a region of interest approach for medical image compression
DESCRIPTION
A Region of Interest Approach For Medical Image Compression. Salih Burak Gokturk Stanford University. OVERVIEW. Motivation Previous Work Comparison Study of Compression Schemes ROI based System Design Conclusion. Motivation. Medical images are huge.(300x512x512x2) - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/1.jpg)
A Region of Interest Approach For Medical Image Compression
Salih Burak Gokturk
Stanford University
![Page 2: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/2.jpg)
OVERVIEW
• Motivation
• Previous Work
• Comparison Study of Compression Schemes
• ROI based System Design
• Conclusion
![Page 3: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/3.jpg)
Motivation
• Medical images are huge.(300x512x512x2)
• High quality imaging is required in diagnostically important regions.
• ROI based approach is the only solution:– Lossless compression in ROI.– Very lossy compression in non-ROI.
![Page 4: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/4.jpg)
OVERVIEW
• Motivation
• Previous Work
• Comparison Study of Compression Schemes
• ROI based System Design
• Conclusion
![Page 5: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/5.jpg)
Previous Work• Lossless Compression Schemes (Takaya95,
Assche00)• DCT based Compression Schemes
(Vlaciu95) • PCA based Compression(Tao96)• Wavelet Transformation(2D and 3D)
(Baskurt93)• ROI based coding (Cosman 94,95)
![Page 6: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/6.jpg)
OVERVIEW
• Motivation
• Previous Work
• Comparison Study of Compression Schemes
• ROI based System Design
• Conclusion
![Page 7: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/7.jpg)
Lossless Compression• Entropy of images – 7.93bpp
• Predictive Coding – 5.9bpp
• Entropy of difference images – 5.76bpp
![Page 8: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/8.jpg)
DCT Compression (1)
![Page 9: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/9.jpg)
DCT Compression (2)
![Page 10: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/10.jpg)
DCT Compression (3)
Quantization Step Size 1 2 4 8 16 32 64 128 256 512 1024
MSE in dB -11.7 -5.7 0.34 6.26 11.9 17.1 21.8 25.7 29.3 32.6 35.9
Rate (without RLC) (bpp)
5.74 4.97 4.09 3.20 2.34 1.57 0.96 0.55 0.31 0.16 0.09
Rate (with RLC) (bpp)
8.04 7.09 5.87 4.51 3.15 1.95 1.07 0.55 0.28 0.14 0.07
![Page 11: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/11.jpg)
PCA Compression - Treat each image block as a vector
MSE ~ 30 dB
Rate ~ 0.54 bpp
![Page 12: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/12.jpg)
Blockwise Vector Quantization(1)
- A simpler decoder is required
![Page 13: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/13.jpg)
Blockwise Vector Quantization(2)
MSE ~ 38 dB MSE ~ 39 dB
![Page 14: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/14.jpg)
Motion Compensated Hybrid Coding (1)
- Lukas Kanade Tracker was used by 0.1 pixel accuracy
![Page 15: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/15.jpg)
Lukas-Kanade Tracker
![Page 16: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/16.jpg)
Motion Compensated Hybrid Coding (2)
- Entropy of the motion vector is 2.28 and 2.45 in x and y.- This brings 0.018 bpp.
MSE ~ 35 dB
![Page 17: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/17.jpg)
OVERVIEW
• Motivation
• Previous Work
• Comparison Study of Compression Schemes
• ROI based System Design
• Conclusion
![Page 18: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/18.jpg)
Segmentation- Thresholding to find the air- Gradient magnitude to extract the colon wall- Grassfire operation to find the ROI around the colon wall
![Page 19: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/19.jpg)
ROI Based System
![Page 20: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/20.jpg)
Experiment with 16 by 16 Blocks- The ratio of ROI ~ %12.2- Entropy of motion vector is 2.28 in x and 2.45 in y- The entropy of the error image is ~ 4.38- average RMS error 33.7 dB with lossless in ROI- Overall rate 0.552 bps
MSE ~ 33.7 dB
![Page 21: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/21.jpg)
Experiment with 8 by 8 Blocks- The ratio of ROI ~ %7.3- Entropy of motion vector is 1.82 in x and 1.96 in y- The entropy of the error image is ~ 4.31- average RMS error 30.3 dB with lossless in ROI- Overall rate 0.37 bps
MSE ~ 30.3 dBMSE ~ 33.7 dB
![Page 22: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/22.jpg)
OVERVIEW
• Motivation
• Previous Work
• Comparison Study of Compression Schemes
• ROI based System Design
• Conclusion
![Page 23: A Region of Interest Approach For Medical Image Compression](https://reader035.vdocuments.us/reader035/viewer/2022081513/56815911550346895dc63da6/html5/thumbnails/23.jpg)
Conclusion
• Effective System (compression rate of %2.3)• Accurate System (lossless in ROI)• Results of ROI based compression over performs
standard compression schemes.• Future work includes lossy compression in ROI.• Case study with the radiologist for determining
rate-diagnosis performance curve.