Healthcare technology continues to advance each year, including IoT-assisted wearable sensor devices, AI, and blockchain, as well as improved medical diagnosis.
Jeff Elton, CEO at ConcertAI has outlined several expert industry healthtech predictions around SaaS, real-world data and AI in evolving healthcare, precision medicine and clinical trials.
AI will enhance life sciences through advanced Digital Twins and AI-designed drugs
Elton explains how accuracy will be the turning point of AI, supported by LLMs over ambient AI. In particular, Elton states: “AI regulations in healthcare will be marked by highly differentiated approaches and AI adoption. Oncology-specific AI and LLM systems will work together across the entire lifecycle – from discovery to clinical trials.”
With digital twins, Elton provides the advantages: “Twins will simulate early-phase clinical trials and will help identify the most beneficial and likely successful drugs in trials. Fully integrated patient-to-trial matching in provider workflows will cut time and costs for late-stage trials by 30 percent. All of this will lower costs and increase the success rates of pharmaceutical R&D processes.”
These different programs will unify discovery, translational, and developmental processes
Accuracy will be the turning point of AI, supported by LLMs over ambient AI.
There will continue to be an increase in the integration of AI in daily workflows and decision-making as AI increases in accuracy and efficiency, opines Elton.
The expert finds: “2025/2026 will see the enormous potential of AI as a ‘decision augmentation’ of expert humans. This will come from context-sensitive solutions, LLMs, that can align other LLMs to collect, analyse, and recommend options to clinical teams that are aligned to that specific decision and the unique characteristics of that patient. This needs to and will happen as there are not enough staff and specialists to provide the needed care.”
AI regulations in healthcare will be marked by highly differentiated approaches and AI adoption.
Elton believes AI regulation currently encourages responsible innovation and self-regulation, allowing space for new advancements.
As examples, Elton provides: “Current Gen 1 solutions, often single-LLM-based, are expected to be short-lived and evolve into multi-model, highly tuned systems with domain-specific models and advanced prompt engineering. Most healthcare AI will be run on data locally, edge deployed, or done within secure, segregated clouds to ensure control, prevent misuse, and protect patient health data.”
Elton concludes with: “Lastly, leading AI SaaS solutions will heavily publish performance metrics, certify against model drift, and provide transparent data flow and model disclosures. This would be the equivalent of certifying drug safety and manufacturing standards.”
Oncology-specific AI and LLM systems will work together across the entire lifecycle – from discovery to clinical trials.
As to the benefits from large language models (LLMs), Elton has formulated: “LLMs specific to oncology will allow for the consideration of Agents as “Interpretation Experts” with performance comparable to the highest-trained humans for 90%+ of patient decisions. There will be a new generation of multi-modal infrastructure with persistent data analysis occurring over the life of the clinical trial. Trial designs, patient matching, data collection, and real-time analysis will have AI enablement throughout the process.”
AI2AI programs will unify discovery, translational, and developmental processes.
This part of the process is one grounded in the future state. The first translational programs will be simulated at a mass scale and will give way to new scientific and AI-based engineering approaches.
According to Elton: “This ensures that the notion of a “funnel” for clinical development is left behind.”
