Deep neural networks are increasingly used in medical imaging for tasks such as pathological classification, but they face challenges due to the scarcity of high-quality, expert-labeled training data.

To solve the problem regarding unbalanced distribution of multi-category Chinese long texts and improve the classification accuracy thereof, a data enhancement method was proposed. Combined with this ...