NeuroMed
by NeuroVision

Artificial intelligence in healthcare

An intelligent healthcare platform that improves the efficiency of clinical diagnostics and administration

Request a demo

Supporting clinicians

Empowering clinicians and improving the quality of care through innovation

Early cancer detection

Enabling early detection of breast and lung cancer to improve the chances of recovery

Disease risk prediction

Predicting disease risk 1–6 years ahead to enable proactive treatment.

Reducing mortality

Reducing mortality through healthcare digitalization and new technologies

Powered by technology

Challenges facing modern clinics

Administrative

  • Inefficient reception processes

    • Slow document processing
    • Manual completion of patient forms
    • Human error
    • Queues
  • Navigation and safety

    • Difficulty finding departments and buildings
    • Incidents in corridors and wards
    • Unauthorized access
    • Infection control
  • Electronic document management

    • Manual review of incoming and outgoing paper documents
    • Disconnected administrative and clinical information systems
  • Limited appointment time

    • Increasing workloads as patient volumes grow
    • Significant time spent collecting routine vital signs

Diagnostic

  • Heavy clinician workloads

    • Clinician overload
    • Declining quality of care
    • Professional and emotional burnout
  • Risk of diagnostic errors

    Human error and limited time per patient can lead to diagnostic mistakes.

  • Growing volumes of medical data

    Growing data volumes require effective processing and analytics tools.

  • Limited healthcare resources

    Specialist shortages delay access to essential care

The NeuroMed platform

The NeuroVision platform

Listed in the Russian Software Registry

№19972

Core

Artificial intelligence that unifies all modules and enables seamless interaction

MedFace

Contactless photoplethysmography measures:

  • Blood pressure
  • Heart rate
  • Emotional state
  • Body temperature
  • Blood oxygen saturation
  • etc.

Additional modules

  • Document recognition and AI OCR
  • Video analytics and monitoring systems
  • Recognition of faces, gender, age, and face coverings
  • Contactless payments

Pathology detection

Ready to detect 200 conditions

  • Lung and breast cancer detection
  • Disease risk assessment

Key capabilities of the NeuroMed platform

A physician working with a holographic model of a patient’s body

AI and computer vision integration

NeuroMed combines AI and computer vision to improve diagnostic accuracy

CT/MRI image
processing

NeuroMed analyzes CT and MRI data for more accurate diagnosis

Automated medical examinations

The system automates pre-shift and periodic medical examinations, improving efficiency

MedFace — contactless diagnostics

Versatile use cases and a wide range of benefits

Contactless measurements

Measurement of heart rate, blood pressure, temperature, and oxygen saturation without physical contact

Contactless measurement of a patient’s vital signs by a camera in a clinic waiting area

High speed

Faster routine diagnostic procedures and more time for patient interaction

A physician working with a patient

Broad applicability

From modernizing medical examinations to optimizing patient pathways

A medical examination using the platform

Coming soon

  • Stroke detection
  • Detection of alcohol and drug intoxication
  • Pupillometry
Expanding the capabilities of the MedFace platform

Lung cancer risk assessment from a single CT scan

Using a single chest CT scan, the NeuroMed model estimates the probability of lung cancer — without additional clinical data or manual annotation by a physician. It analyzes the entire scan volume and provides a calibrated 1–⁠6-year probability together with a risk category. Results are generated in seconds and can be processed in the background on a workstation.

What the model does

  • Detects lesions

    Measurement of heart rate, blood pressure, temperature, and oxygen saturation without physical contact

  • Calculates probability

    Provides a calibrated probability of lung cancer over a 1–⁠6-year horizon.

  • Assigns a risk category

    Classifies the patient as low, medium, or high risk to support follow-up planning.

PERFORMANCE METRICS

  • 0.83 malignancy classification accuracy (AUC)
  • 0.77 prediction of a future
    cancer diagnosis (AUC)
  • 1 CT scan
    needed for assessment
  • > 50K patients with outcomes in the
    training dataset

How the model works in clinical practice

The model distinguishes malignant lesions from normal and benign findings and identifies high-risk patients on earlier CT scans. It supports case triage and follow-up prioritization — assisting clinicians rather than replacing them.

  1. Case triage

    Prioritize scans that require a
    radiologist’s attention.

  2. Follow-up interval

    Determine repeat CT frequency: less often
    for low-risk patients and more often for high-risk patients.

  3. Second opinion

    A quantitative risk assessment to complement
    the physician’s report.

A radiologist examining a CT scan

Risk categories define follow-up protocols, while thresholds can be tailored to the clinic’s workload:

Low
routine interval
Medium
shorter follow-up interval
High
priority review and further testing

Model validation data

Validation used real-world chest CT scans collected from 2021 to 2024. Patients who underwent multiple CT scans over time were selected separately for prognostic evaluation. Physicians manually annotated all lesions into three categories.

16,111 annotated lesions

  • Malignant 4,299 (26.68%)
  • Benign 3,481 (21.61%)
  • Other (scars, calcifications) 8,331 (51.71%)
  • 2021–⁠2024 real-world data period
  • 5–⁠14× higher early CT risk among patients subsequently diagnosed with cancer

Validation was performed on real-world but retrospective data. Cancer patients undergo repeat CT scans more frequently, so some “future diagnoses” represent recurrence rather than newly detected cancer. The results should therefore be interpreted as identifying patients with a future diagnosis and supporting follow-up. Independent prediction of cancer in initially healthy lungs still requires confirmation.

Reduce reception processing time to 30 seconds

Traditional registration takes

up to 10 minutes per patient

AI OCR implementation

  • Automatic document recognition
  • Transfer of all data to the hospital information system
  • Preparation of all documents for signing

6× faster patient registration

AI document verification: Instant. Accurate. Borderless.

  • 16,000+ types of government-issued documents
  • 195+ countries and territories
  • 90+ supported
    languages
  • <1 sec for the complete document recognition and verification cycle

Platform benefits

Scalability

Recognition of any paper documents received by the clinic

Patient loyalty

Shorter queues and higher patient satisfaction

Efficiency and quality

Faster and more accurate data entry, fewer errors and faster workflows

Video analytics for clinics

  • Smart navigation

    The system helps patients navigate the clinic with ease

  • Access control and monitoring

    • Enhanced safety and security
    • Tracking staff and patient movement to optimize workflows
  • Rapid response

    • Automatic detection of unusual incidents in patient areas
    • Alerts when personal protective equipment is missing
Discuss implementation

Platform evolution

  1. Early detection of lung cancer on CT and breast cancer on mammograms

    • Report generation
    • Integration with hospital information systems
    • Smart reception desks
    • Cancer risk prediction 2‑5 years ahead
    • Full-text report generation with an LLM
    • Detection of up to 200 conditions

Outcomes

  • Greater staff efficiency through the automation of routine tasks

    Integration of all clinical diagnostic and administrative processes into a single system

  • Improved quality of patient care and diagnostics

    More accurate diagnostics and predictive AI to help prevent disease

  • Optimized clinical workflows, allowing physicians to focus on patients’ needs and treatment

    Adoption of medical AI innovations for patient consultations, clinical decision support, and clinic management

Future development

Still have questions? Ask us!

Contact us directly or submit an enquiry — we’ll respond within one business day

Not sure how to phrase your enquiry? Schedule a call, and our team will contact you at a convenient time

Schedule a call

The company is listed in the Russian Ministry of Digital Development’s registers of accredited IT companies and personal data operators.

Address: 12 Presnenskaya Embankment, Federation Tower, Moscow City Business Center, Moscow. The nearest metro stations are Delovoy Tsentr and Mezhdunarodnaya.