Text this: Transformer-Based Encoder-Decoder Model for Medical Image Captioning with Concept Embedding