2022: Digital Human Carer Conversation

The Caring Economy Needs Help - Enter Digital Humans

Around the word, the caring economy is massive but failing those who need it most.

This is a real-life interaction between heart patient carer Marie Johnson and the AI powered Digital Human Cardiac Coach.

We believe this is the first carer interaction modelled with a digital human health coach.

The carer is often forgotten in a heart patient’s health journey, but so essential and potentially lifesaving. 

However, the carer is the one constant human presence in the lives of patients. 

And because of this, the carer is an essential participant in frontline healthcare, performing two critical functions.  Firstly, carers help the patient to understand what the health system is telling them. And secondly, the carer is responsible for procedural patient support such as wound care, bathing, remembering to take medications, eating, exercise and so on.

Here is an AI powered digital human helping a carer.

The common corpus of conversations supports multiple contexts, including direct patient conversations as well as carer conversations. 

The topic of this conversation that the carer is having with the coach – finding out what’s involved in going to hospital and then going home – is the same topic of conversation that the patient would have with the coach.  What is different is the context of the carer.

In this conversation, the carer not only asks questions, but also makes statements a number of times – such as “I’m not sure I’ll remember all this”.  This is based on real life.

In both the question interaction and statement interaction, the Coach is able to infer intent from the context and provide information in a natural contextual flow.

The questions (intents) we used are based on years of R&D including in the cardiac rehabilitation setting, patient forums globally, peer reviewed research and general patient education sourced globally.

Also demonstrated in this interaction, is the importance of diversity in the persona of the male digital human coach, with this digital human provided by Trulience. 

Diversity in the co-designed persona of the digital human coach is essential for providing people with choice so as to inclusively meet the full range of gender, age, disability, and cultural diversity across society.

Not Diagnosis and Treatment. Focus is health illiteracy and accessibility.

This conversation is not diagnosis or treatment. The conversation is co-designed to help address the impact of health illiteracy and isolation.

Why does this matter? 

18 million people die each year from cardiovascular disease, an estimated 32% of all deaths worldwide.

In Australia, CVD was the underlying cause of 25% of all deaths.  4 million Australians live with CVD.

And health illiteracy is a major factor in these statistics.

——————————————————————————-

The digital human male persona demonstrating diversity is provided by #Trulience from the United Kingdom. We built the Agent in Google Dialogflow using their free ES license and their ES Console. We created the Agent on the Google Dialogflow US server so that we could use the beta Knowledge Base functionality. This enabled us to upload our existing intents and responses via a csv file.

Trulience provide simple and clear instructions in their documentation on how to integrate a Google Dialogflow Agent with #Trulience. This involved creating a Google Cloud Platform Service Account and JSON key from our Agent. This only took a few minutes and it worked first time. We used the base #Trulience expression mark up as provided. 

The real-life interaction between Marie Johnson and Coach was recorded on a MacBook Pro with #Trulience running in Google Chrome. 

Stay tuned for more.