I have been trying to run your model to predict both lymph node status as well as tumor grade but the model doesn't seem to be performing too well after tuning the hyperparameters. I have a large dataset (1000 images) and I think the data quality is high. I'm wondering if feature extraction process is leaving out valuable information in the slides. Is there a way to run the model directly on the image patches before extraction?
I have been trying to run your model to predict both lymph node status as well as tumor grade but the model doesn't seem to be performing too well after tuning the hyperparameters. I have a large dataset (1000 images) and I think the data quality is high. I'm wondering if feature extraction process is leaving out valuable information in the slides. Is there a way to run the model directly on the image patches before extraction?