AI-Based Lung Diagnosis from CT Scans

An artificial intelligence-based service that identifies the affected lung tissue, assesses the degree of damage on the CT scale and detects cancer at an early-stage

How does it work?

1. The doctor uploads CT scans of the patient’s lungs to the service.


2. The artificial intelligence system highlights the damaged areas, calculates their total percent, and determines whether virus pneumonia is present, including that originating from COVID-19. The neural network assesses the lungs hazard level, from CT1 to CT4.


If cancer is discovered, the system finds even minor tumors in the lungs and highlights all potentially damaged areas to be reviewed by the doctor.


3. Based on the AI opinion, the doctor issues the diagnosis to the patient.



CT Lungs - fast and accurate identification of lung diseases

How the service helps doctors

Automatically determines the hazard level and percent of the lungs damaged.

Minimizes all chances to overlook early-stage tumors. When the AI system detects a tumor exceeding the threshold of 4 mm, it will highlight all tumors regardless of their size. Radiologists see highlighted areas and decide on further diagnostic actions.

The algorithm highlights the damaged areas across all dissections scanned and uploads a relevant opinion draft to decrease the doctor’s workload and relieve from routine tasks.

Handles data most accurately to minimize diagnosing errors.

Competitive advantages

The service was trained on one of the largest sample bases for similar AI-based models

Has the highest accuracy among similar solutions in Russia - 94%

Can be adapted to customer requirements