Healthcare / Radiology

AI-Powered X-Ray Patient Positioning System

An intelligent Raspberry Pi device that automatically detects and tracks patient body parts inside an X-ray machine, then controls motors to align the collimator in real time — replacing manual calibration with AI precision.

$0

Cloud dependency

+40%

Positioning accuracy

3x

Faster calibration

The Challenge

Radiology technicians were manually positioning X-ray collimators for every patient — a repetitive, error-prone process that slowed imaging workflows. Misalignment meant retakes, higher radiation exposure, and wasted time. The client needed an automated solution that could run standalone on low-cost hardware.


Our Approach

We built a compact embedded system on a Raspberry Pi that uses a camera module to detect patient anatomy in real time. An AI object detection model identifies body parts, while a control loop drives stepper motors to align the collimator. The entire stack runs on-device — no cloud, no latency.

The Results

  • Real-time body part detection with 95%+ confidence

  • Automated motor-driven collimator alignment

  • Fully standalone — no internet required

  • Reduced technician workload by 60%

  • Compact, cost-effective hardware footprint


System Architecture

Camera Module

AI Detection Engine

Tracking Algorithm

Collimator

Motor Controller


Embedded Linux

Python

Motor Control

Sensor Fusion

Computer Vision

Object Detection

 Raspberry Pi