Nowadays, it is common for people to take photographs of every beverage, snack, or meal they eat and then post these photographs on social media platforms. Leveraging these social trends, real-time food recognition and reliable classification of these captured food images can potentially help replace some of the tedious recording and coding of food diaries to enable personalized dietary interventions. Although Central Asian cuisine is culturally and historically distinct, there has been little published data on the food and dietary habits of people in this region. To fill this gap, we aim to create a reliable dataset of regional foods that is easily accessible to both public consumers and researchers. To the best of our knowledge, this is the first work on the creation of a Central Asian Food Dataset (CAFD). The final dataset contains 42 food categories and over 16,000 images of national dishes unique to this region. We achieved a classification accuracy of 88.70% (42 classes) on the CAFD using the ResNet152 neural network model. The food recognition models trained on the CAFD demonstrate the effectiveness and high accuracy of computer vision for dietary assessment.

The burden of diet-related diseases is high in Central Asia. In recent years, the field of food computing has gained prominence due to advancements in computer vision (CV) and the increasing use of smartphones and social media. These technologies provide promising potential in many applications by facilitating real-time information retrieval from food images for efficient digital food journaling, smart restaurants, and supermarkets etc. Yet, to develop a robust CV model for food information retrieval, a large-scale high quality dataset is required. Several food dataset have been developed covering Western, Mediterranean, Chinese etc. cuisines. These dataset solve the simpler problem of food classification with single food item per image, which is not practical for real-life scenarios, where meals typically consist of multiple food items. To address this gap, we developed a large-scale high-quality Central Asian Food Scenes Dataset for food localization and detection. The dataset contains 21,306 images across 239 food categories, 69,856 instances. ed images. To evaluate the dataset, we performed the parametric experiments with the object detection models, with the best results achieved using YOLOv8xl (mAP50 score of 0.677).

While large language models like ChatGPT-4 have demonstrated competency in English, their performance for minority groups speaking underrepresented languages, as well as their ability to adapt to specific socio-cultural nuances and regional cuisines, such as those in Central Asia (e.g., Kazakhstan), still requires further investigation. To evaluate and compare the effectiveness of the ChatGPT-4 system in providing personalized, evidence-based nutritional recommendations in English, Kazakh, and Russian in Central Asia. This study was conducted from May 15 to August 31, 2023. Based on fifty mock patient profiles, ChatGPT-4 generated dietary advice, and responses were evaluated for personalization, consistency, and practicality using a 5-point Likert scale. To identify significant differences between the three languages, the Kruskal Wallis Test was conducted. Additional pairwise comparisons for each language were carried out using the post-hoc Dunn's Test. ChatGPT-4 showed a moderate level of performance in each category for English and Russian languages while in Kazakh language outputs were unsuitable for evaluation. The scores for English, Russian, and Kazakh were as follows: for personalization, 3.32±0.46, 3.18±0.38, and 1.01±0.06; for consistency, 3.48±0.43, 3.38±0.39, and 1.09±0.18; and for practicality, 3.25±0.41, 3.37±0.38, and 1.07±0.15, respectively. The Kruskal-Wallis test indicated statistically significant differences in ChatGPT-4's performance across the three languages (P<0.001). Subsequent post hoc analysis using Dunn's test showed that the performance in both English and Russian was significantly different from that in Kazakh. Our findings show that, despite using identical prompts across three distinct languages, the ChatGPT-4 's capability to produce sensible outputs is limited by the lack of training data in non-English languages. Thus, a customized LLM should be developed to perform better in underrepresented languages and to take into account specific local diets and practices.

