Optimal medication dosing from suboptimal clinical examples: A deep reinforcement learning approach

By Shamim Nemati, Marzyeh Ghassemi, G. Clifford. 2016

198 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Emergency Medical Services, Epidemiology, Reinforcement Learning, Medical & Health AI

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Emergency Medical Services

In Epidemiology

In Reinforcement Learning

In Medical & Health AI

Citations in the corpus

Reinforcement Learning in Healthcare: A Survey (2019, methodology): “To address this issue, Nemati et al. [164] inferred hidden states of patients via a discriminative hidden Markov model and applied neural FQI to learn optimal heparin dosages.”
Opportunities and obstacles for deep learning in biology and medicine (2017, methodology): “[274] used deep reinforcement learning to optimize a heparin dosing policy for intensive care patients.”
Supervised Reinforcement Learning with Recurrent Neural Network for Dynamic Treatment Recommendation (2018, methodology): “iscover optimal individualized medications for non-small cell lung cancer (NSCLC) based on simulation data, where a support vector regression (SVR) model is used to estimate Q-function. Nemati et al. [25] leveraged a combination of Hidden Markov Models and deep Q-networks to predict optimal heparin ”
Continuous State-Space Models for Optimal Sepsis Treatment: a Deep Reinforcement Learning Approach (2017, methodology): “To our knowledge, this is the ﬁrst extensive application of novel deep reinforcement learning techniques to medical informatics, building signiﬁcantly on the ﬁndings of Nemati et al. (2016).”
Reinforcement Learning based Recommender Systems: A Survey (2021, methodology): “DRL-based Methods RLRS Year SR PO RF EB Metrics Dataset Application Code Slate-MDP [112] 2015 SR1 DQN R1 Simulation Reward N/A Wolpertinger [24] 2015 SR1 DDPG R1 Simulation Reward N/A Nemati et al. [113] 2016 SR2 DQN R2 Oline Reward MIMIC Healthcare CEI [114] 2017 SR3 REINFORCE-wb R1 Simulation P, F”
Inverse reinforcement learning for intelligent mechanical ventilation and sedative dosing in intensive care units (2019, background): “Nemati et al. developed deep RL algorithms that learn an optimal heparin dosing policy from real trails in large electronic medical records [19, 20].”
Reinforcement Learning with Action-Derived Rewards for Chemotherapy and Clinical Trial Dosing Regimen Selection (2018, background): “In the health context, temporal data in the form of electronic health records and intensive care unit interventions over time have both been represented in an MDP’s state (Nemati et al., 2016; Raghu et al., 2017; Prasad et al., 2017).”
Representation and Reinforcement Learning for Personalized Glycemic Control in Septic Patients (2017, background): “RL is a potential approach for sequential decision making with delayed reward or outcome [Nemati et al., 2016].”
Accelerating Drugs Discovery with Deep Reinforcement Learning: An Early Approach (2018, background): “But DRL has been recently also applied to other areas such as robotics [14, 26], natural language processing [27, 38], or health-care [31, 39].”
Supervised-actor-critic reinforcement learning for intelligent mechanical ventilation and sedative dosing in intensive care units (2020, methodology): “[18] used deep RL methods to calculate optimal unfractionated Heparin from sub-optimal clinical ICU data.”
Artificial intelligence to deep learning: machine intelligence approach for drug discovery (2021, background): “Likewise, Nemati et al. [126] developed a deep reinforcement learning model trained on multipa-rameter intelligent monitoring in intensive care II database (MIMIC II) to find an ideal dose of another anticoagulant drug, heparin.”
Learning from an expert (2016, methodology): “Previous works have addressed the reinforcement learning problem of anesthesia’s automation using the BiS (see e.g. [Martín-Guerrero et al., 2009], [Moore et al., 2014], [Nemati et al., 2016]).”
The Actor Search Tree Critic (ASTC) for Off-Policy POMDP Learning in Medical Decision Making (2018, methodology): “Instead of using OPE to provide theoretical policy evaluation, we focus on empirically evaluating our learned policy by comparing how the similarity between clinicians’ decisions and our suggestions indicate patient outcomes: this provides an empirical validation and is commonly adopted [8, 11] for ”
Deep Learning in Pharmacogenomics: From Gene Regulation to Patient Stratification (2018, methodology): “to learn an optimal heparin dosing policy from sample dosing trails and their associated outcomes in large electronic medical records from the publicly available MIMIC II intensive care unit database [100]. Looking forward, we also envision the incorporation of data from mobile devices and wearable ”