Medical Image Processing Using Optical Fourier Technique

描述（由申请人提供）：该提案的最终目标是开发一种低成本、坚固耐用的便携式全光学医学图像处理系统，用于乳腺癌诊断和乳房X线检查。从并行处理的角度来看，光学图像处理技术固有地快速。该实验室在过去的十年中参与了用于图像处理的全光学傅立叶和小波变换的创新技术。利用细菌视紫红质薄膜的光致二向色性，建立了一种自适应光学傅里叶处理系统，并在真实的实时降噪、边缘增强和带通滤波等方面得到了应用。应用这些强大的技术来增强医学图像是本项目的目标。三个乳房X线片的初步结果证明了可行性。通过使用一个旋钮（在光学系统中旋转分析仪），可以选择性地显示微钙化簇或周围的致密组织。该技术速度快，乳房X线照片的扫描速率仅受CCD相机的帧速率的限制。提出了一种使用电子光学接口的集成架构。最初将研究乳房X线照片和体模的模拟图像。无论是商业上可用的标准以及定制设计的对象将用于一个彻底的对比细节分析。 该系统将是通用的，使它可以接受模拟以及数字格式的医学图像。重点是优化实验安排，并与我们在伍斯特医学院的放射学组合作研究了大量的乳腺X线照片，并进行了相关的统计分析。细菌视紫红质膜在高达140摄氏度的温度下是稳定的，并且是环境友好的。由于实验中不涉及干扰，因此不需要隔振，甚至不需要相干光源。该系统已经设计了一个便携式和节能电池供电的固态激光器是现场部署。随着在小芯片上的半导体激光器的潜在新进展，有可能进一步减小尺寸和成本，从而产生小的方便的信号增强放大器。调查的目的不是为了与昂贵的数字图像处理技术竞争，而是为了开发一种农村地区可以使用的低成本便携式仪器，用于快速筛查大量的乳房X线照片。

DESCRIPTION (provided by applicant): The ultimate goal of the proposal is to develop a low cost rugged portable all optical medical image processing system for breast cancer diagnostics with mammograms. Optical image processing techniques are inherently fast in view of parallel processing. This laboratory is involved during the last ten years in innovating techniques of all optical Fourier and wavelet transforms for image processing. A self adaptive optical Fourier processing system, using photo induced dichroism in a Bacteriorhodopsin film was experimentally demonstrated with applications in noise reduction, edge enhancement and band pass filtering in real time. Application of these powerful techniques for enhancement of medical images is the object of this project. Preliminary results on three mammograms demonstrated the feasibility. By playing with one knob (rotating the analyzer in the optical system) either the micro calcification clusters or the surrounding dense tissue can be selectively displayed. The technique is fast and the scan rate of mammograms is only limited by the frame rate of the CCD camera. An integrated architecture is proposed using electronic optical interfaces. Analog images of mammograms and phantoms will be investigated initially. Both commercially available standard as well as custom designed objects will be used for a thorough contrast detail analysis. The system will be made versatile so that it can accept medical images in analog as well as digital format. The focus is on optimizing the experimental arrangement and study a large number of mammograms in collaboration with the Radiology group of our medicalzchool in Worcester and carry out relevant statistical analysis. Bacteriorhodopsin films are stable up to 140 degrees C and environmental friendly. As no interference is involved in the experiments, vibration isolation and even a coherent light source are not required. The system designed already with a portable and energy efficient battery operated solid state laser is field deployable. With potential new advances in semiconductor lasers on small chips, it may be possible to further decrease the size and cost resulting in a small handy signal enhancing magnifier. The purpose of the investigation is not to compete with the expensive digital techniques for image processing but to develop a low cost portable instrument accessible to rural areas for quick screening of large number of mammograms.