Mobile Computer Vision-Based Applications for Food Recognition and Volume and Calorific Estimation: A Systematic Review

Decoding Heterogeneity in Data-Driven Self-Monitoring Adherence Trajectories in Digital Lifestyle Interventions for Weight Loss: A Qualitative Study

Background

Data-driven trajectory modeling is a promising approach for identifying meaningful participant subgroups with various self-monitoring (SM) responses in digital lifestyle interventions. However, there is limited research investigating factors that underlie different subgroups. This qualitative study aimed to investigate factors contributing to participant subgroups with distinct SM trajectory in a digital lifestyle intervention over 6 months.

Methods

Data were collected from a subset of participants (n = 20) in a 6-month digital lifestyle intervention. Participants were classified into Lower SM Group (n = 10) or a Higher SM (n = 10) subgroup based on their SM adherence trajectories over 6 months. Qualitative data were obtained from semi-structured interviews conducted at 3 months. Data were thematically analyzed using a constant comparative approach.

Results

Participants were middle-aged (52.9 ± 10.2 years), mostly female (65%), and of Hispanic ethnicity (55%). Four major themes with emerged from the thematic analysis: Acceptance towards SM Technologies, Perceived SM Benefits, Perceived SM Barriers, and Responses When Facing SM Barriers. Participants across both subgroups perceived SM as positive feedback, aiding in diet and physical activity behavior changes. Both groups cited individual and technical barriers to SM, including forgetfulness, the burdensome SM process, and inaccuracy. The Higher SM Group displayed positive problem-solving skills that helped them overcome the SM barriers. In contrast, some in the Lower SM Group felt discouraged from SM. Both subgroups found diet SM particularly challenging, especially due to technical issues such as the inaccurate food database, the time-consuming food entry process in the Fitbit app.

Conclusions

This study complements findings from our previous quantitative research, which used data-drive trajectory modeling approach to identify distinct participant subgroups in a digital lifestyle based on individuals' 6-month SM adherence trajectories. Our results highlight the potential of enhancing action planning problem solving skills to improve SM adherence in the Lower SM Group. Our findings also emphasize the necessity of addressing the technical issues associated with current diet SM approaches. Overall, findings from our study may inform the development of practical SM improvement strategies in future digital lifestyle interventions.

Trial registration

The study was pre-registered at ClinicalTrials.gov (NCT05071287) on April 30, 2022.

A fine-grained recognition technique for identifying Chinese food images

As a crucial area of research in the field of computer vision, food recognition technology has become a core technology in many food-related fields, such as unmanned restaurants and food nutrition analysis, which are closely related to our healthy lives. Obtaining accurate classification results is the most important task in food recognition. Food classification is a fine-grained recognition process, which involves extracting features from a group of objects with similar appearances and accurately classifying them into different categories. In a such usage environment, the network is required to not only overview the overall image, but also capture the subtle details within it. In addition, since Chinese food images have unique texture features, the model needs to extract texture information from the image. However, existing CNN methods have not focused on and processed this information. To classify food as accurately as possible, this paper introduces the Laplace pyramid into the convolution layer and proposes a bilinear network that can perceive image texture features and multi-scale features (LMB-Net). The proposed model was evaluated on a public dataset, and the results demonstrate that LMB-Net achieves state-of-the-art classification performance.

Estimating Amount of Food in a Circular Dining Bowl from a Single Image

Unhealthy diet is a top risk factor causing obesity and numerous chronic diseases. To help the public adopt healthy diet, nutrition scientists need user-friendly tools to conduct Dietary Assessment (DA). In recent years, new DA tools have been developed using a smartphone or a wearable device which acquires images during a meal. These images are then processed to estimate calories and nutrients of the consumed food. Although considerable progress has been made, 2D food images lack scale reference and 3D volumetric information. In addition, food must be sufficiently observable from the image. This basic condition can be met when the food is stand-alone (no food container is used) or it is contained in a shallow plate. However, the condition cannot be met easily when a bowl is used. The food is often occluded by the bowl edge, and the shape of the bowl may not be fully determined from the image. However, bowls are the most utilized food containers by billions of people in many parts of the world, especially in Asia and Africa. In this work, we propose to premeasure plates and bowls using a marked adhesive strip before a dietary study starts. This simple procedure eliminates the use of a scale reference throughout the DA study. In addition, we use mathematical models and image processing to reconstruct the bowl in 3D. Our key idea is to estimate how full the bowl is rather than how much food is (in either volume or weight) in the bowl. This idea reduces the effect of occlusion. The experimental data have shown satisfactory results of our methods which enable accurate DA studies using both plates and bowls with reduced burden on research participants.

