vaccine and tuberculosis history with COVID-19 mortality in infected health-care worker population.MethodThis study was conducted in three hospitals to assess the clinical presentations, disease severity and correlation with BCG vaccine and tuberculous history in COVID-19 positive health-care workers by an online questionnaire platform. The relationship between characteristics and tuberculosis history were investigated according to hospitalization status of the patients.ResultTotal of 465 infected healthcare workers included in the study. The rate of history of direct care and contact to tuberculosis patient, presence of previous tuberculosis treatment and BCG scar, presence of radiological infiltrations was significantly higher in hospitalized healthcare workers. The ratio of direct care and direct contact to the patient with tuberculosis, and presence of family history of tuberculosis were statistically significantly higher in patients with radiological infiltrations.ConclusionAlthough COVID-19 risk and incidence are higher among healthcare workers compared to the normal population due to higher virus load, this study brings evidence for the fact that the lower mortality rate seen in infected healthcare workers might be due to healthcare workers’ frequent exposure to tuberculosis bacillus and the mortality-reducing effects of BCG vaccine, despite the higher hospitalization rate and radiological infiltrations due to over-triggered immune system.

scholarly journals Potential Role of Personal Protective Equipment Use in the Protection Against COVID-19 Infection Among Health Care Workers

2020 ◽  
Author(s):  
Wei Wang ◽  
Yuan-Zeng Min ◽  
Chun-Mei Yang ◽  
Hai-Ou Hong ◽  
Tian Xue ◽  
...  
Keyword(s):  
Health Care ◽  
Health Care Workers ◽  
Personal Protective Equipment ◽  
Hubei Province ◽  
Protective Equipment ◽  
Daily Production ◽  
Daily Number ◽  
Care Workers ◽  
The Relationship

AbstractThe coronavirus disease 2019 (COVID-19) pandemic has posed a major challenge for protecting health care workers (HCWs) against the infection. Use of personal protective equipment (PPE) in health care workplace is recommended as a high priority. In order to investigate the relationship between PPE use and the number of COVID-19 cases among HCWs, we conducted a molecular epidemiological study among 142 HCWs who were dispatched from Hefei to work in Wuhan and 284 HCWs who remained in Hefei, China; both provided care for patients with COVID-19. Nucleic acid testing and SARS-CoV-2 specific antibody (IgM, IgG, IgA) detection were performed to confirm SARS-CoV-2 infection among those HCWs. We also extracted publicly released data on daily number of COVID-19 cases among HCWs, daily number of HCWs who were dispatched to Hubei province since January 24, and daily production of PPE in China and daily demand and supply of PPE in Hubei province. Our laboratory testing confirmed that none of the 142 HCWs who were dispatched to work in Wuhan and 284 HCWs who remained in Hefei were infected by SARS-CoV-2. Consistent with these findings, as of April 15, 2020, none of the 42,600 HCWs who were successively dispatched to Hubei province since January 24, 2020 was reported to have COVID-19. These HCWs were provided with adequate supply of PPE as committed by their original institutions or provinces. In contrast, during the early phase of COVID-19 epidemic in Hubei province, a substantial shortage of PPE and an increasing number of COVID-19 infection among HCWs were reported. With the continuing increase in domestic production of PPE in China, the PPE supply started to meet and then exceed the demand. This coincided with a subsequent reduction in the number of reported COVID-19 cases among HCWs. In conclusion, our findings indicate that COVID-19 infection among HCWs could be completely prevented. Appropriate and adequate PPE might play a crucial role in protecting HCWs against COVID-19 infection.

Sign in / Sign up

Export Citation Format

Share Document