Infrared guided smart food formulation: an innovative spectral reconstruction strategy to develop anticipated and constant apple puree products

Zhenjie Wang; Sylvie Bureau; Benoit Jaillais; Catherine M. G. C. Renard; Xiao Chen; Yali Sun; Daizhu Lv; Leiqing Pan; Weijie Lan; Zhenjie Wang; Sylvie Bureau; Benoit Jaillais; Catherine M. G. C. Renard; Xiao Chen; Yali Sun; Daizhu Lv; Leiqing Pan; Weijie Lan

doi:10.48130/fia-0024-0003

Figures (6) Tables (2)

Figure 1.
Experimental scheme of puree reformation, quality characterizations, and spectral acquisition.
Figure 2.
Boxplot of colors (a* and b*), rheological parameters (η₅₀), total sugars (TSC), titratable acidity (TA) and fructose of different formulated puree groups.
Figure 3.
Process of Vis-NIR spectral data by multivariate resolution alternative least square (MCR-ALS) and spectral reconstruction of formulated purees.
Figure 4.
Overview of the applied methodology of Vis-NIR spectra pre-processing, spectral variable selection, spectral reconstruction and multivariate regression.
Figure 5.
Principal component analysis (PCA) of the SNV pre-treated Vis-NIR spectra of four single cultivar purees. (a) Discrimination map of four apple cultivars. (b) Beta-coefficients of first principal component (PC1). (c) Beta-coefficients of second principal component (PC2).
Figure 6.
ANOVA results of the SNV pre-treated (a) visible (500–780 nm) spectra and (b) NIR (780–2,500 nm) spectra of all formulated apple purees.

Parameters	Range	SD	Methods	Variables	PLSR
Parameters	Range	SD	Methods	Variables	$\rm R_c^2 $	RMSEC	$\rm R_p^2 $	RMSEP	RPD
L*	41.6−48.9	1.5	FULL	2,722	0.87	0.6	0.73	0.7	1.76
			SPA	6	0.83	0.6	0.81	0.6	2.13
			CARS	51	0.87	0.6	0.81	0.6	2.14
			UVE	1,259	0.88	0.5	0.80	0.6	2.06
a*	(−4.8)−2.4	2.0	FULL	2,722	0.98	0.3	0.96	0.4	5.17
			SPA	6	0.98	0.3	0.96	0.4	5.24
			CARS	33	0.98	0.3	0.97	0.3	5.56
			UVE	1,596	0.98	0.3	0.97	0.4	5.38
b*	9.6−18.4	1.7	FULL	2,722	0.71	0.9	0.54	1.2	1.48
			SPA	7	0.73	0.9	0.58	1.2	1.54
			CARS	143	0.75	0.8	0.57	1.2	1.53
			UVE	997	0.75	0.8	0.59	1.2	1.55
Viscosity η₁₀₀	834−1,721	210	FULL	2,722	0.83	84.6	0.79	98.2	2.17
			SPA	9	0.79	95.5	0.82	92.4	2.30
			CARS	44	0.84	83.3	0.82	91.2	2.33
			UVE	1,193	0.87	76.0	0.85	85.4	2.49
Viscosity η₅₀	526−1,029	119	FULL	2,722	0.88	40.2	0.83	50.8	2.38
			SPA	11	0.82	49.0	0.82	51.5	2.35
			CARS	166	0.86	43.6	0.85	47.1	2.57
			UVE	1,133	0.88	40.9	0.87	44.5	2.73
DMC (g/g FW)	0.14−0.17	0.01	FULL	2,722	0.74	0.004	0.49	0.006	1.39
			SPA	9	0.56	0.006	0.50	0.006	1.42
			CARS	92	0.75	0.004	0.58	0.005	1.56
			UVE	1,497	0.80	0.004	0.63	0.005	1.66
TSC (g/kg FW)	93.2−145.4	12.6	FULL	2,722	0.96	2.7	0.90	3.6	3.57
			SPA	9	0.96	2.8	0.92	3.6	3.53
			CARS	101	0.97	2.7	0.92	3.5	3.66
			UVE	1,531	0.96	2.7	0.92	3.6	3.63
TA (meq/kg FW)	28.0−94.8	16.2	FULL	2,722	0.96	0.3	0.91	0.5	3.31
			SPA	10	0.94	0.4	0.90	0.5	3.12
			CARS	92	0.96	0.3	0.91	0.5	3.37
			UVE	1061	0.96	0.3	0.91	0.5	3.34
pH	3.39−4.47	0.23	FULL	2722	0.90	0.07	0.86	0.09	2.57
			SPA	8	0.69	0.13	0.69	0.14	1.76
			CARS	51	0.88	0.08	0.88	0.09	2.69
			UVE	1,385	0.88	0.08	0.87	0.09	2.73
Glucose (g/kg FW)	13.2−28.3	3.7	FULL	2,722	0.86	1.2	0.85	1.3	2.42
			SPA	5	0.85	1.3	0.84	1.4	2.51
			CARS	38	0.87	1.1	0.86	1.3	2.58
			UVE	1,181	0.86	1.2	0.85	1.3	2.54
Fructose (g/kg FW)	40.2−80.3	9.1	FULL	2,722	0.71	4.9	0.54	6.0	1.46
			SPA	12	0.59	5.8	0.57	6.2	1.42
			CARS	92	0.64	5.5	0.61	5.5	1.61
			UVE	1,235	0.63	5.6	0.61	5.5	1.61
Sucrose (g/kg FW)	33.2−57.3	5.5	FULL	2,722	0.74	2.8	0.66	3.1	1.72
			SPA	16	0.67	3.2	0.64	3.2	1.65
			CARS	92	0.73	2.9	0.63	3.3	1.61
			UVE	1,279	0.67	3.2	0.64	3.4	1.60
Malic acid (g/kg FW)	3.0−8.8	1.3	FULL	2,722	0.93	0.3	0.91	0.4	3.33
			SPA	20	0.91	0.4	0.90	0.4	3.16
			CARS	92	0.92	0.4	0.91	0.4	3.34
			UVE	952	0.92	0.4	0.92	0.4	3.36
$\rm R_c^2 $: determination coefficient of the calibration test; $\rm R_p^2 $: determination coefficient of the external prediction test; RMSEP: root mean square error of prediction test; RPD: the residual predictive deviation of the prediction test.

