Algorithm 1: Interpretability algorithm for training the dataset of the neural-network models

1. Input: the characteristics of the affected parts of the organ as per the medical image

2. Variables = the set of the characteristics of the affected parts of the organ

/*3. For each variable assign a relative weight*/

$Relativeweight=\frac{Weightofthevariable}{{\displaystyle \sum }Weightsofallvariables}$

/*4. Generate the probabilities of having the disease*/

−LR + = relative weights of the variables

+LR = 1 − (−LR)

5. Output: the positive and negative probabilities in addition to the relative weights of the variable

End