Lucentaa
Early Concept

We're designing a first pass
for every imaging study.

Lucentaa is a concept for an AI tool that would flag candidate findings on X-ray, CT, and MRI studies and move the urgent ones up the worklist. A radiologist would review every flag before anything reached a report. It is at an early stage, and we are seeking pilot partners among healthcare providers.

Early Concept
Seeking Pilot Partners
DICOM Compatibility Planned
Based in Lagos, Nigeria
Planned Design

Three stages between the scanner and the report.

This is the design we are working toward. Each stage has one job: studies come in, the models look at them, and a radiologist decides what stands. None of it is built yet.

01 / Intake (planned)

Imaging Intake

The plan is for studies to arrive by DICOM upload or from a hospital's PACS, with identifying details stripped and images normalized. The radiology team's routine would stay as it is.

PACS IntegrationDICOM ParsingAuto De-identification
02 / Analysis (planned)

AI Analysis Engine

The AI models we are aiming to build would read each X-ray, CT, or MRI series, mark regions that may matter, and attach a confidence score to every detection.

Multi-Modality ModelsRegion DetectionConfidence Scoring
03 / Review (planned)

Clinical Review Layer

In the intended design, flagged findings open in a viewer made for radiologists, with a prioritized worklist and overlays on the image. The radiologist confirms or dismisses each one, then signs off.

Prioritized WorklistAnnotated OverlaysAudit Trail
How It Would Work

From incoming scan to reviewable summary.

The idea: when a study comes in, the pipeline runs on it automatically, so candidate findings are already listed next to the images when a radiologist opens the file. The example below is a mockup with sample data, not real output.

Incoming Study (sample)New
CT_CHEST_0417.dcm

Modality: CT

Patient ID: ****3821

Series: 3 / Slices: 214

AI
AI Summary (sample)Processed
FindingPulmonary nodule, right upper lobe
Confidence94.2%
RegionRUL, segment 3
Recommended ReviewPriority, needs radiologist sign-off
Design Goals

The questions imaging departments ask first.

Will it work with our systems? Where does the patient data go? Can we try it without risk? These are the questions we are designing around.

Interoperability

DICOM Compatibility

The plan is to read standard DICOM studies from X-ray, CT, and MRI, with PACS integration so a team can keep working the way it does now.

Data Protection

Patient Privacy

Data protection practices are being designed in line with applicable regulations, including Nigeria's NDPR, as the platform moves toward pilot testing.

Adoption

Pilot Program

We are seeking pilot partners among healthcare providers. A pilot would run on your own studies, with success criteria agreed up front.

Validation

Planned Validation

Nothing has been clinically tested yet. We plan to work directly with radiologists during pilot testing to validate the platform's outputs.

Interested in being a pilot partner?

We are seeking pilot partners among healthcare providers. Tell us about your imaging department and we'll take it from there.