Foods· 2026Q1
Development of Candied Strawberry Products via Neural Network-Optimized Sugar-Infiltration Processes
- 0citations
- Q1SCImago
- 2026year
Short summary
A back-propagation neural network (BPNN) model achieved superior prediction accuracy (0.95% error for HHP, 0.47% for US) compared to response surface methodology (RSM) in optimizing non-thermal sugar infiltration for candied strawberries, significantly enhancing nutrient retention.
AI-generated from the title and abstract; the full text is not read.
Key points
- Back-propagation neural network (BPNN) demonstrated superior predictive accuracy over RSM for optimizing HHP and US-assisted sugar infiltration.
- Optimal HHP parameters: 7.70 min infiltration, 258.00 MPa pressure, 45.00% sugar solution.
- Optimal US parameters: 50.00 min treatment, 80.00% power, 45.00% sugar solution.
- Non-thermal treatments (US, HHP) yielded 6.8x more total phenolics and 9.5x more total acids than thermal processing.
- Non-thermal methods reduced total sugar content by ~24% compared to thermal processing.
AI-generated from the title and abstract; the full text is not read.
Abstract
Conventional thermal processing of candied strawberries causes severe nutrient degradation and impairs the nutritional quality of the final products. To address this problem, high hydrostatic pressure (HHP) and ultrasound (US) were adopted as non-thermal auxiliary techniques for sugar-infiltration treatment in this study. Single-factor experiments were performed, and response surface methodology (RSM) and back-propagation neural networks (BPNN) were applied to optimize the processing parameters, with the sensory score used as the evaluation indicator. The optimization performance of the two models was then compared. Thermally processed candied strawberries served as the control group, and the moisture, total sugar, total acid, total phenolic, and total flavonoid contents of all sample groups were determined. The BPNN produced smaller relative errors than RSM in both HHP-assisted (0.95% vs. 1.15%) and US-assisted (0.47% vs. 0.68%) sugar-infiltration optimization, which confirms its superior predictive capability. The optimal HHP-assisted parameters were an infiltration time of 7.70 min, a treatment pressure of 258.00 MPa, and a sugar-solution concentration of 45.00%. The optimal US-assisted conditions were a treatment time of 50.00 min, an ultrasonic power of 80.00%, and a sugar-solution concentration of 45.00%. No significant differences in moisture content were observed among the three processing groups (thermal processing, 19.45 ± 0.35 g/100 g; US treatment, 19.51 ± 0.26 g/100 g; HHP treatment, 19.88 ± 0.21 g/100 g). The non-thermal treatments markedly enhanced nutrient retention: the total phenolic contents of the US- and HHP-treated samples reached 4.85 mg/g and 4.89 mg/g, respectively, which were 6.8-fold higher than that of the thermally processed samples (0.71 mg/g), and the corresponding total acid contents were 5.93 g/kg and 6.19 g/kg, approximately 9.5-fold higher than that of the thermal group (0.64 g/kg). In contrast, the thermally processed samples showed significantly higher total sugar contents than the non-thermal treatments. Both HHP and US processing effectively suppressed the thermal degradation of phenolic compounds and organic acids and reduced the total sugar content by approximately 24%. These results indicate that non-thermal processing is an effective route for manufacturing low-sugar, high-nutrition candied strawberry products.
The authors' abstract, as published at the source. Foods, 2026 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
BiotechnologyBiochemistry, Genetics and Molecular Biology