Deep learning in systems medicine.

Joao Luis De Miranda,Ivan Chorbev,Vojtech Spiwok,Massimiliano Zanin,Steven Watterson,Tiago De Melo Malaquias,Roisin Mcallister,Paolo Tieri,Estelle Pujos-Guillot,Blandine Comte,Filippo Castiglione,Taranjit Singh Rai,Haiying Wang,Huiru Zheng

doi:10.1093/bib/bbaa237

Abstract

Systems medicine (SM) has emerged as a powerful tool for studying the human body at the systems level with the aim of improving our understanding, prevention and treatment of complex diseases. Being able to automatically extract relevant features needed for a given task from high-dimensional, heterogeneous data, deep learning (DL) holds great promise in this endeavour. This review paper addresses the main developments of DL algorithms and a set of general topics where DL is decisive, namely, within the SM landscape. It discusses how DL can be applied to SM with an emphasis on the applications to predictive, preventive and precision medicine. Several key challenges have been highlighted including delivering clinical impact and improving interpretability. We used some prototypical examples to highlight the relevance and significance of the adoption of DL in SM, one of them is involving the creation of a model for personalized Parkinson’s disease. The review offers valuable insights and informs the research in DL and SM.

Highlights

Systems medicine (SM) has emerged as an interdisciplinary field, which promotes an integrative and holistic approach to studying the human body at the systems level with the aim of improving our understanding, prevention and treatment of complex diseases [1, 2]
deep learning (DL) is a branch of machine learning (ML) and artificial intelligence (AI) that employs a layered structure of computation to learn data representation with multiple levels of abstraction [17]
These set of studies contributed to show the importance of DL methods for precision medicine; in addition, they were associated in a good manner with clinical approaches

Summary

Introduction

Systems medicine (SM) has emerged as an interdisciplinary field, which promotes an integrative and holistic approach to studying the human body at the systems level with the aim of improving our understanding, prevention and treatment of complex diseases [1, 2]. It has been suggested that in order to tackle complicated tasks such as the discovery of complex disease patterns with multiple facets from data and realize the full potential of machine learning (ML) in the era of big data, learning models need to go deep and various deep learning (DL) architectures hold great promise in this endeavour [5,6,7]. Other DL reviews are being published under various approaches; some reviews are addressing models and/or methodologies [8,9,10]; others are focusing either general applications [11] or specific tools (e.g. embedding graphs [12, 13]), or even DL works targeting a certain field (e.g. pharmaceutical research and drug design/discovery [14,15,16])

Introduction to DL

Conclusions

Findings

Key Points