Artificial intelligence (AI) is strongly revolutionizing health care. That is not true only in terms of cutting costs, improving test quality and reducing time. AI is definetely finding biomarkers that can lead to precision medicine, recruiting patients faster than before, reading text volumes in seconds, with revolutionary impact on discoveries involving new diagnostic tools and drugs for many diseases.
According to the BenSci platform, there are at the moment 106 start-ups dedicated to AI in the medical field, but the number is growing daily.
Drug Discovery and Artificial Intelligence
Aggregating and summarizing information
Understand the mechanisms underlying diseases
Generate data and models
Finding new solutions and existing drugs
Generate new drugs
Validate new molecules
Draw new drugs
Draw preclinical experiments
Manage preclinical experiments
Draw clinical processes
Recruit patients in clinical judgment
Optimize clinical trials
Artificial intelligence and clinical trials
Artificial intelligence can influence every phase of clinical experimentation. At the moment less than 5% of cancer patients are enrolled in clinical trails, although it is estimated that to have great innovation in drug analisys at least 25% of patients must be included in the studies.
The reasons are not only usual: times and costs greatly influence the choice of the subjects, who may find themselves in difficulty and fulfill the requests underpinned by an experimental clinical protocol. According to the report “The future of clinical trials: how AI and Big Tech could make cheaper, faster and more effective drug development”, published in August 2018 by CBInsight, it happens in over 80% of cases that clinical trials fail to meet the expected times. In the same way, one third of the interruptions of the phase 3 studies is due to problems of enrollment of patients.
Artificial Intelligence can help both sponsor companies and patients. AI can dare a hand when an individual identifies a process, extending information from the medical records that provide for inclusion and exclusion criteria. The technology can also be useful for adherence to treatment, which can be monitored with different solutions already today.