Physician Dictation Audio Data datasets for Machine Learning

257,977 hours of Real-world Physician Dictation Speech Dataset from 31 specialties’ to train Healthcare Speech models. Our de-identified dataset for healthcare include 31 different specialties audio files dictated by physicians describing patients’ clinical condition and plan of care based on physician-patient encounters in the hospital/clinical setting.

本数据集包含257,977小时的现实世界医师语音记录数据，涵盖31个专业领域，旨在训练医疗语音模型。该去标识化医疗数据集包含31个不同专业领域的音频文件，由医师在医院/临床环境中对患者的临床状况及基于医患互动的护理计划进行描述。