EUVIMEDEUVIMED entdecken
Crossref

An Explainable Deep Learning Framework for Transparent and Personalized Recommendations in Clinical Decision-Making

Anand V, Venkata Subramanian N, S. Bhaskaran

International Journal of Drug Delivery Technology · 19.05.2026 · 16 (30s)

DOI: 10.25258/ijddt.16.30s.92 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 19.05.2026. Verfasst von Anand V und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We develop an explicable system of deep learning-assisted drug delivery recommendations in clinical decision-making to meet the important requirement of interpretability in medical applications. The framework combines structured longitudinal clinical data such as patient-medication interactions, clinical contents, and contextual data to produce specific therapeutic guidance. The patient history is represented as learned embeddings, the temporal position signal, and candidate drug options are represented as structured attributes; finally, they are combined using a Transformer layer to ensure that a complicated dependence is captured. The model calculates compatibility scores of the state of a patient and a drug choice, making it possible to recommend them personally with an explanation at multi-level. In addition, the model uses temporal explanations based on attention, SHAP-based feature attribution, integrated gradients, and rule-based clinical reasoning to provide better interpretability to clinicians. The training goal is a co-optimization of the accuracy of recommendations, interpretability, regularization, and wall is safe which guarantees clinically plausible results. Experiments indicate that not only does the framework enhance the performance of the recommendations, but the actionable insights into the decisionmaking process are also provided as well. The given approach makes the divide between black-box models of deep learning and clinical application, providing a useful place of personalized medicine. The design is modular and can integrate easily with current clinical processes and may minimize adverse drug events and enhance the patient outcomes. The work contributes to the field by integrating the state-of-the-art methods of deep learning with thorough explainability, which leads to building trust and acceptance in real-life healthcare environments.

Zitieren

APA

V, A., N, V. S., & Bhaskaran, S. (2026). An Explainable Deep Learning Framework for Transparent and Personalized Recommendations in Clinical Decision-Making. International Journal of Drug Delivery Technology, 16(30s). https://doi.org/10.25258/ijddt.16.30s.92

Vancouver

V A, N VS, Bhaskaran S. An Explainable Deep Learning Framework for Transparent and Personalized Recommendations in Clinical Decision-Making. International Journal of Drug Delivery Technology. 2026;16(30s). doi:10.25258/ijddt.16.30s.92

BibTeX
@article{V2026,
  author = {V, Anand and N, Venkata Subramanian and Bhaskaran, S.},
  title = {An Explainable Deep Learning Framework for Transparent and Personalized Recommendations in Clinical Decision-Making},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {30s},
  doi = {10.25258/ijddt.16.30s.92}
}
Kostenlos starten →