Applied AI
Applied Artificial Intelligence (Applied AI) involves the pragmatic deployment of advanced Artificial Intelligence methodologies to address tangible challenges across manifold domains. It encompasses the use of cutting-edge techniques, including among others machine learning algorithms, natural language processing, computer vision, and robotics, to devise hardware and software systems proficient in executing tasks conventionally requiring human-like cognitive abilities. In the context of medicine, Applied AI offers the opportunity to analyse large datasets (e.g., medical images, genomic data, and other health records) to extract knowledge, personalise and discover treatments, automate and speed processes, and, in general, assist and support healthcare professionals.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Support and automation towards effectiveness and efficiency
- Physician's focus on the narrative and holistic dimension of the patient
- Prompt analysis and detection of large quantities of collected data improving prevention and treatment
Requirements
- Environmental/economic sustainability
- New risk models associated to collected/processed data
- Reliability, trustability and explainability
- Human in the loop approach
Risks
- Poor outcomes (bias, low-quality data, lack of transparency, use in inappropriate contexts)
- Leaks of personal data and unclear accountability
- Bias leading to safety-critical errors
Impacts
- Availability of high-profile competences at a larger scale, sharable among different communities
Example of potential fields of application
- Health monitoring data interpretation
- AI-enabled prosthetics
- Better sense of patient data
- Enhanced visualization during surgery
- Improved data collection from medical settings
- AI-enabled artificial organs
Bioimaging
Biomedical imaging techniques serve diagnostic and, occasionally, therapeutic functions, including the acquisition and analysis of data across the electromagnetic spectrum. These modalities play a pivotal role in elucidating the structural and functional intricacies of biological systems, thus facilitating comprehensive understanding and targeted intervention in pathophysiological processes. Its non-invasiveness is crucial in various fields of research, diagnostics, and medical treatments, enabling longitudinal studies over time.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Monitoring of metabolites that can be used as biomarkers for disease identification, progress and response to treatment
- Improved diagnosis and therapy
- Expansion of imaging applications to new areas (e.g., gene expression, functional brain organisation)
Requirements
- Environmental/economic sustainability
Risks
- Risk of stratification and treatment strategies
- Biomarkers validation
Impacts
- More effective and timely diagnosis, and treatment improvement
- Risks vs benefits and costs vs effectiveness trade-off balance
Example of potential fields of application
- Diagnostic
- Intelligent disease mapping
- Computer-aided diagnosis
- Hyperspectral imaging system
- Bio-electromagnetic technologies
- Bioanalytical sensors
- Image-guide surgery
- Magnetic Resonance Imaging
- Tracers and contrast devices
- Biomedical visualisation and data augmentation
Biosensors prompt, Bing Image Creator, DALL-E 2, 26.03.2024
Biosensors
A biosensor is a device measuring biological or chemical reactions through the generation of signals proportionate to the concentration of an analyte participating in the reaction. Biosensors have a very wide range of applications, including but not limited to disease monitoring, pharmaceutical research, pathogenic microorganisms, disease-related biomarkers, and detection of environmental pollutants. Noteworthy subcategories of biosensors encompass wearables, portable devices, and ingestible sensor systems, each tailored to address distinct operational contexts and analytical requirements within the scope of biomedical and environmental sciences.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Accurate, precise, and early diagnosis of emerging diseases
- Collection of clinical parameters without the need for sampling and laboratory test
- Real-time longitudinal data collection
Requirements
- Extreme precision
- Biodegradability
Risks
- The complexity of the technologies required
- Risks related to the individual use of sensors (anxiety, stress and/or hypochondria)
Impacts
- Optimise and speed up both clinical development and longitudinal monitoring of drug efficacy and safety
Example of potential fields of application
- Blood parameters
- Cardiac data
- Metabolic markers
- Sleep quality
- Stress and cognitive decline
Data Science
In the realm of healthcare, substantial datasets replete with pertinent information on patient demographics, treatment modalities, diagnostic outcomes, insurance particulars, and other relevant variables are routinely generated. Within this context, data science assumes a critical role, serving as a foundational framework for facilitating the systematic processing, management, analysis, interpretation, and integration of these datasets. Through the application of rigorous scientific methodologies, advanced data mining techniques, machine learning algorithms, and big data analytics, enable healthcare practitioners to extract actionable insights essential for informed decision-making and optimised patient care delivery.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Data-driven insights can:
- Enhance clinical decision-making
- Advance our understanding of human health and disease
- Personalised medicine
Requirements
- Data availability
- Creation of predictive models
- Integrated data management
- Management structures and criteria for discerning useful data
Risks
- Lack of data, data quality / accuracy
- Bias and errors in data interpretation
- Use of unvalidated and uninterpreted data
- Continuous monitoring
- Excessive self-diagnosis and development of anxiety and hypochondria
- Privacy/cybersecurity
- Lack of skills on the medical side to interpret huge amounts of data
