3D Comparator
Company
Jessenius Faculty of Medicine in Martin, Comenius University
Project Duration
6 months
Implemented System
Evaluation system for talent exams
Area
Education
Our customer
The Jessenius Faculty of Medicine is one of the three medical faculties of public universities in the Slovak Republic. It was established as an independent faculty in 1969 and is an integral part of Comenius University. Each year, the admissions committee must select only 11 of the 100 to 200 applicants seeking to study dental medicine at the faculty in Martin. Due to human limitations and subjective evaluation, we at BRAIN:IT, in cooperation with the Jessenius Faculty of Medicine in Martin of Comenius University, decided to create an objective evaluation system for talent exams for dental technician applicants.
Our challenge and goals
Imagine finally receiving your new zirconium tooth — more precisely, a dental crown — crafted by a dental technician. But how would you feel if, after a painful fitting, it started to irritate you? Or imagine eagerly looking into the mirror only to see an asymmetrically shaped tooth, turning your new smile into a nightmare. Unfortunately, this is today’s reality. We have excellent dental technicians who excel in theoretical knowledge, yet their fine motor skills and talent needed to sculpt a perfect, natural-looking tooth are not always sufficient.
At the Jessenius Faculty of Medicine in Martin of Comenius University, leading dental experts decided that enough was enough. They chose to introduce motor skill talent exams as part of the admissions process.
But what should such exams look like? The dental experts decided to test all year groups of the program. Students were given the task of sculpting a chess piece from a plaster block using dental tools within a limited amount of time. The results were alarming. Many students performed very poorly, and their skills improved only slowly as they progressed through their studies.
For this reason, the faculty decided to introduce a talent exam that included carving a plaster figurine. However, they quickly encountered a major challenge. In some years, more than 200 applicants applied, yet only 11 could be admitted. How can such a large number of applicants be evaluated quickly and objectively? How can the process ensure impartiality, reproducibility of results, and fairness for all applicants in a given year — and across all future exams as well?
Speed
The evaluation of results and the digitalization of objects into a 3D model must be fast and time-efficient.
Objectivity
All objects must be evaluated under the same conditions, and the outputs must be reproducible to ensure objectivity.
Variability
The system must allow prioritization of the required model parameters and evaluation criteria using weights to ensure flexibility.
Our solution
At BRAIN:IT, we developed a generic tool called 3D Comparator, which objectively evaluates and ranks applicants each year based on their results in the dental technician talent exams. In doing so, we apply knowledge from statistics, fundamental object properties, symmetry analysis, as well as heatmaps and voxels.
Thanks to our research team, dental experts, and countless hours spent in dental laboratories, we created something far beyond a tool for the admissions committee. 3D Comparator is widely applicable for any comparison and evaluation of digitized 3D objects, with the ability to customize the reference model.
The exam
The examination process may sound simpler than it truly is, but it is extremely demanding. In addition to the well-known National Comparative Exams (NPS), applicants must also undergo the Manual skills test (MST), the so-called talent admission exam. The manual skills test assesses applicants’ fine motor skills. They must carve a replica of a given model (template), such as a chess piece, from a plaster block using provided dental modeling tools (modeling knives, Lecron instruments) within a set time limit of 180 minutes.
Process
At the end of the exam, anonymized figurines are collected from all applicants and sent to a dental laboratory. The first step is 3D scanning of the figurines using the same scanner used for producing dental crowns and implants. A dental technician carefully mounts each carved piece in a specialized holder and inserts it into the 3D scanner. The scanner then begins the digitalization process, creating a 3D model in STL format. This process takes several minutes, after which a unique identifier is assigned to each newly created 3D model.
Next, the models arrive at BRAIN:IT. We load the prepared 3D models into our 3D Comparator application together with the reference model used for the talent exam. The first task is alignment to the reference model, which we perform using the CPD (Coherent Point Drift) algorithm. CPD is based on statistical principles, specifically the maximum likelihood method.
If this algorithm interests you, look up Coherent Point Drift.
Basic characteristics such as height, width, depth, circumference, surface area, and volume are compared with the characteristics of the reference model, allowing us to identify the first deviations. This is followed by symmetry analysis, which is important for dental experts as it is closely linked to the subjective perception of aesthetics, and therefore must also have its place in our evaluation.
These are thermal heat maps. In our case, we used the SSCD (Spatial Structure Circular Descriptor) algorithm to generate them, and then we compare them against the reference model. If you want to learn more, look up Spatial Structure Circular Descriptor.
And finally, everything must be evaluated by comparing the results with the reference model. A weighted error is determined, from which the final percentage match is calculated. All data — especially the results — are clearly displayed in graphical form, and an Excel file is generated for further processing and exam evaluation.
Benefits for the customer
And how did we contribute to improving the manual skills of future dental technicians—and thus to all future patients?
Improved manual skills of students
Thanks to our solution, the first pilot year achieved approximately double the improvement in manual skills compared to previous years.
More effective training of dental technicians
Students from the pilot cohort were able to create two teeth instead of one in the same amount of time, demonstrating a faster development of their practical abilities.
Higher quality of future professionals
By increasing the accuracy and objectivity of applicant selection, the quality of dental technicians has improved, which directly translates into better healthcare outcomes.
Project complexity
Size
2/6
Time
4/6
Finances
2/6
Complexity
6/6
What did our customer say about the solution?
“Students show a significantly higher level of manual skill compared to previous years.”
– prof. MUDr. Mgr. Juraj Mokrý, PhD, Vice-Dean, Jessenius Faculty of Medicine in Martin, Comenius University
Are you interested in a similar solution?
We will be happy to help you create a system tailored to your needs – from design to implementation.