<p>Currently, providing privacy to the Internet of Things (IoT) data and maintaining its security are the main concerns in the medical sector. Sophisticated medical systems utilized IoT to gather information, distribute it via smart devices, analyze it, and store it. It is a challenging endeavor to analyze a large volume of data from a variety of IoT devices within a short period. Consequently, it is vital to develop an effective healthcare system featuring secure storage of data along with retrieval. Recently, protecting confidential health information has been elevated to the list of priorities. Therefore, this research unveils an approach to encrypting information through an innovative method named Pseudo-random Obfuscation With Enhanced Randomization (POWER). It consists of several intermediary steps, comprising the generation of a unique master key by utilizing an Artificial Neural Network (ANN) technique, rotation, and shuffling. Upon carrying out several experimental and security analyses, the proposed algorithm exhibits a very randomized byte frequency distribution in the ciphertext with higher entropy and less autocorrelation, i.e., close to zero, and resists frequency-based attacks, statistical-based attacks, differential cryptanalytic attacks, and brute-force attacks. Along with this, IND-CPA and IND-CCA security analyses have also been carried out to assess the semantic security of the POWER algorithm. Again, the avalanche effect was analyzed to verify confusion-diffusion and non-linearity of the proposed algorithm. Moreover, the practicality of the algorithm is evaluated through the investigation of multiple performance metrics, including encryption and decryption time and average memory consumption during execution, under an experimental setting. This study enhances medical data security while offering useful insights to healthcare providers that are dedicated to protecting confidential information about patients in the age of digitization.</p>