How we use AI

Heidi works by transcribing speech into text from a healthcare encounter such as conversations between clinicians and patients or by clinicians dictating their clinical findings, impression and/or management plans before, during and after the healthcare encounter. The clinician can also add additional contextual notes about the healthcare encounter. This system is designed to alleviate the administrative burden on healthcare professionals, allowing them to focus more on patient care rather than paperwork. The Heidi Scribe will leverage natural language processing (NLP), speech recognition technology, and machine learning algorithms to understand and interpret complex medical dialogue, identify key health information, and categorise data into the appropriate sections of an Electronic Health Record (EHR).

Your consent will be sought for consultations that are transcribed using the Heidi AI tool.  Heidi also uses aggregated de-identified information from these consults to improve its models and outputs, ultimately improving both patient care and clinician experience.

All Data that identifies you stays within the practice and its servers which are UK based, no identifiable data is used by the Heidi tool for machine learning.

Heidi AI will not make decisions about your care, it only transcribes verbal interactions with the practice, with your consent.

More information about the model can be found on the Heidi website here

Accurx Scribe works in a similar way, by transcribing audio from a patient contact or free-dictation and uses AI to summarise them into structured form for medical notes. The clinician can instantly modify notes and save these notes to the patient record.

The audio stream is processed in real-time during conversations and automatically deleted as soon as the audio is transcribed by Accurx Scribe. This ensures that the audio, which is highly sensitive, is not retained any longer than necessary to fulfil their purpose.

Your consent will be sought for consultations that are transcribed using the Accurx scribe tool.

Accurx Scribe does not make any automated decisions that affect patients or their care. While the system uses AI to assist with transcribing and summarising clinical conversations, it is not designed to function independently or replace clinical judgement. All outputs generated by the AI, such as transcriptions and summaries must be reviewed and approved by a healthcare professional before being saved to record or acted upon.

This ensures that healthcare professionals remain fully in control of the information recorded in a patient’s health record and that no decisions are made without appropriate human oversight. As such, the system does not engage in solely automated decision-making under the UK GDPR definition.

More information can be found on the Accurx website here.