Akademik Veri Yönetim Sistemi
16th Medical Informatics Congress, Ankara, Türkiye, 22 - 23 Mayıs 2025, cilt.3, ss.242-245, (Özet Bildiri)
Multi-omics data are produced by merging many biological data types, including
proteomics, metabolomics, and genomes. These data help us gain a more thorough
understanding of people's biological states and are essential for the creation of individualized
healthcare treatment plans and diagnostic techniques. However, a major obstacle to the
security and accuracy of health data is the safe and effective transfer of multi-omics data.
For the efficient transfer of multi-omics data in wire-free health systems, a method based on
non-orthogonal multiple access (NOMA) has been suggested in this paper. NOMA is a
technique that uses the same frequency band for several users to share transmission power
in an equitable manner. Three distinct users' distances from the base station were used to
determine the channel gains for the study, and power allocation was done in accordance with
these gains. The inverse channel gain approach has been used to optimize power allocation,
while machine learning techniques (K-Nearest Neighbors and Random Forest) have been
used to predict power allocation. In addition, the Successive Interference Cancellation (SIC)
algorithm has been used to reduce user interference. The results reveal that the proposed
NOMA-based approach is very reliable and efficient for multi-omics data transmission, with
digital success rates surpassing 95%. Furthermore, with a short processing time (2.18
seconds) and low memory utilization (314.24 MB → 304.49 MB), the suggested system has
shown considerable computational efficiency. This study describes an important strategy for
increasing the integration of multi-omics data into health information systems.