Blog

Radiology workflow. AI annotation. Real numbers.

Notes from the Neurmorph team: pulmonary nodule detection, DICOM infrastructure, Fleischner guidelines, radiologist workflow, and the practical realities of deploying annotation AI in a working radiology department.

Empty radiology workstation with multiple monitors representing workflow challenges
Workflow
Radiologist Burnout Starts with Mechanical Tasks. Here's What Automation Actually Helps.
The problem isn't volume alone — it's the proportion of each shift spent on annotation mechanics rather than diagnosis.
Jan 14, 2025
 Abstract visualization of medical scan analysis and measurement precision
AI & Evidence
What "94% Accuracy" Means in Pulmonary Nodule Detection — and What It Doesn't
Sensitivity, specificity, and AUC — what those numbers mean for a radiologist reading 80 chest CTs a day, and what they don't.
Feb 11, 2025
 Hospital server room with network infrastructure representing PACS integration
Integration
Integrating an AI Tool into Your PACS Without a 6-Month IT Project
DICOM DIMSE, DICOMweb, and what actually determines deployment time in a real hospital network.
Mar 18, 2025
 Abstract timeline visualization representing workflow timing and bottlenecks
Operations
The Three Places Where Radiology Turnaround Time Actually Breaks Down
TAT optimization usually focuses on the wrong stages. Where time actually disappears in a typical department.
Apr 22, 2025
 CT scanner in a hospital imaging department representing growing scan volumes
Industry
CT Chest Screening Volume Is Growing 12% Year-Over-Year. Radiologist Headcount Isn't.
The capacity gap that's driving interest in AI annotation — and why it won't close by hiring alone.
May 20, 2025
 Abstract dual-view medical imaging concept representing second read analysis
AI & Evidence
AI as a Second Read: What the Evidence Actually Shows
Prospective studies on AI second-read in chest CT and mammography — what improved, what didn't, and the methodological caveats procurement teams overlook.
Jun 24, 2025
 Abstract data metrics visualization representing medical workflow measurement
Operations
How Do You Measure Radiologist Productivity Without Penalizing Thoroughness?
RVU-based metrics reward speed and penalize thoroughness. How radiology operations teams are rethinking productivity measurement without sacrificing read quality.
Aug 5, 2025
 Abstract regulatory document and approval concept in clinical context
Regulatory
510(k) Clearance and AI Radiology: What Buyers Need to Know
510(k) clearance tells you a tool is substantially equivalent to a predicate — not that it's accurate on your scanner population. What radiology buyers should ask instead.
Sep 16, 2025
 Abstract workflow diagram representing incidental findings tracking in radiology
Workflow
Incidental Findings Are Everywhere. Most Departments Don't Have a Workflow for Them.
An incidental pulmonary nodule flagged in a chest CT ordered for a different reason — what happens to it without a tracking workflow? In most departments: not much.
Oct 28, 2025
 Abstract checklist and verification concept in clinical environment
Implementation
Before You Pilot a Radiology AI Tool: An IT and Clinical Checklist
DICOM connectivity, BAA requirements, on-premise vs hosted, false-negative accountability — a checklist for radiology IT and clinical leads before agreeing to a pilot.
Dec 2, 2025