There is a necessity to protect the confidential medical images data from an unauthorized access when the exchange of medical information is taken place among the patients and doctors. We can generate a common secret key using neural cryptography, which is based on synchronization of Tree Parity Machines (TPMs) by mutual learning. In the proposed TPMs replacing random inputs with queries are considered, which depend on the current state of the neural network. Then, TPMs hidden layer of each output vectors are compared. That is, the output vectors of hidden unit using Hebbian learning rule, leftdynamic hidden unit using Random walk learning rule and right-dynamic hidden unit using Anti-Hebbian learning rule are compared. Among the compared values, one of the best values is received by the output layer. The medical image is encrypted using Rijndael Encryption and compressed using Huffman Coding, in order to fully dissimulate the visual  information of the medical image, which is produced as Crypto-Compressed and Encrypted Medical Images (CCEMI). A network with queries generates a secret key, which can be used to encrypt and decrypt of medical images. We have shown that it is more difficult to break a secret key by brute force attack.

Key words: Neural Cryptography, Medical Images, Rijndael algorithm, Crypto-Compression.