Reversible Watermarking Method with Low Distortion for the Secure Transmission of Medical Images

In telemedicine,the realization of reversible watermarking through information security is an emerging research field.However,adding watermarks hinders the distribution of pixels in the cover image because it creates distortions(which lead to an increase in the detection probability).In this article,we introduce a reversible watermarking method that can transmit medical images with minimal distortion and high security.The proposed method selects two adjacent gray pixels whose least significant bit(LSB)is different from the relevant message bit and then calculates the distortion degree.We use the LSB pairing method to embed the secret matrix of patient record into the cover image and exchange pixel values.Experimental results show that the designed method is robust to different attacks and has a high PSNR(peak signal-to-noise ratio)value.The MRI image quality and imperceptibility are verified by embedding a secret matrix of up to 262,688 bits to achieve an average PSNR of 51.657 dB.In addition,the proposed algorithm is tested against the latest technology on standard images,and it is found that the average PSNR of our proposed reversible watermarking technology is higher(i.e.,51.71 dB).Numerical results show that the algorithm can be extended to normal images and medical images.

Impacts of distorted fishery statistical data on assessments of three surplus production models

We evaluated the effect of various error sources in fishery harvest/effort data on the maximum sustainable yield （MSY） and corresponding fishing effort （EMsv） using Monte Carlo simulation analyses. A high coefficient of variation （CV） of the catch and effort values biased the estimates of MSY and EMsv. Thus, the state of the fisheries resource and its exploitation was overestimated. We compared the effect using three surplus production models, Hilborn-Waters （H-W）, Schnute, and Prager models. The estimates generated using the H-W model were significantly affected by the CV. The Schnute model was least affected by errors in the underlying data. The CVof the catch data had a greater impact on the assessment than the CV of the fishing effort. Similarly, the changes in CV had a greater impact on the estimated maximum sustainable yield （MSY） than on the corresponding estimate of fishing effort （EMsY）. We discuss the likely effect of these biases on management efforts and provide suggestions for the improvement of fishery evaluations.