Table 1.

PLS prediction of physical, chemical and rheological parameters of all formulated purees using Vis-NIR (500–2,500 nm) spectra or their selected spectral variables based on SPA, CARS and UVE methods.

Parameters	Range	SD	LVs	PLSR
Parameters	Range	SD	LVs	$\rm R_c^2 $	RMSEC	$\rm R_p^2 $	RMSEP	RPD
L*	41.6−48.9	1.5	10	0.77	0.7	0.62	0.8	1.57
a*	(−4.8)−2.4	2.0	8	0.91	0.6	0.92	0.6	3.30
b*	9.6−18.4	1.7	9	0.58	1.0	0.42	1.4	1.31
Viscosity η₁₀₀	834−1721	210	10	0.81	87	0.81	96	2.22
Viscosity η₅₀	526−1029	119	10	0.82	48	0.82	54	2.26
DMC (g/g FW)	0.14−0.17	0.01	10	0.57	0.005	0.43	0.006	1.38
TSC (g/kg FW)	93.2−145.4	12.6	10	0.91	3.6	0.86	4.8	2.64
TA (meq/kg FW)	28.0−94.8	16.2	8	0.92	0.4	0.85	0.6	2.55
pH	3.39−4.47	0.23	10	0.84	0.09	0.85	0.10	2.47
Glucose (g/kg FW)	13.2−28.3	3.7	10	0.85	1.3	0.82	1.5	2.25
Fructose (g/kg FW)	40.2−80.3	9.1	10	0.68	5.0	0.60	5.6	1.58
Sucrose (g/kg FW)	33.2−57.3	5.5	13	0.82	2.3	0.77	2.6	2.08
Malic acid (g/kg FW)	3.0−8.8	1.3	8	0.93	0.3	0.86	0.5	2.67
$\rm R_c^2 $: determination coefficient of the calibration test; $\rm R_p^2 $: determination coefficient of the external prediction test; RMSEP: root mean square error of prediction test; RPD: the residual predictive deviation of prediction.

Table 2.

Prediction results of chemical and rheological parameters of all formulated purees from the reconstructed spectra computed by the concentration of MCR-ALS and the selected spectral variables of single-cultivar purees.