“Lucy Conversation Agent For Healthcare” aims to revolutionize the way healthcare is delivered by enabling users to obtain accurate diagnosis for health disorders online. Our current target is the diagnosis of gastrointestinal disorders and heart diseases. Ultimately we aim to provide a platform for that will include diagnosis for various other disorders as well.
Why Online Diagnosis?
Cost of training a General Practitioner (GP) is high and even they can care only for a limited number of patients a day and are not always available. Medical professionals interacting with users online in real time can interact with only a few patients at a time and they could be interacting with users that are only curious and not actually sick, which would be a waste of their time.
eHealth is a term for healthcare that is supported by electronic communication. eHealth would greatly benefit developing countries and remote areas where there is a lack of proper infrastructure/equipment or availability of medical professionals as eHealth requires only a satellite broadband to be established and this can provide vast coverage. There are various approaches to online diagnosis such as blogs, wikis, patient doctor match, online appointment systems and clinical decision support.
Conversational Agent and its benefits
A Conversation/Intelligent/Virtual Agent is a program that interprets and responds to statements made by the users in natural language. It is designed to communicate with a real person efficiently based on the natural language interaction interfaces. These agents could also be text or voice based. When users search for health information online they may be presented with vast information that could be misleading and users may find it difficult to find relevant information in context of their medical condition. There is a lot of medical jargon that users that are not medical professionals cannot understand. Having real time conversation about symptoms and obtaining health information would greatly benefit the users. Another major benefit is that they are available all the time. Conversation agents can be designed to be friendly to encourage people to be more honest and open about their health conditions.
Ours is chat based application with a conversational agent called Lucy which converses with users about their lifestyle, diet and other factors to determine if the users are prone to the diseases. The service can be availed from any device available on the internet. Our approach combines questionnaire based diagnosis with objective diagnosis using heart/bowel sounds. The flow of the conversation will depend on the disease the users wish to be diagnosed for (gastrointestinal or heart disease). The combined approach helps users get a more reliable diagnosis in real time without having to visit a doctor for diagnosis.
Why heart diseases?
31% of all deaths worldwide is due to cardiovascular diseases with 80% occurring in low and middle income countries as of 2013. According to American Health Association, 2013, cardiovascular disease is the primary cause of death worldwide. It is also the leading cause of preventable death. Cardiac auscultation is the widely used technique to evaluate cardiac functions and detect. Although auscultation is a cheap and efficient method, diagnosis based on heart sounds is a difficult skill to master. A wrong diagnosis could not only incur unnecessary medical expenses but also cause emotional stress to patients and their families.
Why gastrointestinal diseases?
About 60 to 70 million US citizens suffer from Gastrointestinal (GI) diseases annually. In 2004, there were an estimated 4.6 million hospitalizations, 72 million ambulatory care visits, and 236,00 deaths that are attributed to Gastrointestinal disease. The expenditure of GI disease in the US, are estimated to be around $142 billion per year.8 Some causes of gastrointestinal disorders are food intake, sedentary lifestyle, intolerance/allergy to certain food items or a change in lifestyle. Bowel sounds caused due to a GI disease is different from normal bowel sounds. Having an online diagnosis with a combination of questionnaire and bowel sound analysis would greatly benefit the patients.
The current functionalities include uploading an application greeter, help menu, file uploader and heart rate detector.
The project is being developed in python by developers Christina, Huang and Noe.