Jun 09, 2026
Protocol for the Generation and Statistical Validation of the SMARTCLOTH-Database for Dietary Control in Diabetes
- María García-Rodríguez1,
- Fernando León-García2,3,4,
- Rafael Molina-Luque5,3,4,
- María Pilar Villena-Esponera6,3,4,
- Manuel Romero-Saldaña5,3,4,
- Guillermo Molina-Luque5,3,4
- 1Faculty of Health Sciences. UTAMED (Universidad Tecnológica Mediterráneo), Campanillas, Málaga, Spain.;
- 2Department of Electronic and Computer Engineering, University of Córdoba, Spain.;
- 3Lifestyles, Innovation and Health research group (GA-16). Maimonides Biomedical Research Institute of Cordoba (IMIBIC), Córdoba, Spain.;
- 4Advanced Informatics Research Group (GIIA), TIC-252. University of Córdoba, Spain.;
- 5Nursing, Pharmacology and Physiotherapy Department. University of Córdoba, Spain.;
- 6Universidad Internacional de La Rioja, Logroño, Spain.
Protocol Citation: María García-Rodríguez, Fernando León-García, Rafael Molina-Luque, María Pilar Villena-Esponera, Manuel Romero-Saldaña, Guillermo Molina-Luque 2026. Protocol for the Generation and Statistical Validation of the SMARTCLOTH-Database for Dietary Control in Diabetes. protocols.io https://dx.doi.org/10.17504/protocols.io.x54v9q5ypl3e/v1
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: Working
We use this protocol and it's working
Created: March 18, 2026
Last Modified: June 09, 2026
Protocol Integer ID: 313446
Keywords: database for dietary control, statistical validation of the smartcloth, execution of the smartcloth validation pipeline, smartcloth validation pipeline, diabetes this protocol, calculation of nutritional value, dietary control, food group system, nutritional value, statistical validation, smartcloth model, smartcloth, final statistical analysis in jamovi, intake control in people, diabetes, generation of synthetic menus, synthetic menus, validation
Funders Acknowledgements:
Instituto de Salud Carlos III (ISCIII), Spanish Ministry of Science and Innovation, and co-funded by the European Union. Strategic Action in Health (Acción Estratégica en Salud, AES 2021)
Grant ID: PI21/01602: SMARTCLOTH. Design, implementation and efficacy study of a digital tablecloth for dietary self-management in patients with diabetes.
Abstract
This protocol describes the steps necessary to reproduce the results described in the paper "Development and validation of a food group system for intake control in people with diabetes. SMARTCLOTH-Database.". The workflow includes execution of the SMARTCLOTH validation pipeline from the project root, generation of synthetic menus, calculation of nutritional values under the BDCA/FCDB reference model and the SMARTCLOTH model, and export of the resulting outputs as JSON and Excel files. The protocol also describes how to use the exported `validation_dataset.xlsx` file as the structured input for the final statistical analysis in JAMOVI, including descriptive comparisons, intraclass correlation analysis, and Bland–Altman agreement assessment at the aggregation levels reported in the manuscript.
Run the SMARTCLOTH validation pipeline
34m 50s
In the terminal, clone the public code repository from GitHub. You must have Git installed on your computer
2m
Change to the root directory of the repository
cd smartcloth-gendie
10s
Create a Python virtual environment in the project root:
python -m venv .venv
45s
Activate the virtual environment.
On Linux/macOS:
source .venv/bin/activate
On Windows:
.venv\Scripts\activate
10s
Install the required dependencies:
pip install -r requirements.txt
1m
Verify that `models.json` and `generator_rules.json` are located in the `data/` directory.
30s
Run the pipeline:
python main.py --seed 42
4m
Confirm that `output.json`, `1_models.xlsx`, `2_menus.xlsx`, `3_analysis.xlsx`, `frecuencies.xlsx`, and `validation_dataset.xlsx` are generated in the `outputs/` directory.
30s
Open `validation_dataset.xlsx` and verify that it contains the worksheet `validation_dataset`.
30s
Verify that the `validation_dataset` worksheet contains rows labelled by analysis level in the `Level` column, including:
- `FOOD`,
- `DISH`,
- `GROUP`,
- `DAY`.
2m
Verify that each row contains the nutritional values for both models (BDCA/FCDB reference values and SMARTCLOTH values) for:
- energy (kcal),
- carbohydrates (g),
- proteins (g),
- lipids (g),
together with the corresponding differences:
- `Dif. Kcal`,
- `Dif. Carb`,
- `Dif. Prot`,
- `Dif. Lip`.
2m
Install JAMOVI Desktop (version 2.3.28 or a compatible version) on your computer.
3m
Import `validation_dataset.xlsx` into JAMOVI for the final statistical analys
30s
In JAMOVI, use the `Level` column to subset the dataset according to the aggregation level required for each analysis:
- `FOOD` for individual food items,
- `DISH` for composite dishes,
- `GROUP` for food-group level summaries,
- `DAY` for complete daily menus.
2m
In JAMOVI, assess the distribution of the differences using:
- Kolmogorov–Smirnov test with Lilliefors correction,
- P–P plots,
- Q–Q plots.
3m
In JAMOVI, calculate the intraclass correlation coefficient using the ICC(3,k) model with:
- two-way mixed effects,
- absolute agreement,
- 95% confidence intervals.
3m
Perform the ICC analysis independently at the aggregation levels reported in the manuscript.
3m
Generate the summary tables reporting the nutritional differences between SMARTCLOTH and FCDB, including mean (SD) and range (minimum–maximum).
3m
Generate Bland–Altman plots to evaluate agreement between both methods.
3m
Export the statistical outputs and figures to obtain the final validation results reported in the manuscript.
45s