Client is a group of physicians delivering healthcare services across U.S. We helped Physicians automate the clinical documentation process with digital scribes for faster transcriptions, as well as real-time diagnostic analysis using speech recognition and NLP techniques.

ClientHealthcareServicesSpeech Recognition, NLPYear 2022

Key Challenges

  • Physicians were spending most of their time on clinical documentation of patient encounters and this was impacting the quality of healthcare
  • Only around 40% of doctor’s time was left to meeting patients, while the rest was spent on administration. Freeing up doctor’s time for patients was their focus​
  • They also wanted to reduce the manual process of transcription which at times can be error prone leading to huge operation costs

Technologies

Python
microsoft sql server
tensor-flow
opennlp

Solution

  • To address the challenges of our client, we suggested that we implement a digital scribe/AI Medical assistant that can: ​
    • Record the Physician – Patient conversation
    • Convert the audio into text and
    • Extract the relevant information for the summarization of the patient encounter
  • Developed a digital scribe using Speech Recognition libraries, AI and ML techniques to record and transform clinical interactions into meaningful medical notes through ambient listening and subsequent voice-to-text conversion
  • Implemented NLP techniques to mine useful data from transcribed medical notes such as symptoms, diseases, and prescription etc.
  • This data was used for better report generation which helped physicians save time when going through patient’s health records
  • With the solution, physicians and clinicians can document and record the patient interactions on their mobile, laptop or directly into the EHR systems

Benefits

  • Our client experienced a 20% increase in appointment attendance by sending out reminders and obtaining confirmations
  • Reduced waiting period of patients by providing 24/7 patient service with automated calls, messages at low operational costs
  • Our solution saved more than 20hrs per month of doctor’s time to efficiently use it for better patient care rather than focus on support tasks
  • Increased patient satisfaction with faster, easier appointment scheduling