NeuroImager
NeuroImager: AI Clinical Decision Support Software for Medical Imaging
Executive Overview
NeuroImager is an enterprise Clinical Decision Support System (CDSS, a system that assists physicians in decisions) from AI-Systems. It uses computer vision and natural language processing to detect pathologies and other anomalies on digital medical images, and alerts the doctor when preset critical values are exceeded. It integrates into existing hospital IT without changing standard diagnostic workflows. Results are shown as an annotated DICOM series (DICOM, the standard format of medical images) with a draft structured report for the radiologist to verify and sign. It supports clinicians and does not replace their judgment.
Neural & Technological Stack
Computer vision (CV) models that parse 3D DICOM series slice by slice: localization of pathology and calculation of the percentage of affected tissue
Classification of pathological changes in accordance with medical standards, such as lung involvement severity (CT 0–4) and mammography BI-RADS
Synchronization of processed results with the original images for clearer analysis by specialists
Intelligent worklist triage: studies with critical findings move to the top of the radiologist's queue
Voice-to-text (NLP) dictation for medical terminology, auto-completing DICOM SR protocols
Processing time: a full 3D series in under 3 minutes
Input: multi-frame DICOM 3.0 (CT, LDCT, MRI, MG, DX, CR, RG, FLG)
Output: annotated DICOM series, DICOM SR, Secondary Capture, PDF export, REST/JSON
Integration: PACS, VNA, RIS, HIS/EMR and regional radiology and health information services via DICOM and REST API
Clinical Coverage
Pulmonology (CT, low-dose CT, X-ray, fluorography): viral pneumonia, tuberculosis, lung consolidation, emphysema
Cardiology (CT, angiography): coronary calcification, aortic aneurysm, paracardial fat volume
Oncology (CT, mammography, MRI): lung nodules, breast screening, adrenal, liver, brain and prostate findings
Neurology (brain and spine MRI): demyelinating plaque tracking, spine and disc analysis
Emergency and trauma (X-ray, CT): stroke localization, fracture detection
Reported model metrics (developer data): sensitivity 0.9945, specificity 0.9945, area under the ROC curve (AUC) 0.96
Expected outcomes for healthcare organizations: fewer diagnostic oversights and faster diagnosis
Target Industry Verticals
Hospitals, general medical and surgical hospitals, diagnostic imaging centers.
Deep-Dive ReferenceExplore the full architectural whitepaper and engineering case studies in My Projects: https://www.loranjacobs.com/my-projects/smart-clinic-cdss-neuroimager
Deployment & Licensing Matrix
Product tiers: Enterprise, Enterprise Ultra, Enterprise Ultima, Infrastructure (critical information infrastructure)
Platform: Smart Clinic
Delivery model: standalone AI software product
Licence: enterprise software licence, out-licensing with milestone-based payments
Your input: DICOM studies via PACS and integration with your hospital information system
Our scope: deployment, integration, model support
Installation: on-premises air-gapped GPU servers (NVIDIA) or a hybrid edge gateway
Maturity: TRL 9 (proven in operation)
Price: see the selected tier, excluding sales tax
