1
COMPUTERIZED SELF-ASSESSMENT DIETARY SYSTEM Soubhik Biswas1,Dr. Swapan Kumar Biswas2
Abstract — Computerized dietary assessment programs are often used for nutrition education research and practice. This article provides an overview of the dietary assessment program mentioned here, along with the components and capabilities of these programs derived from additional sources as needed. According to the literature, the advantages of using computers for dietary assessment include standardization of the questioning sequence, fast and easy processing, immediate results, and increased flexibility .The disadvantages include the need for typing skills and computer literacy, as well as potential bias in the responses if an interviewer is required. Index Terms — diet, nutrition assessment, computer, selfassessment, software system
I. MOTIVATION In this evolving cosmopolitan urban time there has been a recent trend in maintaining fitness. It has been observed that often many people start planning their dietary and exercise profile unscientifically, which ultimately it leads to unhealthy physiological conditions. The basic purpose of this literature is the development of software tool to help in creating nutritional awareness among the common people. In the global campaign of Health for All, promotion of proper nutrition is one of the eight elements of primary health care. Greater emphasis is now placed on integrating nutrition into primary health care systems whenever possible, and formulation of national dietary goals to promote health and nutritional status of families and communities. II. INTRODUCTION This paper addresses the use of computer technology for dietary assessment. Computer-assisted self-assessment is in which a respondent uses the computer to complete their own assessment. There are a number of cognitive advantages to using computers for dietary assessment, rather than noncomputerized approaches. These include enhanced communication through pictures, standardisation of the questioning sequence , decreased bias toward socially undesirable questions and the ability to collect data in a neutral environment (in the case of self-assessment). 1 Soubhik Biswas, is currently doing Msc in Computer Science from Department of Computer Science, Pondicherry University, Kalapet, Puducherry – 605014,India. (e-mail:
[email protected]). 2 Dr. Swapan Kumar Biswas(MBBS - Cal) is an ex-house physician of NRS Medical College, Sealdah, Kolkata – 700014, West Bengal, India. (e-mail:
[email protected]).
Disadvantages include the need for typing skills and computer literacy as well as potential bias in the responses if an interviewer is required. Speech recognition and touch screen technology may enhance computerised assessment as they are incorporated into newer programs. Computerised assessment can maximize effectiveness of dietary advice because it provides automatic feedback, tailored to the individual. Feedback may be personalised by drawing relevant recommendations from an ordered message archive. Automated feedback of this nature has been shown to be more effective than generalized feedback in diabetes management. Tailored feedback can have both motivating and reinforcing effects. Feedback may take the form of graphs or tables representing the adequacy of a person’s nutritional intake, health risks associated with low or high intakes identified by the assessment, and related nutrition recommendations and recipes. Some even generate related shopping lists. Tailored feedback should provide recommendations based on an individual’s usual eating habits, food preferences and stage of change. In most cases the computerised dietary assessment programs don’t have an advice component, and it simply reports results of the dietary assessment which the dietician or nutritionist uses to formulate advice or to assess change in dietary intake. Though it may contain some form of ‘memory enhancement’ features to help respondents remember all details of their usual diet. Food descriptions, probe questions and prompts in text format are employed in assessment programs to prompt the respondent’s memory. Meal-based questions have been shown to result in more accurate reporting than questions regarding individual foods. With newer computerized programs, clients may select a food from photographs integrated into the program and drag the image of that food to a plate representing the foods as they are eaten together. Approaches such as these enable respondents to focus on the timing, setting and task to be remembered. Manual methods of assessment limit the accuracy of reporting actual consumption unless direct visual representations of the food and plate waste can be conducted. Portion sizes visualised through realistic images can aid recall of dietary habits (generic memory) and casual encounters with food (episodic memory). Computerized assessment programs are often judged on ease of data entry; ability to preview single nutrients while
2 entering food names; optional expression of food portion by weight, volume, or household measure; whether food lists can be edited, and ability to compare results to a variety of dietary standards. The ease of averaging multiple days of intake and exporting data for statistical analyses may be important as well.
literacy, incomplete questionnaires, and disinterest resulting in low participation rates caused by the tedious nature of traditional methods were successfully overcome. The cognitive advantages of this approach were self-evident. The potential gains from the use of Computer Assisted Selfinterviewing technology was evaluated in validation studies.