Background: Lower BMI cutoffs as compared to standard cut-offs have been recommended to reduce the risk of obesity-related co-morbidities in some ethnic populations (e.g. south Asian and Chinese populations). Recent attempts have also been made to establish ethnicity-specific BMI cutoffs to identify individuals affected with obesity in relation to type 2 diabetes risk in multi-ethnic populations based in the UK and North America. However, to date, there is yet to have any published work done to identify these cut-offs in Central Asia populations nor specify any difference for genders even though the fat distribution varies amongst the different ethnic groups as well as between the genders. To the best of the authors’ knowledge, this is the first study exploring new BMI and WC cut-offs in this population.Methods: To address this gap, we used a database of secondary care electronic health records from the National Research Cardiac Surgery Center to identify BMI and waist circumference cutoffs for obesity based on the risk of developing diabetes and other cardiometabolic disorders among 297 adults in Kazakhstan. Bivariate and multivariable logistic regression analysis were utilized to investigate the relationships between risk factors and type 2 diabetes mellitus (T2DM). BMI and WC thresholds were predicted using the Youden index.Results: For an equivalent age-adjusted and sex-adjusted incidence of type 2 diabetes at a BMI of 30·0 kg/m² in White populations, we found higher BMI cutoffs for Kazakhstani women (30.5 kg/m²) but lower cut-offs for men (28·9 kg/m²). As for waist circumference, the cut-off points for females were 95cm and 104 cm for males.Conclusions: For Central Asia populations, the current recommended BMI and WC cutoffs may not be suitable and further work is needed to establish specific cut-offs for this population.

3D food printing (3DFP) is emerging as a vital innovation in the food industry’s pursuit of sustainability. 3DFP has evolved to significantly impact food production, offering the capability to create customized, nutritionally balanced foods. Central Asia has a higher than global average level of meat consumption per capita, which might be influenced by its historical and cultural background of nomadism. This dietary trend might potentially result in negative impacts on both the environment and human health outcomes, as it leads to increased greenhouse gas emissions and increased risk of chronic diseases. Reducing meat consumption holds the potential to address these sustainability and health issues. A possible strategy to reduce meat consumption and promote plant-based foods is 3D Food Printing (3DFP), which can rely on plant-protein sources from the region to create appealing and tasty alternatives for these populations. This review summarizes recent studies on plant protein-rich materials for 3DFP as a substitute to meet the growing global demand for meat as well as the 3DFP printing parameters associated with the different plant-based proteins currently used (e.g., lentils, soybeans, peas, and buckwheat). The findings revealed that buckwheat, a dietary staple in Central Asia, can be a promising choice for 3DFP technology due to its widespread consumption in the region, gluten-free nature, and highly nutritious profile.

Buckwheat (Fagopyrum esculentum) is a gluten-free crop valued for its protein, fiber, and essential minerals. This study investigates the rheological properties of buckwheat (BW) dough, both with and without the addition of gums (no gum, guar (GG), xanthan (XG)), at varying barrel temperatures (25, 55, and 85 °C) of the rheometer and at different water content levels (45, 50, and 55% w/w) to optimize dough formulations for 3D food printing. Using high shear stress capillary tests, the consistency coefficient (K) and flow behavior index (n) were measured. The results indicated that GG significantly increases the apparent viscosity of buckwheat dough across shear rates ranging from 200 to 2000 s⁻¹, under all temperature and water content conditions. XG also enhanced viscosity but to a lesser extent at moderate temperatures (55 °C, 85 °C). All BW dough formulations exhibited a non-Newtonian shear-thinning behavior, crucial for 3D printing applications. In addition, computational fluid dynamics (CFD) simulations were conducted to analyze the extrusion process of BW dough formulations (50% W, 50% W + XG, and 50% W + GG), focusing on shear rate, viscosity, and pressure distribution. The simulations demonstrated that shear rates increased as the dough moved through the nozzle, while viscosity decreased, facilitating extrusion. However, gum-added formulations required higher pressures for extrusion, indicating an increased difficulty in dough flow. The study highlights the complex interactions between temperature, water content, and additive type on the rheological properties of buckwheat dough, while also incorporating CFD simulations to analyze the extrusion process. These insights provide a foundation for developing nutrient-dense, gluten-free 3D-printed foods tailored to specific dietary needs.