EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A modular deep learning architecture for interpretable disease prediction across tabular clinical and biometric datasets

Vijay U Rathod, Siddhesh Sanjay Amrutkar, Kirti A Patil, Aoudumber D Londhe, Sandip Y Bobade, Virendrakumar A Dhotre, Melkamu Workie Kebede iD

PLOS One · 08.05.2026 · 21 (5) · e0348670

DOI: 10.1371/journal.pone.0348670 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 08.05.2026. Verfasst von Vijay U Rathod und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of CSE (Artificial Intelligence and Machine Learning), Vishwakarma Institute of Technology (Affiliated to Savitribai Phule Pune University…. Hauptthemen laut MEDLINE: Deep Learning, Biometry. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate disease prediction using clinical datasets is essential for improving early diagnosis and clinical decision-support systems; however, many existing deep learning approaches are disease-specific, computationally intensive, and difficult to generalize across heterogeneous biomedical datasets. This study addresses this challenge by proposing a unified and dataset-aware deep learning framework that enables accurate and interpretable disease prediction across diverse clinical datasets. The framework adopts a modular architecture that selects appropriate models based on dataset characteristics such as feature dimensionality, sample size, and class imbalance. It integrates multiple deep learning architectures, including MLP, one-dimensional CNN, FT-Transformer, autoencoder-based classifiers, and ensemble strategies. Robust preprocessing, fold-safe feature selection, and nested cross-validation are incorporated to ensure reliable performance evaluation. The framework is evaluated on three heterogeneous benchmark datasets: the UCI Heart Disease dataset (303 samples, 13 clinical features), the PIMA Indians Diabetes dataset (768 samples, 8 metabolic features), and the Parkinson’s disease voice dataset (195 recordings, 22 acoustic features). Experimental results demonstrate competitive predictive performance relative to classical baselines across the diverse tasks. The FT-Transformer + autoencoder ensemble achieved an AUC of 0.8980 (±0.0483) for heart disease prediction, while the CNN + Autoencoder ensemble obtained an AUC of 0.8451 (±0.0270) for diabetes classification. For Parkinson’s disease detection, the MLP achieved an AUC of 0.7538 with perfect specificity. Overall, all models achieved AUC values comparable to ML baselines. The study contributes a scalable and interpretable deep learning framework that improves reliability, generalization, and practical applicability for multi-disease prediction in real-world healthcare environments.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningBiometryHumansAutoencoderConvolutional Neural NetworksPrediction AlgorithmsClassification AlgorithmsDatabases, FactualPredictive Learning Models

Autor:innen und Institutionen

  • Department of CSE (Artificial Intelligence and Machine Learning), Vishwakarma Institute of Technology (Affiliated to Savitribai Phule Pune University, Pune), Pune, Maharashtra, India.
  • Department of Information Technology, MET's Institute of Engineering (Affiliated to Savitribai Phule Pune University, Pune), Nashik, Maharashtra, India.
  • Department of Artificial Intelligence and Data Science, Vishwakarma Institute of Technology (Affiliated to Savitribai Phule Pune University Pune), Pune, Maharashtra, India.
  • Department of Computer Science and Engineering (Artificial Intelligence), Vishwakarma Institute of Technology (Affiliated to Savitribai Phule Pune University Pune), Pune, Maharashtra, India.
  • Department of Computer Science and Engineering (Artificial Intelligence and Machine Learning), Vishwakarma Institute of Technology (Affiliated to Savitribai Phule Pune University, Pune), Pune, Maharashtra, India.
  • Department of Information Technology, Institute of Technology, Debre Markos University, Debre Markos, Ethiopia.

ORCID-Profile: Melkamu Workie Kebede

Zitieren

APA

Rathod, V. U., Amrutkar, S. S., Patil, K. A., Londhe, A. D., Bobade, S. Y., Dhotre, V. A., & Kebede, M. W. (2026). A modular deep learning architecture for interpretable disease prediction across tabular clinical and biometric datasets. PLOS One, 21(5), e0348670. https://doi.org/10.1371/journal.pone.0348670

Vancouver

Rathod VU, Amrutkar SS, Patil KA, Londhe AD, Bobade SY, Dhotre VA, et al. A modular deep learning architecture for interpretable disease prediction across tabular clinical and biometric datasets. PLOS One. 2026;21(5):e0348670. doi:10.1371/journal.pone.0348670

BibTeX
@article{Rathod2026,
  author = {Rathod, Vijay U and Amrutkar, Siddhesh Sanjay and Patil, Kirti A and Londhe, Aoudumber D and Bobade, Sandip Y and Dhotre, Virendrakumar A and Kebede, Melkamu Workie},
  title = {A modular deep learning architecture for interpretable disease prediction across tabular clinical and biometric datasets},
  journal = {PLOS One},
  year = {2026},
  volume = {21},
  number = {5},
  pages = {e0348670},
  doi = {10.1371/journal.pone.0348670}
}
Kostenlos starten →