**2. Methodology**

#### **2.1 Data obtention, selection and homogenization**

The data used for the present study was obtained from a search of different previously published information by other authors. A total of 11 studies were preselected and a final group of only 4 studies (8 data sets) was used to obtain data for the food systems analyzed (**Table 1**). The selection was based on the type of food (solid foods), number of points in the curves (at least six points), number of repetitions (at least three repetitions with mean and standard deviation) and inactivation as a function of Total Dose (*T.D.,* J cm�<sup>2</sup> ).

In some cases, the inactivation curves were given as a function of treatment times; therefore, to obtain the total dose (*TD*), Eq. (1) was used [18]. Where *e*<sup>1</sup> (Jcm�<sup>2</sup> ) is the energy of a single pulse, *t* (s) is the treatment time and *f* (Hz) is the frequency of the light pulses.

$$\mathbf{TD} = \mathbf{e}\_1 \cdot \mathbf{t} \cdot \mathbf{f} \tag{1}$$

### **2.2 Data modeling using the Weibull model**

The logarithmic reduction values (log (*N/N*0)) were obtained from the interpretation of the inactivation curves reported on each article, where *N* is the survival population after HILP treatment and *N*o is the initial number of microorganisms.

The inactivation of microorganisms was fitted using the Weibull model (Eq. (2)), where log (*N/N*0) is the survival ratio after HILP treatment,*TD* is the total dose (Jcm�<sup>2</sup> ) required to achieve a certain amount of microbial log reduction, *b* (min�<sup>1</sup> ) is equivalent


#### **Table 1.**

*Summary of food data sources and corresponding authors, used for the HILP modeling analysis.*

to the inactivation rate constant while the exponent *n* indicates the shape of the curve (*n* < 1, upward-concave curve; *n >* 1*,* downward-concave curve) [12, 19, 20].

$$
\log\left(\frac{N}{N\_0}\right) = -bTD^n\tag{2}
$$
