Effectiveness of neural language models for word prediction of textual mammography reports

Radiologists are required to write free paper paper text reports for breast screenings in order to assign cancer diagnoses in a later step. The current procedure requires a lot of time and needs efficiency. To streamline the writing process and keep up with the specific vocabulary, a word prediction tool using neural language models was developed. Challenges as different languages (English,Dutch), small data sizes and low computational power have been overcome by introducing EnDuRLM process, able to improve by 25\% the current workflow according to RPE measurement. After defining model architectures, EnDuRLM process involves data preparation, hyperparameters optimization, configuration transfer and evaluation. This work supports future research involving other languages and also an extensive set of real-world applications.