Privacy-Preserving AI for Encrypted Medical Imaging: A Framework for Secure Diagnosis and Learning
Abdullah Al Siam, Sadequzzaman Shohan · 2025
Read Privacy-Preserving AI for Encrypted Medical Imaging: A Framework for Secure Diagnosis and Learning by Abdullah Al Siam, Sadequzzaman Shohan (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-21060.txt.
Description
Abdullah Al Siam and Sadequzzaman Shohan put forth a way to read medical images for diagnosis without decrypting them, in this two thousand twenty-five paper. AI diagnosis needs scans to be sent, stored, and worked on, raising risks for patient privacy. Their frame runs a convolutional neural network, called Masked-CNN, on changed or ciphered image forms. Images are kept safe with AES encryption in CBC mode plus JPEG2000 compression, and stay fit for inference.
The tests used chest image sets: the NIH ChestX-ray14 collection and the LIDC-IDRI lung data set. The team read accuracy, inference latency, storage efficiency, and how well privacy held up. Inference on kept-safe images did as well as on plain ones, with loss in how exact it was and in speed. The team calls the frame a bridge between privacy rules and what use clinics get from AI diagnosis.
Catalog record
| CREATOR | Abdullah Al Siam, Sadequzzaman Shohan |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-21060