Objectives

Because of the large amount of medical imaging data, the transmission process becomes complicated in telemedicine applications. Thus, in order to adapt the data bit streams to the constraints related to the limitation of the bandwidths a reduction of the size of the data by compression of the images is essential. Despite the improvements in the field of compression, the transmission itself can also introduce errors. For this reason, it is important to develop an adequate strategy which will help reduce this volume of data without having to introduce some distortion and resist the errors introduced by the channel noise during transmission. Thus, in this paper, we propose a ROI-based coding strategy and unequal bit stream protection to meet this dual constraint.

Material and methods

The proposed ROI-based compression strategy with unequal bit stream protection is composed of three parts: the first one allows the extraction of the ROI region, the second one consists of a ROI-based coding and the third one allows an unequal protection of the ROI bit stream.

First, the Regions Of Interest (ROI) are extracted by hierarchical segmentation of these regions according to a segmentation method based on the technique of Marker-based-watershed combined with the technique of active contours by level set. The resulting regions are selectively encoded by a 3D coder based on a shape adaptive discrete wavelet transform 3D-BISK, where the compression ratio of each region depends on its relevance in diagnosis. These obtained regions of interest are protected with an error-correcting code of Reed-Solomon type with a code rate that varies according to the relevance of the region by an unequal protection strategy (UEP).

Results

The performance of the proposed compression scheme is evaluated in several ways. First, tests are performed to study the impact of errors on the different bit streams. In the first place, these tests are carried out in order to study the effect of the variation of the compression rates on the different bit streams. Secondly, different Reed Solomon error-correcting codes of different code rates are tested at different compression rates on a BSC channel. Finally, the performances of this coding strategy are compared with those of SPIHT 3D in the case of transmission on a BSC channel.

Conclusion

The obtained results show that the proposed method is quite efficient in transmission time reduction. Therefore, our proposed scheme will reduce the volume of data without having to introduce some distortion and resist the errors introduced by the channel noise in the case of telemedicine.

中文翻译：

---

用于无线通信的3D MRI脑数据集基于ROI的压缩策略

目标

由于大量医学成像数据，因此在远程医疗应用中传输过程变得复杂。因此，为了使数据比特流适应与带宽限制有关的约束，通过图像压缩来减小数据大小是必不可少的。尽管在压缩领域有所改进，但变速器本身也会引入误差。因此，重要的是要开发一种适当的策略，该策略将有助于减少此数据量，而不必引入某些失真并抵抗传输过程中由通道噪声引入的错误。因此，在本文中，我们提出了一种基于ROI的编码策略和不相等的比特流保护来满足这种双重约束。

材料与方法

所提出的具有不等比特流保护的基于ROI的压缩策略由三部分组成：第一部分允许提取ROI区域，第二部分包括基于ROI的编码，第三部分允许对ROI进行不平等的保护位流。

首先，根据基于标记的分水岭技术与基于级别集的活动轮廓技术相结合的分割方法，通过对这些区域进行分层分割来提取感兴趣区域（ROI）。通过形状自适应离散小波变换3D-BISK，由3D编码器选择性地对生成的区域进行编码，其中每个区域的压缩率取决于其在诊断中的相关性。通过不等保护策略（UEP），利用Reed-Solomon类型的纠错码来保护这些获得的关注区域，该纠错码的编码率根据该区域的相关性而变化。

结果

所提出的压缩方案的性能可以通过几种方式进行评估。首先，进行测试以研究错误对不同比特流的影响。首先，进行这些测试是为了研究压缩率变化对不同比特流的影响。其次，在BSC信道上以不同的压缩率测试不同码率的不同里德所罗门纠错码。最后，在BSC信道上传输的情况下，将这种编码策略的性能与SPIHT 3D的性能进行比较。

结论

所得结果表明，该方法在减少传输时间方面是非常有效的。因此，在远程医疗的情况下，我们提出的方案将减少数据量，而不必引入一些失真，并抵抗由信道噪声引入的错误。