Show simple item record

AuthorAbdelfattah, Sherif
AuthorBadr, Mahmoud M.
AuthorMahmoud, Mohamed M. E. A.
AuthorAbualsaud, Khalid
AuthorYaacoub, Elias
AuthorGuizani, Mohsen
Available date2024-03-26T11:56:46Z
Publication Date2023
Publication NameIEEE Internet of Things Journal
ResourceScopus
ISSN23274662
URIhttp://dx.doi.org/10.1109/JIOT.2023.3303429
URIhttp://hdl.handle.net/10576/53518
AbstractDecision tree (DT) models are widely used in medical applications where the size of the data sets is usually small or medium. Moreover, DT ensemble models are preferred over single DT models because of their higher accuracy in spite of the need for more overhead due to using multiple trees. Several schemes have been proposed for privacy-preserving cloud-based medical diagnosis using ensemble models. However, these schemes suffer from several limitations. First, they suffer from high computation/communication overheads due to using inefficient public-key cryptosystems. Second, none of them can simultaneously protect the intellectual property of the model and preserve the privacy of the patients' data and diagnosis results. Finally, they do not provide inherent access control for the outsourced model and micropayment, in which only the registered patients can use the model and pay for the service. In this article, we develop a lightweight and privacy-preserving cloud-based medical diagnosis scheme using ensemble models with high accuracy and acceptable overhead. Using our scheme, the model owner can control the patients who can use the model. Also, for each classification operation, patients must make a micro-payment to pay for the diagnosis service. Our analysis indicates that our scheme can protect the model's intellectual property and diagnose diseases without leaking any sensitive information about the patients' medical data and the diagnosis results. Our experimental results demonstrate that our scheme requires less communication/computation overhead compared to the existing schemes.
Languageen
PublisherInstitute of Electrical and Electronics Engineers Inc.
SubjectAccess control
ensemble classification
inner product
medical diagnosis
privacy preservation
TitleEfficient and Privacy-Preserving Cloud-Based Medical Diagnosis Using an Ensemble Classifier With Inherent Access Control and Micro-Payment
TypeArticle
Pagination22096-22110
Issue Number24
Volume Number10
dc.accessType Abstract Only


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record