NeoScan_Radiologist/.cursor/rules/appflow.mdc

193 lines
5.2 KiB
Plaintext

---
alwaysApply: true
---
Radiologist App Flow Guide
1. Onboarding Flow
There are two ways for a radiologist to join the platform:
A. Self-Registration (via App)
Start App Signup → Radiologist enters email address
Email Check
If email already exists → Show message "Already Registered" → Redirect to Login
If new email → Continue to next step
Set Password → Create secure password
Enter Personal Info → Complete profile information
Upload Hospital ID → for verification
Select Hospital/Institution → Choose affiliated organization → Wait for admin approval
Note: Radiologist registration requires additional verification of Uploaded hospital-ID. Status remains 'Inactive' until admin approval.
After Review → Login → Reset Password → Radiologist Dashboard
B. Admin-Created Radiologist
Admin creates radiologist account in the system
Login credentials sent to radiologist's registered email address
Radiologist logs in using received credentials
Upload Hospital ID → Upload the ID-Card issued by the hospital
Reset Password → Access Radiologist Dashboard
2. Dashboard Overview
When the radiologist logs in, they access a specialized Dashboard displaying:
Cases Prediction Review (AI predictions over all review)
Total Cases Reviewed (lifetime review count like feedback given by radiologist)
Feedback Statistics (Agreed/Disagreed with AI predictions)
Priority Cases (High-severity cases requiring immediate attention)
3. Patient List & Review Queue
Accessible from the Patient List Tab and Review Queue Tab.
Patient List View
Displays patient information in a structured table/card layout with the following details:
Patient ID & Demographics
Clinical History (if available)
Scan Details (date, type, protocol)
Previous Radiology Reports (if any)
Modality
Report Status (available/not available)
Series Information
Filtering Options:
All Cases
Processed (Data already proceessed by AI)
Pending (Data need to proceesss)
4. Case Review Interface
Clicking a case opens the Case Review Interface with:
Patient Information Panel:
Patient ID & Demographics
Clinical History (if available)
Scan Details (date, type, protocol)
Previous Radiology Reports (if any)
AI Analysis Panel:
AI Predictions Summary
Confidence Scores for each finding
Severity Classifications
Location Mappings
Processing Time and Algorithm Version
DICOM Viewer Panel:
Full-screen DICOM image viewer with professional tools
Advanced viewing controls:
Zoom functionality
Frame navigation with slide bar control
Local DICOM Upload - Option to choose and load DICOM files from local system
Feedback Panel:
Overall Assessment
✅ Agree with AI - AI prediction is accurate
🔶 Partially Agree - Some findings correct, others missed/incorrect
❌ Disagree with AI - AI prediction is inaccurate
Detailed Feedback by Finding
True Positive ✓ (AI correctly identified finding)
False Positive ✗ (AI incorrectly identified finding)
False Negative ✗ (AI missed actual finding)
Additional Findings (Findings not detected by AI)
5. AI Predictions Analysis & DICOM Viewer
Series Navigation:
Each Series contains multiple images
Thumbnail navigation strip
Scroll through series (e.g., Image 15 of 120)
Detailed AI Analysis:
Hemorrhage Detection
Epidural
Subdural
Subarachnoid
Intraparenchymal
Intraventricular
Stroke Analysis
Midline shift measurement
Confidence Scoring:
Overall Confidence: 85.6%
Per-finding Confidence:
Subdural hematoma: 94.2%
Midline shift: 78.9%
No acute stroke: 91.5%
6. Completed Reviews & History
Review History Tab:
List of all previously reviewed cases
Review statistics and performance metrics
Case Features:
AI System Performance (Radiologist Perspective):
Algorithm Accuracy based on radiologist feedback
Common AI Errors and patterns
Suggested Improvements submission portal
8. Profile & Account Management
Profile Card:
Display Name
Role (Radiologist)
Email Address
Profile Picture (Optional)
Account Management Options:
Edit Profile:
Update personal details
Update dispaly name
Change Password:
Reset login password
Navigation Flow Summary
App Launch
Registration/Login
Radiologist Dashboard (Main Hub)
├── Review Queue
│ └── Case Review Interface
│ ├── Patient Info
│ ├── AI Analysis
│ ├── DICOM Viewer
│ └── Feedback Panel
├── Completed Reviews
│ └── Review History & Analytics
├── Performance Dashboard
│ └── Personal & AI Metrics
├── Profile Management
│ ├── Edit Profile
│ └── Change Password
└── Logout
Key Features
Comprehensive Case Review with advanced DICOM viewing capabilities
Detailed AI Feedback System for continuous algorithm improvement
Performance Analytics for personal and system-wide metrics
Professional-grade Tools for accurate radiological assessment
Streamlined Workflow optimized for radiology practice
Quality Assurance features for maintaining diagnostic standards
Educational Components for continuous learning and improvement
This flow guide ensures radiologists can efficiently review AI predictions, provide meaningful feedback, and maintain the highest standards of diagnostic accuracy while contributing to AI system improvement.