III. COMPUTERS IN DIETARY ASSESSMENT
Next was the arrival of fully automated systems for selfassessment where one no longer required a dietician to assist in the interviewing process or to validate the results obtained from the system. Though there are a lot of controversies regarding this new technology, but research are going on to remove fallacies those are still present.
In the past three decades, computers have become an integral part of everyday life as they have gotten faster, more versatile, more reliable, and more user-friendly. Today's userfocused computers talk, show movies, correct mistakes, and check electronic mail automatically. The spread of computers from the scientist's laboratory to the family room, workplace, bank, and public library has been dramatic. Many applications now being developed requires citizens to deal with a computer to file records on obtain permits. Because of these changes, the feasibility of using computer-based interactive programs in the population at large is no longer an issue. In dietary assessment, computers have long been used extensively to simplify the tedious task of analyzing intake data. They were first used to calculate nutrient intakes from data on the types and amounts of foods eaten. In this role, they replaced a dietician with an adding machine, who needed several hours to analyze the intake of 12 essential nutrients in each dietary record from one person. Computers were next used to code dietary data. Two decades ago, food coders were employed to look up appropriate food codes in books and write them on paper forms for subsequent computerized foodto-nutrient conversion. People manually entered the data from the paper forms into the computers. Since then, optically readable forms have been developed to allow automated and direct coding of dietary data obtained by means of standard sequences of questions, such as those on food-frequency questionnaires. In a few instances, computers have been involved in guiding the wording, sequence, and depth of questions asked by interviewers but now technology has interviewers in expressing dietary assessment questions, making it a fully automated process. But the validity of such methods of full-automation is yet to be proved. So we are looking into a new age where computers will replace unnecessary labours. In the past, dietary assessment used to rely largely on the subjective assessment of habitual diet but then a few technological tools became available to help nutritionists and overcome many of the weaknesses of dietary assessment methods. Application of these tools used to motivate the subjects of studies and thereby improved participation and increased the quality of the information gathered. Generalized visual stimulation, video presentations, soundtracks with narration and music, and interactive components with immediate feedback were well received by respondents to dietary assessment interviews and enhance their interest and concentration. Computer Assisted Self-interviewing technology helped nutritional epidemiologists obtain more accurate and more detailed dietary information from diverse populations. Problems such as language interpretation, low
IV. CALCULATIONS The Harris–Benedict equation has been used for calculating daily energy requirements in Kcal for adult men & women. For men, e = 66.4730 + 13.7516 * w + 5.0033 * h - 6.7550 * a
For women, e = 655.0955 + 9.5634 * w + 1.8496 * h - 4.6756 * a
Where, e = Daily energy needs in Kcal/day, w = weight in kg, h = height in cm, a = age in years.
*Special Cases Special Groups 1 – 3 years 4 – 6 years 7 – 9 years 10 – 12 years 13 – 15 years 16 – 18 years Pregnant Lactation (first 6 months) Lactation (6 - 12 months)
Energy Requirements (in Kcal) 1240 1690 1950 2190(males) 1970(females) 2450(males) 2060(females) 2640(males) 2060(females) + 300 + 550 + 400
Daily Water Intake (in ml) = 50 * weight (in kg)
3 For calculation of Protein Energy Malnutrition among children, Waterlow classification has been used.
Given below are some screenshots of the prototype software.
