| name | quantum-medical-image-encoding |
| category | quantum-medicine |
| description | Quantum image encoding and compression methodology for medical imaging using Fourier-based methods. Reduces quantum gate requirements by factor of 4+ compared to existing approaches. Based on arXiv:2505.06471 |
| activation | quantum medical imaging, quantum image encoding, quantum compression, QIMP, Fourier quantum, medical image quantum |
| arxiv_id | 2505.06471 |
| authors | Taehee Ko, Inho Lee, Hyeong Won Yu |
| created | 2025-06-10 |
| tags | ["quantum-imaging","medical-imaging","quantum-compression","fourier-transform","gate-reduction"] |
Quantum Medical Image Encoding and Compression
Overview
This skill provides a methodology for encoding classical medical images into quantum circuits with significantly reduced gate complexity, using Fourier-based techniques. The approach reduces quantum gates by at least 4x compared to existing encoding methods, making large-scale quantum medical imaging computationally feasible on near-term quantum hardware.
Core Methodology
1. Fourier-Based Quantum Image Encoding
Traditional quantum image encoding methods require approximately 2× the number of gates as pixels in the image. This Fourier-based approach reduces gate requirements by:
- Decomposing images into frequency domain using Fourier transforms
- Encoding frequency coefficients instead of pixel values
- Exploiting sparsity in frequency representation of medical images
- Achieving at least 4× gate reduction compared to direct pixel encoding
2. Compression Techniques
Two complementary compression methods:
2.1 Frequency Domain Compression
- Discard high-frequency components below quality threshold
- Maintain diagnostic quality while reducing qubit requirements
- Achieves additional 2-4× compression with negligible quality loss
2.2 Pre-processing Optimization
- Apply classical pre-processing to reduce quantum circuit depth
- Optimize measurement strategy for medical features
- Reduce overall quantum resource requirements
3. Validation Framework
- Tested on 1024×1024 high-quality medical images
- Validated with BABA (Bilateral Axillo-Breast Approach) robotic thyroidectomy surgical images
- Quality metrics: PSNR, SSIM, gate count, circuit depth
- Demonstrated feasibility for large-scale medical imaging applications
Implementation Steps
Step 1: Image Pre-processing
import numpy as np
from scipy.fft import fft2, ifft2
def prepare_medical_image(image_path, threshold=0.01):
"""Load and prepare medical image for quantum encoding"""
img = load_and_normalize(image_path)
freq_domain = fft2(img)
freq_domain[np.(freq_domain) < threshold * np.(np.(freq_domain))] =
freq_domain
