cs.CYApr 14, 2026

From Prediction to Intervention: The Evolution of AI in Biomedicine

Authors: Andrew Feinberg, Aleksandr Sarachakov, Viktor Svekolkin, Alexander Bagaev, Ferran Prat, Michael Feinberg

Organizations: BostonGene Corporation, Waltham, MA, USA

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Artificial intelligence has advanced rapidly in biomedicine through large-scale multimodal data integration, enabling increasingly accurate prediction of clinical outcomes and patient stratification. These systems, however, remain fundamentally observational: they learn statistical associations from historical data and operate within previously observed biological and clinical states, limiting their ability to generalize to novel therapies or unobserved interventions. We argue that AI in biomedicine is undergoing a structural transition. As biomedical decision-making increasingly depends on reasoning about intervention rather than extrapolation from past observations, predictive architectures become structurally insufficient. Systems that learn from historical data cannot, by construction, represent how biological systems evolve under perturbation, and therefore cannot reliably support decision-making in the presence of novel interventions. We introduce a conceptual framework distinguishing observational and interventional intelligence and define disease-level models as systems that explicitly represent the state, dynamics, and intervention response of biological processes. These models enable a shift from inference to simulation -- reasoning about what will happen under intervention rather than what is likely based on the past. This transition also implies a shift in where value is created: from data processing and prediction toward systems that support and define decision-making under intervention. It follows directly from the structure of biomedical decision-making and defines the next stage of AI in medicine. Systems that cannot model intervention will be structurally excluded from decision-making.

Figures & tables

Figure 1: From observational to interventional intelligence in biomedicine. Current AI systems in biomedicine operate within an observational paradigm, learning statistical associations from historical multimodal data to generate predictions such as risk scores, classifications, and patient stratification. While highly effective within previously observed distributions, these systems are inherently constrained in their ability to reason about novel interventions. The emerging paradigm shifts toward interventional intelligence, in which models explicitly represent disease as a dynamic biological process and simulate its evolution under therapeutic perturbations. This transition—from prediction to simulation, and from correlation to modeled dynamics—enables reasoning about counterfactual scenarios and supports decision-making under conditions not directly observed in training data.
Figure 2: The Disease Model Triad. Disease-level models are defined by three interdependent components: state representation, which encodes the biological configuration of the system across tumor, immune, and microenvironmental dimensions; state dynamics, which describe how this state evolves over time under endogenous processes; and the intervention operator, which maps therapeutic actions to transitions in system state. Together, these components define a minimal computational architecture for interventional intelligence, enabling models to simulate disease trajectories under perturbation and to reason about counterfactual outcomes beyond observed data.
AspectObservationalInterventional
Core questionWhat is likely?What will happen if?
Data basisHistoricalModeled + data
OutputPredictionSimulation
LimitationDistribution-boundModel-bound
Table 1: Observational AI vs Interventional AI
Figure 3: From state representation to decision through simulation. Disease-level models operate by representing the biological state of a system, modeling its dynamics, and applying interventions as perturbations to that state. Simulation of these perturbations generates trajectories that reflect how disease evolves under different therapeutic actions. Decisions are then derived from these simulated trajectories rather than inferred from historical correlations. This framework captures the transition from observational to interventional intelligence, where reasoning about intervention replaces prediction based on past observations.

Explore similar work

CardsList

  1. Towards World Models in Biomedical Research

    Jun 4, 2026Guangyu Wang, Jingkun Yue, Siqi Zhang +19AI for ScienceWorld Model Learning

    State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing, China · Department of Engineering Science, University of Oxford, Oxford, United Kingdom · Institute of Medical Artificial Intelligence, South China Hospital, Medical School, Shenzhen University, Shenzhen, Guangdong, China +8

  2. Medical world models: representing medical states, modelling clinical dynamics and guiding intervention policies

    Jun 15, 2026Ke Liu, Mengxuan Li, Yanyi Bao +4World Model LearningClinical Decision Support

    College of Computer Science, Zhejiang University, Hangzhou, China. · School of Medicine, Zhejiang University, China. · Department of Biomedical Informatics, Harvard University, USA.

  3. From Static Risk to Dynamic Trajectories: Toward World-Model-Inspired Clinical Prediction

    May 16, 2026Pujun Feng, Xiaoyu Guo, Seyed Ehsan Saffari +10Causal Effect EstimationOff-Policy Evaluation

    Faculty of Applied Sciences, Macao Polytechnic University, Macao SAR 999078, China. · Medin.ai, Beijing 100871, China. · School of Software & Microelectronics, Peking University, Beijing 100871, China. +8