|
Development of an AI-Based Predictive Model for Assessing Genetic Vulnerability to Marine Toxins in Seafarers and Divers
|
Sakineh Khanamani Falahatipour   , Salimeh Khanamani Falahati-pour , Amirreza Ghasemloo , Somayyeh Karami-Mohajer , Zohreh Oghabian , Soudeh Khanamani Falahati-pour , Motahareh Soltani * |
| Pistachio Health Research Center, Rafsanjan University of Medical Sciences, Rafsanjan, Iran , motahareh.soltani@gmail.com |
|
|
Abstract: (1499 Views) |
Background and Aim: Marine chemical and biotoxin exposures pose substantial health risks to seafarers and divers, notably in the Persian Gulf and the Sea of Oman. Individual susceptibility varies with genetic polymorphisms (SNPs) in detoxification genes such as the CYP450 family. This study develops and specifies a conceptual AI-based framework to predict genetic vulnerability and defines its evaluation protocol. The key novelty is the multi-source integration of genetic and environmental data with optional links to telemedicine and real-time alerts.
Methods: The framework integrates: (1) genetic data from seafarers/divers (SNP profiles in toxin-metabolism genes), (2) environmental measurements from marine sensors and satellite imagery (e.g., ciguatoxin/saxitoxin/ domoic-acid levels, water quality, temperature), and (3) toxin knowledge bases such as ToxCast and T3DB. Three model families -deep neural networks, random forests, and support vector machines- are considered. Evaluation follows train/validation/test splits with cross-validation; performance metrics include accuracy, precision, recall, and F1-score (AUC may be reported for binary setups). Target geographical scope for field data is the Persian Gulf and the Sea of Oman.
Results: This is a conceptual/framework study; no numerical or experimental results are reported at this stage. Main outputs are: (a) the model architecture and processing flowchart, (b) executable data specifications and sources (marine sensors, satellite, ToxCast/T3DB, genetic sampling), (c) a quantitative evaluation protocol with explicit metrics, and (d) a clear separation between innovation (genetic+environmental+AI integration) and applications (diving/occupational medicine, maritime assistance via telemedicine, harmful algal bloom early warnings). Key challenges (at-sea compute constraints, data quality/completeness, and genetic-data ethics) and their design implications are documented.
Conclusion: The proposed framework delineates a pragmatic path to implement and evaluate a personalized risk-prediction system, with potential to enhance occupational health, reduce hazardous exposures, and support remote clinical decisions. Empirical validation with real-world cohorts and formal reporting of quantitative metrics constitute the next research step. |
|
| Keywords: Genetic predisposition to disease, Marine toxins, Artificial intelligence, Seafarers, Health prediction |
|
|
Full-Text [PDF 526 kb]
(641 Downloads)
|
Type of Study: Original Article |
Subject:
Marine Medicine
Received: 2025/08/8 | Accepted: 2025/11/19 | Published: 2025/12/1
|
|
|
|
|
|
