Helseboka provides the patient with an overview of their own health. It allows easy access to their journals, future appointments, lab-results and so on.
When a patient schedules a doctor’s appointment via Helseboka, the system asks the patient a series of questions, about their symptoms. This way, the patient can describe their symptoms at depth, from the comfort of their own home, without time-pressure or fear of forgetting important details. Furthermore, the doctor won’t have the spend the first 5 out of 20 minutes on initial diagnosis and will have a better basis for preparation.
Intelligent diagnostic tool
An underlying intelligence, Intmed, was built to allow patients to get the help effectively and accurately they need.
At its core, Intmed provides an intelligent question-engine that supports doctors in diagnosing patients following a brief interview. The process is fully automated – the engine will pose questions about symptoms to the patient and depending on their answers throughout the interview, will provide a trustworthy diagnosis. The algorithm takes the patient’s existing health data into account. For example, if the patient has diabetes the questions are adjusted to fit this.
The diagnosis of a patient
In fact, trustworthiness is a fundamental part of Intmed’s design. The diagnosis of a patient may be a question of life and death. As such, transparency and trustworthiness are paramount in making such a solution viable. Although deep learning has made incredible strides in machine intelligence and has become the new default approach to most things intelligent, this explainability and transparency is simply not possible using standard deep learning methods. So, a novel Bayesian-network-based prediction model with accuracy and explainability as the key acceptability metrics was developed.
As a result, every decision made by the system – from individual question selection to the final diagnosis – is completely transparent and directly supported by empirical probabilities learned from real-world patient health data.
Empirically supported report
At any time, a doctor can receive an empirically supported report of the decision-making process conducted by the model to investigate how and why a diagnosis was reached.
Not only does this increase trustworthiness – it directly enables and incentivises the user to become part of the decision-making process rather than a simple bystander. Currently, Helseboka is used by several hundred healthcare professionals in Norway, including doctors, nurses, therapists, chiropractors, etc. The solution has a 5-digit number of users in total.
Currently, Helseboka is used by several hundred healthcare professionals in Norway, including doctors, nurses, therapists, chiropractors, etc. The solution has a 5-digit number of users in total.