Using tabular data in addition of image data in efficientnet

Hi I have some MRI images from kidneys. I am using effiecientnet for image classification to classify images in one of the stage 0, stage1 or stage 2 cancer classes. Because I didnt get good results just from image classification I wanted to use one column of clinical data that has tumor size alongside the images. Because I am not proficient in pytorch I faced this question that how can I merge one column of clinical data with images? Because I am using effiecientnet I can not use two inputs to the model how can I give tumor sizes to efficientnet? Should I change the architecture of efficientnet? Do you have another idea for this problem?