医学图像处理专业由于数据的隐私性，一般都是不公开的。我投了一篇论文到Medical Image Analysis，2个审稿人分别给出了一些意见，其中第二个审稿人没有其他什么意见，仅仅我要提供代码和数据。因为我们的实验数据来自其它医疗机构，对方不允许我们以任何方式将数据提供给审稿人，因为我提供了源代码，并说明了数据不能提供。结果第一个审稿人让我过了，第二个审稿人以“没有数据无法验证”为由把我拒了。由于审稿时间比较长，编辑又给了一次机会，让我说服第二个审稿人，下面是第2个审稿人的原话：
Authors have addressed the concerns in a detailed response, which I appreciate. However, some of the responses are not quite satisfactory. if the data is not available, not much can be done to verify the claims. Even if the data is available, without the ground truth corrections on segmentation, the training of models can not be done. So there is no way anyone can reasonable reproduce the results even on the original data as the model itself is not available. Since there is no principled theory behind the model trained on so few data points (other than use of LSTM), I am not convinced that it adds much to the knowledge base unless data and models are validated properly.
请问现在该怎么办？审稿人坚持要数据，数据方坚持不许提供数据，夹在中间很难受。 返回小木虫查看更多