Skip to content

Our Technology Stack

Core technologies powering our research and applications.

Internal Use

Clinical Workflow
From patient scan to treatment observation
Clinical workflow diagram

Clinical workflow from pre-treatment day (Day 0) to each treatment day (Day n)

State of the Art: Pose Estimation
Industry-wide progress in human pose estimation accuracy over the past decade

This chart represents the field-wide advancement in pose estimation technology based on published research benchmarks, not our specific lab results. We leverage these state-of-the-art methods in our clinical applications.

Field-Wide Progress in Pose Estimation (2014-2024)

This chart shows the advancement in human pose estimation accuracy from published research papers, not our specific lab results. We leverage these state-of-the-art methods in our applications.

80%85%90%95%100%201420162018202020222024YearAccuracy %81.3%
MPII Human Pose dataset (PCKh@0.5) - Based on published research benchmarks

Implementation Strategy

We prioritize open-source solutions with established benchmarks, balancing innovation with reliability.

EasyMocap

Multi-view motion capture system.

Motion Capture
Multi-view Geometry

Camera calibration techniques.

Camera Calibration
Motion Capture
Foundation for patient digital twins

Motion Capture Process

1. Data Capture

Multiple cameras capture patient positioning

2. Pose Estimation

AI estimates 3D pose from images

3. Model Fitting

Pose data fits body models to patient