Dentistry · Artificial Intelligence

Ehsan
Shirdel

DDS  ·  AI Developer  ·  Digital Dentist
Teaching machines to reveal what clinicians never knew they needed.
Scroll
01 / Biography
A clinical path,
turned computational.
From a college program in St. Petersburg to dental school in Moscow, his journey evolved into a research practice guided by one conviction: dentistry deserves intelligent tools created by those who understand its real clinical challenges.
Ehsan Shirdel — DDS, DMD \2 Portrait of Ehsan Shirdel

In 2021, after completing a general college course in St. Petersburg, Ehsan entered Sechenov First Moscow State Medical University to study dentistry — a path he would soon redefine on his own terms.

From his earliest years as a student, Ehsan refused to limit himself to coursework. He was drawn to ideas that could become real systems, especially at the point where clinical understanding meets advanced technology.

By his third year of dental school, he had already begun developing his own concepts for applying artificial intelligence to dental image analysis.

What first existed as research sketches gradually evolved through study, programming, experimentation, and the demanding process of training deep learning models into functional systems at the intersection of dentistry and artificial intelligence.

2020-2021
Saint Petersburg State Pediatric Medical University.
2021 → present
Sechenov First Moscow State Medical University.
2023-2024
First AI concepts in dental imaging drafted.
Today
Deployable AI models for clinical dentistry.
02 / Vision
Two disciplines, one practice.
The future of dentistry will not be defined by clinical skill alone. The integration of clinical knowledge, imaging, and AI can transform the speed, precision, and quality of every treatment decision.
The Clinician
human
insight
Human hand Digital hand
The Engineer
machine
reasoning

As a dentist, he understands the patient — the symptoms, the oral structures, the diagnostic needs, and the human complexity behind every clinical decision. He sees what an AI model cannot: the patient in the chair.

As an AI developer, he translates those same clinical challenges into data, models, and tools. He brings back to the operatory what computation can uniquely offer: pattern recognition at scale, consistency, and the discipline of measurement.

03 / Research Thesis
What is out of place?
While current AI approaches in dentistry have focused on what’s visible in the dental image, Ehsan turned to a fundamental question.
Can an AI system be developed to detect not only what is present in the image, but also what is absent from it?

This distinction matters. An AI system designed to reason about absence cannot rely on visible features alone. It must understand the spatial relationships within the oral cavity, the anatomy of the dental arch, the position and alignment of existing teeth, and the clinical logic that defines where each tooth should be.

It must reason from the data toward absence, recognizing something whose only proof is the shape of the space it left behind.

This shifts the role of AI in dentistry from pure recognition toward a kind of clinical and spatial reasoning. It allows the system to understand structure more deeply, analyze patterns, and infer the absence of an element from indirect evidence.

04 / AI Systems
From concept
to working models.
The AI systems he has developed include models for identifying edentulous spaces with precision, automating dental charting and tooth numbering, supporting initial patient triage, and recognizing dental arch types.
Heatmap visualization of jaw detection model
Attention heatmap over maxillary arch
Inferred positions of missing teeth highlighted
Tooth numbering with FDI bounding boxes
Detected and missing teeth visualization
ROC curves across tooth classes with AUC values
Where Dentistry Meets Deep Learning.
05 / Publications
Selected work.
A growing record of research at the intersection of dentistry and artificial intelligence. Full bibliography available via Google Scholar and ResearchGate.
— — — —

Transitioning from a single-model baseline to a multi-stage deep learning pipeline: a head-to-head evaluation for tooth numbering in intraoral photographs

First author · Published
— — — —

Deep learning applications for automated dental arch form diagnosis

First author · Published
— — — —

A novel AI system for preliminary triage support in single-tooth edentulous spaces using intraoral images

First author · Under review
— — — —

A novel AI system for edentulous localization

First author · Manuscript in preparation
/ The full and updated publication list is available through the Google Scholar and ResearchGate profiles in Section 07.
06 / Life
Beyond the lab.
Research and clinical practice are demanding; balance is its own discipline.
Ehsan on a golf course at sunrise
"Focus, patience, and precision — three things both a swing and a model demand in equal measure."

Alongside research and clinical work, Ehsan keeps an active life — one that protects the equilibrium between learning, building, and being.

Golf, for the quiet of concentration. Basketball, for energy and the rhythm of a team. And the everyday companionship of his dog — a source of calm, responsibility and steady joy that no AI has yet learned to replace.

Let's build the next layer.
He believes the future of AI in dentistry will be built through multi-center, international and interdisciplinary collaboration. He invites dentists, orthodontists, prosthodontists, implantologists, oral and maxillofacial radiologists, and AI researchers to join the work.
— Contact
Email Google Scholar ResearchGate