Other Metrics to Collect and fix the image view

[Ivan-E-Johnson]

Implement a custom loss function that is a weighted combination of metrics. something like loss = .8 diceCE + .1 hassdorfs + .1 other

implement:

https://lightning.ai/docs/torchmetrics/stable/classification/label_ranking_average_precision.html#label-ranking-average-precision

https://lightning.ai/docs/torchmetrics/stable/pages/overview.html

https://github.com/Project-MONAI/tutorials/blob/main/modules/transform_visualization.ipynb

[Ivan-E-Johnson]

Something like this

class CustomLoss(nn.Module):
    def __init__(self, weights):
        super(CustomLoss, self).__init__()
        self.weights = weights
        # Initialize other necessary components here

    def forward(self, outputs, targets):
        # Calculate each metric
        dice_loss = DiceLoss(...)
        hausdorff_distance = HausdorffDistanceMetric(...)
        # Combine them with weights
        total_loss = self.weights['dice'] * dice_loss + self.weights['hausdorff'] * hausdorff_distance
        return total_loss