This paper presents a DNA-chaos governed cryptosystem for secure storage and sharing of medical images in cloud-based repositories. The system ensures confidentiality, integrity, and availability (CIA) of medical images. The Region of Interest (RoI) is isolated for integrity checks, and a two-tier security framework is proposed, including an additional layer for RoI. A 3D Lorenz chaotic attractor is used to generate keys, increasing the keyspace. DNA-based image diffusion provides high entropy and low correlation, ensuring security. Metadata is encrypted to protect client authentication. A Graphical User Interface (GUI) is developed in Python 3.8 for non-technical users and medical professionals. The system uses a 1D Tent map to generate random pixel indexes for encryption. The proposed approach enhances resistance to chosen plaintext attacks and improves entropy through DNA-based diffusion. The system also includes a secure method for embedding RoI digest in noise-insensitive regions of medical images. The use of chaotic maps and DNA-based encryption ensures high security and resistance to brute-force attacks. The system is tested using DICOM medical images and provides efficient encryption with minimal computational time. The proposed framework ensures secure storage and sharing of medical images in the cloud, protecting patient privacy and data integrity. The system's use of chaotic attractors and DNA-based encryption offers a robust solution for secure medical image repositories.This paper presents a DNA-chaos governed cryptosystem for secure storage and sharing of medical images in cloud-based repositories. The system ensures confidentiality, integrity, and availability (CIA) of medical images. The Region of Interest (RoI) is isolated for integrity checks, and a two-tier security framework is proposed, including an additional layer for RoI. A 3D Lorenz chaotic attractor is used to generate keys, increasing the keyspace. DNA-based image diffusion provides high entropy and low correlation, ensuring security. Metadata is encrypted to protect client authentication. A Graphical User Interface (GUI) is developed in Python 3.8 for non-technical users and medical professionals. The system uses a 1D Tent map to generate random pixel indexes for encryption. The proposed approach enhances resistance to chosen plaintext attacks and improves entropy through DNA-based diffusion. The system also includes a secure method for embedding RoI digest in noise-insensitive regions of medical images. The use of chaotic maps and DNA-based encryption ensures high security and resistance to brute-force attacks. The system is tested using DICOM medical images and provides efficient encryption with minimal computational time. The proposed framework ensures secure storage and sharing of medical images in the cloud, protecting patient privacy and data integrity. The system's use of chaotic attractors and DNA-based encryption offers a robust solution for secure medical image repositories.