Validation of a novel image-weighed technique for monitoring food intake and estimation of portion size in hospital settings: a pilot study

Navigating next-gen nutrition care using artificial intelligence-assisted dietary assessment tools-a scoping review of potential applications

Introduction

Recent developments in Artificial Intelligence (AI) and Machine Learning (ML) technologies have opened new avenues for their applications in dietary assessments. Conventional dietary assessment methods are time-consuming, labor-driven, and have high recall bias. AI-assisted tools can be user-friendly and provide accurate dietary data. Hence, this review aimed to explore the applications of AI-assisted dietary assessment tools in real-world settings that could potentially enhance Next-Gen nutrition care delivery.

Materials and methods

A total of 17,613 original, full-text articles using keywords such as "artificial intelligence OR food image analysis OR wearable devices AND dietary OR nutritional assessment," published in English between January 2014 and September 2024 were extracted from Scopus, Web of Science, and PubMed databases. All studies exploring applications of AI-assisted dietary assessment tools with human participation were included; While methodological/developmental research and studies without human participants were excluded as this review specifically aimed to explore their applications in real-world scenarios for clinical purposes. In the final phase of screening, 66 articles were reviewed that matched our inclusion criteria and the review followed PRISMA-ScR reporting guidelines.

Results

We observed that existing AI-assisted dietary assessment tools are integrated with mobile/web-based applications to provide a user-friendly interface. These tools can broadly be categorized as "Image-based" and "Motion sensor-based." Image-based tools allow food recognition, classification, food volume/weight, and nutrient estimation whereas, Motion sensor-based tools help capture eating occasions through wrist movement, eating sounds, jaw motion & swallowing. These functionalities capture the dietary data regarding the type of food or beverage consumed, calorie intake, portion sizes, frequency of eating, and shared eating occasions as real-time data making it more accurate as against conventional dietary assessment methods. Dietary assessment tools integrated with AI and ML could estimate real-time energy and macronutrient intake in patients with chronic conditions such as obesity, diabetes, and dementia. Additionally, these tools are non-laborious, time-efficient, user-friendly, and provide fairly accurate data free from recall/reporting bias enabling clinicians to offer personalized nutrition.

Conclusion

Therefore, integrating AI-based dietary assessment tools will help improve the quality of nutrition care and navigate next-gen nutrition care practices. More studies are required further to evaluate the efficacy and accuracy of these tools.

Automated food intake tracking requires depth-refined semantic segmentation to rectify visual-volume discordance in long-term care homes

Malnutrition is a multidomain problem affecting 54% of older adults in long-term care (LTC). Monitoring nutritional intake in LTC is laborious and subjective, limiting clinical inference capabilities. Recent advances in automatic image-based food estimation have not yet been evaluated in LTC settings. Here, we describe a fully automatic imaging system for quantifying food intake. We propose a novel deep convolutional encoder-decoder food network with depth-refinement (EDFN-D) using an RGB-D camera for quantifying a plate's remaining food volume relative to reference portions in whole and modified texture foods. We trained and validated the network on the pre-labelled UNIMIB2016 food dataset and tested on our two novel LTC-inspired plate datasets (689 plate images, 36 unique foods). EDFN-D performed comparably to depth-refined graph cut on IOU (0.879 vs. 0.887), with intake errors well below typical 50% (mean percent intake error: [Formula: see text]%). We identify how standard segmentation metrics are insufficient due to visual-volume discordance, and include volume disparity analysis to facilitate system trust. This system provides improved transparency, approximates human assessors with enhanced objectivity, accuracy, and precision while avoiding hefty semi-automatic method time requirements. This may help address short-comings currently limiting utility of automated early malnutrition detection in resource-constrained LTC and hospital settings.

An Artificial Intelligence-Based System for Nutrient Intake Assessment of Hospitalised Patients

Regular nutrient intake monitoring in hospitalised patients plays a critical role in reducing the risk of disease-related malnutrition (DRM). Although several methods to estimate nutrient intake have been developed, there is still a clear demand for a more reliable and fully automated technique, as this could improve the data accuracy and reduce both the participant burden and the health costs. In this paper, we propose a novel system based on artificial intelligence to accurately estimate nutrient intake, by simply processing RGB depth image pairs captured before and after a meal consumption. For the development and evaluation of the system, a dedicated and new database of images and recipes of 322 meals was assembled, coupled to data annotation using innovative strategies. With this database, a system was developed that employed a novel multi-task neural network and an algorithm for 3D surface construction. This allowed sequential semantic food segmentation and estimation of the volume of the consumed food, and permitted fully automatic estimation of nutrient intake for each food type with a 15% estimation error.