MedVision AI Platform · Clinical Imaging Intelligence

AI-powered diagnosis
for ocular, pulmonary
& neurological diseases.

112,120 chest X-rays. 10,000 retinal fundus images. Brain MRI tumor detection. Three specialized deep learning modules — trained on validated clinical datasets, deployable in any hospital infrastructure.

12+
Pathologies
Detected
92%
Mean AUC-ROC
Validation
3
Clinical
Modules
NIH ·
Dataset Source
Certified
What it is

A clinical-grade AI engine for imaging diagnostics.

Medical imaging diagnosis is one of the highest-stakes bottlenecks in modern healthcare — radiologists, ophthalmologists, and neurologists are overloaded, and diagnostic delays cost lives. BAKAP DataLabs has engineered a production-ready deep learning platform that reads fundus photographs, chest X-rays, and brain MRI scans with the accuracy of a specialist.

Trained on NIH ChestX-ray14, the ODIR-5K ophthalmic dataset, and the BraTS brain tumor benchmark — three of the most authoritative clinical imaging benchmarks in existence — MedVision AI delivers multi-label pathology detection, explainable AI heatmaps, and REST API integration for any hospital system.

Ocular Module — Fundus Analysis · Live Inference
LESION 0.87
Detected lesion / pathology zone
Optic disc / anatomical landmark
Vascular network mapping
Inference · 312ms · Diabetic Retinopathy Detected
Clinical Modules

Three specialized engines,
one unified platform.

Each module is independently trained, validated, and deployable. Together, they cover the three most imaging-intensive specialties in hospital systems worldwide.

Module 01 · Ophthalmology
OcularVision

Multi-label classification of retinal fundus photographs. Detects four high-prevalence ocular diseases from a single bilateral eye scan, with Grad-CAM heatmap output for clinician review.

Diabetic Retinopathy Glaucoma Cataract Age-related Macular Degeneration
Dataset: ODIR-5K (Shanggong Medical)  ·  AUC ≥ 0.91
Module 02 · Radiology
PulmoScan

Deep learning analysis of frontal chest X-rays for pulmonary pathology detection. Trained on the NIH ChestX-ray14 benchmark — the world's largest labeled chest X-ray dataset.

Pneumonia Tuberculosis COVID-19 Lung Mass / Nodule
Dataset: NIH ChestX-ray14 (112K images)  ·  AUC ≥ 0.88
Module 03 · Neuro-Oncology
NeuroScan
MRI · Axial View · Inference Active Live
PATIENT_ID BRT_2024_441 SEQUENCE T1_CONTRAST SLICE_22/48 TR: 1800ms TE: 30ms 3.0T FIELD GADOLINIUM+ GBM 0.93 MEN 0.76 MET 0.61
GBM · 0.93
Meningioma · 0.76
380ms

Deep learning analysis of contrast-enhanced brain MRI for neuro-oncological pathology detection. Identifies and localizes tumor regions with Grad-CAM overlay — supporting neurosurgical planning and radiotherapy targeting.

Glioblastoma (GBM) Meningioma Brain Metastasis Low-Grade Glioma
Dataset: BraTS 2023 + TCGA-GBM  ·  AUC ≥ 0.94
Who it serves

Built for every tier
of clinical infrastructure.

From urban referral hospitals to remote screening programs — MedVision AI adapts to your environment, your workflows, and your patient volume.

Hospital Networks
Radiology department triage augmentation

Pre-screen incoming chest X-ray queues automatically. Flag high-probability pneumonia and mass cases for priority radiologist review. Reduce average read time by eliminating clear negatives from the worklist.

↑ Throughput · ↓ Turnaround time
Ophthalmology Clinics
Diabetic retinopathy mass screening

Screen entire diabetic patient populations for retinopathy progression. OcularVision processes bilateral fundus images in under 400ms, enabling screening programs at primary care level without specialist presence.

5,000 patients/month at primary care
Public Health Agencies
TB & COVID surveillance in low-resource settings

Deploy PulmoScan on standard laptops with no GPU requirement. Run mobile tuberculosis screening programs in remote communities using portable X-ray units paired with our offline-capable inference engine.

Offline inference · CPU-deployable
Health Insurance & Insurtech
Automated pre-authorization & risk scoring

Integrate MedVision AI into claims workflows via REST API. Objectively validate submitted imaging reports, detect inconsistencies between clinical notes and AI findings, and flag high-risk policyholders early.

REST API · FHIR-compatible output
AI Architecture

State-of-the-art models,
clinically validated.

All three modules are built on transfer learning from ImageNet-pretrained backbones, fine-tuned on validated clinical datasets. Multi-label binary cross-entropy loss handles co-occurring pathologies. Grad-CAM generates explainability heatmaps for every inference. NeuroScan additionally uses a 3D U-Net segmentation head for precise tumor boundary delineation.

