. 2023 Feb 28;13(2):203–213. doi: 10.1007/s12553-023-00738-2

Table 2.

Examples of AI applications within Recruitment

Opportunity	Therapeutic Area	AI application
Patient selection and access fairness	Oncology and Cardiovascular	• Oncology: • Facilitation of cohort selection (e.g., AI technology applied to medical records to ameliorate recruitment and identification of suitable patients) [20, 44] • An AI-enabled clinical decision support system (CDSS) -based on natural language processing of cancer specific values and ML methods- to accurately identify eligible subjects with a high degree of sensitivity and specificity during a retrospective review of four breast cancer focused trials [20, 45] • Cardiovascular: AI/ML-based fairness metrics established for the purpose of equity in trial access [21]
Biomarker refinement	Neurology (Alzheimer) and Amyotrophic Lateral Sclerosis (ALS))	• In Alzheimer disease: an AI classifier was optimized to detect asymptomatic cases for CT recruitment (otherwise not identified using the biomarker amyloid plague) [19] • In ALS: It has been shown that a robust ML survival model includes a broader approach to patient inclusion in CTs, by identifying patients that could have still benefitted from a trial despite originally being excluded [18]
Large scale analytics to support trial matching search engine	Infectious diseases (HIV)	• Large public database of interventional trials developed using AI, to support a search engine for a trial matching system to be used by HIV patients [22]

Opportunity

Therapeutic Area

AI application

Patient selection and access fairness

Oncology and Cardiovascular

• Oncology:

• Facilitation of cohort selection (e.g., AI technology applied to medical records to ameliorate recruitment and identification of suitable patients) [20, 44]

• An AI-enabled clinical decision support system (CDSS) -based on natural language processing of cancer specific values and ML methods- to accurately identify eligible subjects with a high degree of sensitivity and specificity during a retrospective review of four breast cancer focused trials [20, 45]

• Cardiovascular: AI/ML-based fairness metrics established for the purpose of equity in trial access [21]

Biomarker refinement

Neurology (Alzheimer) and Amyotrophic Lateral Sclerosis (ALS))

• In Alzheimer disease: an AI classifier was optimized to detect asymptomatic cases for CT recruitment (otherwise not identified using the biomarker amyloid plague) [19]

• In ALS: It has been shown that a robust ML survival model includes a broader approach to patient inclusion in CTs, by identifying patients that could have still benefitted from a trial despite originally being excluded [18]

Large scale analytics to support trial matching search engine

Infectious diseases (HIV)

• Large public database of interventional trials developed using AI, to support a search engine for a trial matching system to be used by HIV patients [22]