Table 3.

Summary for e-health data security and privacy studies.

Ref.	Target Security Concern	Target Security Requirement	Security Solution	Strength	Weakness
[21]	- Unauthorized access - Data disclosure - Data tampering	- Access restriction - Confidentiality - Anonymity - Accountability	- Access Control - Anonymization	Anonymization and access control according to the sensitivity of e-health data	The current anonymization method has some vulnerabilities that could lead to re-identification
[22]	- Unauthorized access - Data disclosure - Data loss - Data tampering - Data forgery	- Access restriction - Confidentiality - Integrity - Availability - Accountability	- Access Control - Blockchain (Hyperledger Fabric)	No single point of failure (SPoF) problem and the emergency scenario is considered	Emergency access permission could be abused and access control requires a transaction fee
[23]	- Unauthorized access - Data disclosure - Data loss - Data tampering - Data forgery	- Access restriction - Confidentiality - Integrity	- Access Control - Blockchain (Ethereum)	No SPoF problem and employing an off-chain scaling method to solve the scalability problem of blockchain	Lack of fine-grained access control and access control requires a transaction fee
[24]	- Unauthorized access - Data disclosure - Data loss - Data tampering - Data forgery	- Access restriction - Confidentiality - Integrity - Accountability	- Access Control - AES - Blockchain	A user can revoke permission to access their e-health data from medical staff at any time	Symmetric key was shared to control access; therefore, once a key has been shared with someone or compromised, the key should be updated and related e-health data should be re-encrypted
[27]	- Data disclosure	- Confidentiality	- AES - RSA - Steganography	Discrete wavelet transform is compatible with compression and has resistance to geometric distortions	IoT is a target environment; however, AES and RSA that require high computational power were used
[28]	- Data disclosure	- Confidentiality	- Quaternion-based Encryption	Fast computation speed for the encryption of a large volume of e-health data	The computation speed could be increased if the decomposition process were omitted
[29]	- Data disclosure	- Confidentiality	- ECEM-based Encryption	Differential attack resistance	Performance should be evaluated on diverse medical images
[36]	- Data disclosure - Data tampering - Data loss - Data forgery	- Confidentiality - Integrity - Anonymity - Accountability	- AES - Blockchain	No SPoF problem and medical data can be securely preserved with the blockchain	The transaction fee is relatively high compared to conventional data storage
[37]	- Data disclosure - Data tampering - Data loss - Data forgery	- Access restriction - Confidentiality - Integrity - Anonymity - Accountability	- Access Control - Blockchain	Efficient consensus mechanism and access control protocol for e-health data were also proposed	Access should be able to be delegated to related medical staff or other people in a secure manner for flexible data sharing
[38]	- Data disclosure - Data tampering - Data loss - Data forgery	- Integrity - Accountability	- Blockchain	Blockchain only records URLs instead of medical images that have a large data size	Since real images are stored in the hospital’s database, both the blockchain and endpoints should be properly protected
[39]	- Data disclosure	- Confidentiality	- Rijndael encryption - Steganography	Fast processing time and high embedding capacity based on LSB	Message capacity could be increased with noise cancellation and data reduction
[40]	- Data disclosure - Data tampering	- Confidentiality - Integrity	- AES - Steganography	It considered both confidentiality and integrity and had high capacity, robustness, and imperceptibility	An error control mechanism should be adopted for a robust steganography method
[41]	- Data disclosure - Data tampering - Data forgery	- Confidentiality - Integrity	- Watermarking	Resistance to sharpening and blurring attacks while maintaining acceptable imperceptibility	A tracking key that makes the proposed scheme reversible has to be transmitted with each medical image
[42]	- Data disclosure - Data tampering - Data forgery	- Confidentiality - Integrity	- Watermarking	Proposes an effective scheme to localize and restore tampered pixels and regions	Various tampering attacks on image resizing, skewing, and rotating should be studied