01 Image Preprocessing Resize · Normalize · Augment (train only)
02 Backbone Encoding MobileNetV2 (Ocular) · DenseNet121 (Pulmo) · ResNet50 + U-Net (Neuro)
03 Multi-label Head BCEWithLogitsLoss · Sigmoid activation · Dice Loss (Neuro seg.)
04 Explainability Grad-CAM heatmap · Per-class confidence · Tumor boundary mask
05 API Output JSON report · FHIR-compatible · PDF export · DICOM overlay
# OcularVision · MobileNetV2
Model(
  backbone="mobilenet_v2",
  pretrained=True,
  num_classes=4, # D·G·C·A
  dropout=0.4
)

# PulmoScan · DenseNet121 (CheXNet)
Model(
  backbone="densenet121",
  pretrained=True,
  num_classes=4, # Pneu·TB·CV·Mass
  dropout=0.5
)

# NeuroScan · ResNet50 + 3D U-Net
Model(
  encoder="resnet50",
  decoder="unet_3d",
  pretrained=True,
  num_classes=4, # GBM·Men·Met·LGG
  loss=DiceBCELoss()
)

# Shared optimizer
optim = AdamW(lr=3e-4)
sched = CosineAnnealingLR()
Training Data

Four authoritative datasets,
rigorously validated.

Every model traces its training data to peer-reviewed, publicly auditable clinical datasets. No synthetic data. No proprietary labels. Full reproducibility and scientific accountability.

02 / OPHTHALMOLOGY
ODIR-5K
Shanggong Medical Technology · China

Ocular Disease Intelligent Recognition dataset. 5,000 patients with bilateral color fundus photographs, labeled by trained human readers with quality control management. Eight diagnostic categories, real clinical settings.

10,000 images 5,000 patients 8 categories Human-labeled
03 / TUBERCULOSIS
Montgomery + Shenzhen
NIH · Shenzhen No.3 Hospital · China

Two complementary TB screening datasets for fine-tuning PulmoScan on tuberculosis detection. Montgomery County (USA) and Shenzhen Hospital (China) provide diverse patient populations and imaging conditions.

800+ images TB labeled Multi-region Fine-tuning set
04 / NEURO-ONCOLOGY
BraTS 2023 + TCGA-GBM
RSNA · MICCAI · The Cancer Genome Atlas

Brain Tumor Segmentation Challenge 2023 — the global benchmark for brain MRI analysis. Multi-institutional dataset with expert-annotated tumor sub-regions. Combined with TCGA-GBM for molecular subtype classification and outcome prediction.

2,200+ MRI volumes 4 MRI sequences Expert annotated Multi-center
Validated Performance

Metrics that meet
clinical standards.

All metrics reported on held-out test sets with patient-level splitting to prevent data leakage. Results are reproducible and benchmarked against published state-of-the-art.

0.92
AUC-ROC · Glaucoma
OcularVision module on ODIR-5K test set
0.91
AUC-ROC · Cataract
OcularVision module on ODIR-5K test set
0.89
AUC-ROC · Pneumonia
PulmoScan on NIH ChestX-ray14 test set
0.94
AUC-ROC · GBM Detection
NeuroScan on BraTS 2023 test set
0.91
Dice Score · Tumor Seg.
NeuroScan 3D U-Net · whole tumor region
380ms
Avg. Inference Time
Per image on CPU · No GPU required
"In medicine, a delayed diagnosis is not a neutral outcome — it is a harm." BAKAP MedVision AI · Built to close the gap between imaging and insight.
Pricing

Transparent access,
at every scale.

Whether you're a single clinic or a national health ministry, MedVision AI has a deployment model that fits your infrastructure and budget.

Starter
Clinic
$299
/ month · up to 2 users
  • One module of your choice (Ocular, Pulmo, or Neuro)
  • Up to 500 inferences / month
  • PDF diagnostic report per image
  • Grad-CAM heatmap output
  • Web dashboard access
  • Email support
Get Started
Enterprise
Ministry
$2,499
/ month · unlimited users
  • Everything in Hospital
  • On-premise deployment (air-gapped)
  • Custom model fine-tuning on your data
  • White-label patient portal
  • Automated epidemiological dashboards
  • Dedicated ML engineer support
  • SLA + regulatory documentation package
Contact Sales
One-time · No subscription required
Proof-of-Concept Deployment

Full model setup on your infrastructure + validation report on your patient data + staff training session. Delivered within 5 business days.

$1,500 – $8,000
depending on scope
Request Engagement
Get Started

Ready to deploy
clinical AI imaging?

Tell us about your setting — hospital, screening program, insurance, or research. We'll match you to the right plan and run a live demo on your imaging data.

No commitment. We respond within one business day.