Weight/Height (%) = ((Weight of the child) / (Weight of a normal child of the same height)) * 100 Height/Age (%) = ((Height of the Child) / (Height of a normal child of the same age)) * 100 According to Waterlow’s classification,
Normal Mild Moderate Severe
Weight for Height (wasting) > 90 80 – 90 70 – 80 < 70
Height for Age (stunting) > 95 90 - 95 85 - 90 < 85
Before inputting the fields by the user
Only calculations for the severe and moderate group has been performed and respective diets for has been prescribed. V. IMPLEMENTATION We have implemented a prototype in Java that can give the user a self-assessed report, when he/she input details like age, sex, height, weight, type of culinary preferred. And also details like, if the user is overweight, then how much should they exercise to remain healthy. In the prototype model the scope for customization for patients with chronic disease, and even detecting malnourished kid, and administering special diet for them. Only calories for a few Indian foods have been calculated along with their respective Kcal and added with the system. And a few basic exercises added. But the framework is there for expanding it later.
Inputting data from the user
Data Flow Diagram for the computerized self-assessment dietary system Computer Generated Diet Report for the user according to the specifications made by the user.
4 VI. TESTING OF THE PROTOTYPE PROGRAM
VII. CONCLUSION
Design is a critical aspect of a GUI program. Although it is valuable to have many capabilities available during the design process, it is an additional challenge to use them properly. Decisions concerning users are best made with input from users. In the development of our prototype program, end-user input about its design was obtained through focus group testing to evaluate menu structure and whether the program logic was appropriate and understandable. The dietary assessment program was tested in a group of 30 college-age adults, 20 high-school students and 20 primary school students, most of them used the program on their own. 12 students from the primary school even with basic computer literacy opted for guidance during the process, while four other showed complete lack of interest. Overall, the response to the program was positive in the three groups and the groups had no difficulty understanding or responding to the questions. The subjects from the first two groups felt quite comfortable completing the program independently. This group also expressed interest in the extending the prototype program as an android & iOS release for their smartphones. It maybe noted that the test subjects were not suffering from any chronic disease, or were in any special conditions like pregnancy or lactation. The accuracy was above 90%, and later verified by Dr. Swapan Kumar Biswas, ex-house physician NRS Medical College & Hospital.
In this study we have discussed about the complete automation of nutritional profile creator for a certain user. We have discussed the existing approaches in the field and proposed an alternative solution to completely automate the system. The process shown in the control flow structure has been implemented in a prototype in Java. The prototype was evaluated by testing it on 30 university students. A natural extension of this project is to expand the nutritional database used and include more prepared food that Indians generally have after getting feedback from the actual use of the prototype. The size of the database needs to be increased and better data structures needs to be used that can make the extension easier and the maintenance of the nutritional database simpler. Another future direction is to develop nutritional database for specific cases of chronic diseases, and quick detection of malnourished children and prescribing diet suited for them.
REFERENCES [1]
[2]
[3] [4] [5] [6] [7]
[8]
Accuracy of computerized self-assessment dietary system among various groups
Lenore Kohlmeier, Michelle Mendez, Jennifer McDuffie, and Mary Miller. Computer assisted self-interviewing: a multimedia approach to dietary assessment, 1997. Y. Probst and Linda C. Tapsell. Overview of Computerized Dietary Assessment Programs for Research and Practice in Nutrition Education, 2005. J. C. Waterlow (1972). "Classification and Definition of Protein-Calorie Malnutrition". British Medical Journal (5826): 566–569. J. Arthur Harris and Francis G. Benedict, A Biometric Study Of Human Basal Metabolism, 1918. Recommended Dietary Intakes for Indians: ICMR (Indian Council of Medical Research), 1990. K.Park, Park’s Textbook of Preventive and Social Medicine: Bhanot, 2009. Correy Foster, Neville F. Mistry, Parvin F. Peddi and Shivak Sharma,The Washington Manual Of Medical Therapeutics: Wolters Kluwer, 2010. John Macleod, Davidson’s Principles and Practice of Medicine: Churchill Livingstone, 1977.
