Deep Neural Networks for Predicting Post-Surgical Complications Using Multimodal Clinical Data

Authors

  • Shraddha Gendlal Vaidya
  • K. M. Gaikwad
  • Pragati Patil Bedekar
  • Manoj L. Bangare
  • Shalaka Prasad Deore
  • Ramesh Adireddy

DOI:

https://doi.org/10.52783/jns.v14.2750

Keywords:

Multimodal Data, Medical Imaging, Clinical NLP, AI In Healthcare, Risk Prediction, Patient Outcomes, Complication Detection

Abstract

Post-surgical complications significantly impact patient recovery, hospital resource utilization, and healthcare costs. Early prediction of such complications enables timely interventions, reducing morbidity and mortality rates. Traditional predictive models rely on handcrafted features and domain-specific knowledge, often limiting their accuracy and adaptability. In this study, we propose a deep neural network (DNN) approach for predicting post-surgical complications using multimodal clinical data, including structured electronic health records (EHRs), unstructured clinical notes, and medical imaging. Our framework integrates convolutional neural networks (CNNs) for imaging analysis, recurrent neural networks (RNNs) for sequential patient records, and transformer-based models for clinical text processing. A multimodal fusion layer combines these diverse data representations, capturing intricate relationships between different modalities. We trained and validated our model on a large hospital dataset containing records from 50,000 surgical patients. Experimental results show that our approach outperforms traditional machine learning models, achieving an accuracy of 92.3%, a precision of 91.7%, and an AUC-ROC score of 0.94. We also employ SHAP (SHapley Additive exPlanations) to enhance model interpretability, identifying key predictive factors such as preoperative vitals, surgical procedure details, and early post-operative lab results. Our findings demonstrate that deep learning models, particularly multimodal fusion networks, significantly improve the prediction of post-surgical complications. Future research will focus on expanding datasets, addressing data imbalance, and improving model explainability for clinical adoption. Our study highlights the potential of deep learning to transform surgical outcome prediction and improve patient care.

Downloads

Download data is not yet available.

Metrics

Metrics Loading ...

References

XU H, LIU Q, CAO K, Distribution, Characteristics, and Management of Older Patients with Valvular Heart Disease in China: China-DVD Study [J]. JACC: Asia, 2022, 2(3, Part 2): 354-65.

MAENO Y, ABRAMOWITZ Y, ISRAR S, Prognostic Impact of Permanent Pacemaker Implantation in Patients With Low Left Ventricular Ejection Fraction Following Transcatheter Aortic Valve Replacement [J]. The Journal of invasive cardiology, 2019.

KHATRI P J, WEBB J G, RODéS-CABAU J, Adverse effects associated with transcatheter aortic valve implantation: a meta-analysis of contemporary studies [J]. Annals of Internal Medicine, 2013.

KAILASH K, PRADEEPA M, MIROSLAV M, A Deep Learning Approach for Kidney Disease Recognition and Prediction through Image Processing [J]. Applied Sciences, 2023.

LUO P, LI Y, TIAN L-P, Enhancing the prediction of disease-gene associations with multimodal deep learning [J]. Bioinformatics, 2019.

S. B. Baker, W. Xiang, and I. Atkinson, ‘‘Internet of Things for smart healthcare: Technologies, challenges, and opportunities,’’ IEEE Access, vol. 5, pp. 26521–26544, 2017.

S. Sikdar and S. Guha, ‘‘Advancements of healthcare technologies: Paradigm towards smart healthcare systems,’’ in Recent Trends in Image and Signal Processing in Computer Vision. Singapore: Springer, 2020, pp. 113–132.

K. B. DeSalvo, A. N. Dinkler, and L. Stevens, ‘‘The U.S. Office of the national coordinator for health information technology: Progress and promise for the future at the 10-year mark,’’ Ann. Emergency Med., vol. 66, no. 5, pp. 507–510, Nov. 2015.

B. Norgeot, B. S. Glicksberg, and A. J. Butte, ‘‘A call for deeplearning healthcare,’’ Nature Med., vol. 25, no. 1, pp. 14–15, Jan. 2019.

D. Mahapatra, P. K. Roy, S. Sedai, and R. Garnavi, ‘‘Retinal image quality classification using saliency maps and CNNs,’’ in Proc. Int. Workshop Mach. Learn. Med. Imag. Singapore: Springer, 2016, pp. 172–179.

WANG J, DING H, BIDGOLI F A, Detecting Cardiovascular Disease from Mammograms with Deep Learning [J]. IEEE Transactions on Medical Imaging, 2017.

JILAIHAWI H, MAKKAR R R, KASHIF M, A revised methodology for aortic-valvar complex calcium quantification for transcatheter aortic valve implantation [J]. European Heart Journal - Cardiovascular Imaging, 2014.

TSOI M, TANDON K, ZIMETBAUM P J, Conduction Disturbances and Permanent Pacemaker Implantation After Transcatheter Aortic Valve Replacement: Predictors and Prevention [J]. Cardiology in Review, 2022.

SAMMOUR Y, KRISHNASWAMY A, KUMAR A, Incidence, Predictors, and Implications of Permanent Pacemaker Requirement After Transcatheter Aortic Valve Replacement [J]. JACC: Cardiovascular Interventions, 2021.

GAMET A, CHATELIN A, MERGY J, Does aortic valve calcium score still predict death, cardiovascular outcomes, and conductive disturbances after transcatheter aortic valve replacement with new-generation prostheses? [J]. Journal of Cardiovascular Echography, 2020.

Downloads

Published

2025-03-28

How to Cite

1.
Gendlal Vaidya S, Gaikwad KM, Bedekar PP, Bangare ML, Prasad Deore S, Adireddy R. Deep Neural Networks for Predicting Post-Surgical Complications Using Multimodal Clinical Data. J Neonatal Surg [Internet]. 2025Mar.28 [cited 2025Oct.2];14(9S):749-61. Available from: https://www.jneonatalsurg.com/index.php/jns/article/view/2750

Most read articles by the same author(s)