Assessing daily energy intake in adult women: validity of a food-recognition mobile application compared to doubly labelled water

Accurate dietary assessment is crucial for nutrition and health research. Traditional methods, such as food records, food frequency questionnaires, and 24-hour dietary recalls (24HR), have limitations, such as the need for trained interviewers, time-consuming procedures, and inaccuracies in estimations. Novel technologies, such as image-based dietary assessment apps, have been developed to overcome these limitations. SNAQ is a novel image-based food-recognition app which, based on computer vision, assesses food type and volume, and provides nutritional information about dietary intake. This cross-sectional observational study aimed to investigate the validity of SNAQ as a dietary assessment tool for measuring energy and macronutrient intake in adult women with normal body weight (n = 30), compared to doubly labeled water (DLW), a reference method for total daily energy expenditure (TDEE). Energy intake was also estimated using a one-day 24HR for direct comparison. Bland-Altman plots, paired difference tests, and Pearson's correlation coefficient were used to assess agreement and relationships between the methods. SNAQ showed a slightly higher agreement (bias = -329.6 kcal/day) with DLW for total daily energy intake (TDEI) compared to 24HR (bias = -543.0 kcal/day). While both SNAQ and 24HR tended to underestimate TDEI, only 24HR significantly differed from DLW in this regard (p < 0.001). There was no significant relationship between estimated TDEI and TDEE using SNAQ (R2 = 27%, p = 0.50) or 24HR (R2 = 34%, p = 0.20) and there were no significant differences in energy and macronutrient intake estimates between SNAQ and 24HR (Δ = 213.4 kcal/day). In conclusion, these results indicate that SNAQ provides a closer representation of energy intake in adult women with normal body weight than 24HR when compared to DLW, but no relationship was found between the energy estimates of DLW and of the two dietary assessment tools. Further research is needed to determine the clinical relevance and support the implementation of SNAQ in research and clinical settings. Clinical trial registration: This study is registered on ClinicalTrials.gov with the unique identifier NCT04600596 (https://clinicaltrials.gov/ct2/show/NCT04600596).

The Nutritional Content of Meal Images in Free-Living Conditions-Automatic Assessment with goFOODTM

A healthy diet can help to prevent or manage many important conditions and diseases, particularly obesity, malnutrition, and diabetes. Recent advancements in artificial intelligence and smartphone technologies have enabled applications to conduct automatic nutritional assessment from meal images, providing a convenient, efficient, and accurate method for continuous diet evaluation. We now extend the goFOODTM automatic system to perform food segmentation, recognition, volume, as well as calorie and macro-nutrient estimation from single images that are captured by a smartphone. In order to assess our system's performance, we conducted a feasibility study with 50 participants from Switzerland. We recorded their meals for one day and then dietitians carried out a 24 h recall. We retrospectively analysed the collected images to assess the nutritional content of the meals. By comparing our results with the dietitians' estimations, we demonstrated that the newly introduced system has comparable energy and macronutrient estimation performance with the previous method; however, it only requires a single image instead of two. The system can be applied in a real-life scenarios, and it can be easily used to assess dietary intake. This system could help individuals gain a better understanding of their dietary consumption. Additionally, it could serve as a valuable resource for dietitians, and could contribute to nutritional research.

Guidelines for microbiome studies in renal physiology

Gut microbiome research has increased dramatically in the last decade, including in renal health and disease. The field is moving from experiments showing mere association to causation using both forward and reverse microbiome approaches, leveraging tools such as germ-free animals, treatment with antibiotics, and fecal microbiota transplantations. However, we are still seeing a gap between discovery and translation that needs to be addressed, so that patients can benefit from microbiome-based therapies. In this guideline paper, we discuss the key considerations that affect the gut microbiome of animals and clinical studies assessing renal function, many of which are often overlooked, resulting in false-positive results. For animal studies, these include suppliers, acclimatization, baseline microbiota and its normalization, littermates and cohort/cage effects, diet, sex differences, age, circadian differences, antibiotics and sweeteners, and models used. Clinical studies have some unique considerations, which include sampling, gut transit time, dietary records, medication, and renal phenotypes. We provide best-practice guidance on sampling, storage, DNA extraction, and methods for microbial DNA sequencing (both 16S rRNA and shotgun metagenome). Finally, we discuss follow-up analyses, including tools available, metrics, and their interpretation, and the key challenges ahead in the microbiome field. By standardizing study designs, methods, and reporting, we will accelerate the findings from discovery to translation and result in new microbiome-based therapies that may improve renal health.