

{module 163}
2020 and COVID-19 had an incredible effect on the IT sector, a huge variety of services took off and gained momentum that were once deemed non-viable. Besides the education and service sectors, healthcare was one of the first areas where this effect was felt, and innovators were quick to take advantage of opportunities.
As our partner has a successful dental practice, he had a first-hand experience in witnessing the changing needs, so he decided to offer a solution for those needs.
It can never be said enough to novice businesses that an idea is nothing without a problem to solve. In this case the problem was quite obvious. COVID-19 caused closures, restrictions, and quarantine which meant that healthcare services were only available in urgent situations. The same conditions were present in dental care. There was one question though - are people capable of deciding whether certain complaints can be defined as urgent? Based on his years of experience, our partner didn’t think so... An interesting tendency started to develop among patients where they contacted dentists on various social platforms, described their symptoms with a photo attached, and inquired whether they are entitled to a dental care.
This obviously wasn’t a viable model for various reasons, pictures weren’t of good quality, symptoms weren’t described accurately, and patients often contacted dentists who they found on social media but had an office far away from the patient, so the possibility of actual dental care was very low.
To provide a solution for these problems, our client came up with the idea for a service where patients can describe their symptoms, state their COVID-19 situation, and they can take photos of the affected area and send them to their doctor, or if their don’t have one, to another dentist who is in the system. On this other end of the solution, dentists can see the incoming issues and can decide if the patient requires urgent care, and they can give tips for at-home remedies.
Our partner contacted us to help design the solution and develop an MVP that can be immediately used involving a few doctors.
We started the project with a UX design process, and the final plans were used by the development team for the development process. For a project where end-users will use the application it’s vital to employ a design methodology that is based on understanding the users, their behavior and their motivation, as they are needed for creating a successful product and for minimizing the risks related to market entry. The UX design process which puts the user at the forefront substantially decreases those risks by involving real user at several stages during the process.
The solution consisted of 2 parts, a mobile application for patients and a web application for dentists.
After understanding the target group, we defined the following criteria for the mobile application:
We had several discussions about the first point and explored various possibilities, but in the end, we decided together with our client that a custom-made form is the best way to go about it. The form presents the questions depending on the answers to the previous questions, so the dentist gets the most accurate report of the problem.
Our client compiled the questions in cooperation with several dental specialists, and they put the questions into 8 basic groups. There are overlaps as well as division in the question groups for which we created a logical explanation in Flowmapp.
Naturally, users don’t see the elaborate structure, they simply get the relevant questions based on their answers, so they can fill out the form easily. During the design process we made an effort to create a simple form, where most of the answers can be given with a click.
For an application like this, a smooth user communication is vital and can make or break a solution, so we always consider the following:
There is nothing revolutionary in our solution, but we still managed to create a process where the user feels safe and assured. To achieve this, we used elements like:
summary pages - we placed summary pages to certain sections of the process where the user can see the answers they have given.
The third criteria brought a technical challenge. What can we do to make sure that the user can take clear pictures of their teeth (even their back teeth)? We definitely needed to consider the different types of phone cameras. The solution ended up being a multi-secured process which we expect to be the most effective.
The other end of the solution was designed for dentists. Besides their own patients, dentist can treat anonymous patients as well. Those are the patients that didn’t chose a particular doctor but would like to receive care as soon as possible. Dentists can filter for patients based on location, so they have the opportunity to bring new patients to their practice.
When they choose a patient, they can see their COVID-19 form, their symptoms, and the attached photos. Based on those they can decide if the problem requires urgent care or it’s not an emergency situation. Dentists use the information for diagnostic reasoning and provide the patient with at-home care instruction that help manage the problem until the patient is able to see a doctor in person. They are also able to add further comments to the patient, or ask for other photos if the original is not clear enough.
If the solution proves to be a success, an enormous amount of data will be available which then can be used to teach an algorithm to perform the diagnostic work of dentists, or at least to give recommendations.
It could obviously be a long-term goal that is not part of the MVP, but our client thought it was important that the system included elements that will facilitate the teaching process for the algorithm. As a result, we developed a teaching module for creating a visual diagnosis.
In this module, problematic parts of the patient’s teeth can be marked on the pictures, and specific dental attributes, such as diagnostics report, area, diagnostic reasoning can be added. With a special image processing algorithm, the system will be able to recognize patterns after entering a substantial amount of similar data, and then will be able to recommend categorization.