Impacts
- Models’ development
- Improvement of the management, prevention, and prediction based on data
- Healthcare expenditure related to treatment reduction
Example of potential fields of application
- Data cleaning
- Data mining
- Data preparation
- Data analysis
- Practical insights
- Decision-making
Digital Twin Human Body prompt, Bing Image Creator, DALL-E 2, 26.03.2024
Digital Twin/Modelling
A Digital Twin is a set of virtual information constructs that mimics the structure, context, and behaviour of a natural, engineered, or social system. It is dynamically updated with data from its physical twin, has a predictive capability, and informs decisions that realise value. Health digital twins (HDTs) are virtual representations of patients, derived from diverse patient data sources, population data, and real-time updates concerning patient and environmental variables. Through meticulous application, HDTs can simulate random deviations in the digital twin to elucidate anticipated behaviours of the physical counterpart, thereby offering revolutionary implications in precision medicine, clinical trial methodologies, and public health initiatives
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Prediction of disease progression
- Identification of high-risk individuals
- Recommendation of preventive measures
- Monitoring without the implications of presence checks
- Relief for the healthcare system
- Personalised medicine
Requirements
- Creation of predictive models
- Data collection and data availability
- Interoperability of data
- Development of decision models (with probabilistic formulation)
Risks
- Lack of systems allowing for interoperability of data
- Privacy of data and non-medical uses
- Overconfidence in the adoption of the tool within the clinical decision-making process
Impacts
- Reshaping industries to increase efficiency and identify issues
- Treatment of patients as virtualised standalone assets
- Improvement in treatment and diagnostics within hospitals and for individual patients
Example of potential fields of application
- Access to data-driven insights regarding operational strategies, capacity, staffing and care models
- Monitoring
- Development of a unique model for each patient
Biological Spiral prompt, Bing Image Creator, DALL-E 2, 26.03.2024
Drug Delivery Smart Systems
Drug delivery pertains to the methodologies, formulations, technologies, and mechanisms employed to transport therapeutic agents within the body with precision, ensuring their safe and effective delivery to achieve therapeutic objectives. Drug delivery smart systems have demonstrated the capacity to reduce dosing frequency while maintaining therapeutic drug concentrations within targeted anatomical sites over prolonged periods. In this context, these controlled-release systems offer valuable insights and remarkable attributes for attenuating fluctuations in drug concentrations, mitigating drug-related toxicities, and enhancing therapeutic efficacy.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Precise targeting of diseased cells or tissues, minimising harmful effects on healthy cells
- Release at optimal intervals and dosages to maintain therapeutic levels in the body
- Minimisation of off-target effects
- New drug development
- Pharmacokinetics and pharmacodynamics of therapeutic agents improvement
- New modes of administration (transdermal, ocular, nasal)
Requirements
- Good stability of the encapsulated drug
- Biocompatibilty of materials
- Targeted release and delivery
Risks
- Lack of necessary precision
- Risk related with the long-term release of microscopic objects is not yet known
Impacts
- Personalised treatments based on patient data, genetic profiles, and real-time monitoring, through the combination of drug delivery systems with artificial intelligence
- Personalised medicine
Example of potential fields of application
- Medications (e.g., cancer therapy, imaging)
DNA prompt, Bing Image Creator, DALL-E 2, 22.03.2024
Genetic Editing
Genetic editing is a methodology used to enact precise modifications to the DNA of cells or organisms, facilitating targeted additions, deletions, or alterations within the genome. This technique is applicable to somatic cells, where resultant modifications are non-heritable, as well as to germline cells, where edits may be intended for non-reproductive or reproductive purposes. CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) represents a technology leveraged by research scientists to selectively manipulate the genetic material of living organisms. Derived from naturally occurring genome editing mechanisms observed in bacteria, CRISPR has been adapted for use in laboratory settings, offering a powerful means for targeted genetic editing
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Genetic tests could provide a basis for patient information, without deterministically directing the evolution of a person’s health
- Personalised medicine
- Creation of criteria for classifying people’s phenotypes by cross-referencing family heritage, medical records and daily data
- Profiling and prevention of potential risks
Requirements
- Definition of criteria for ethical and legal applications
Risks
- Genetic tests do not always make it possible to determine with absolute certainty whether, when and at what level of severity the person will become ill
- Healthy people may become “pre-patients” for a long time before developing the disease
- Possible bias in classification
Impacts
- Implementation of a series of prevention and early diagnosis measures to prevent or delay the onset of the disease
Example of potential fields of application
- Single gene testing
- DNA sequencing
- DNA shearing
- Exome sequencing
- Genome sequencing
- DNA bar coding
- Large-scale genetic or genomic testing
Implantable Devices prompt, Bing Image Creator, DALL-E 2, 22.03.2024
Implantable Devices
An implantable device is an active medical device intended to be fully or partially inserted into the human body for diagnostic or therapeutic purposes, with the intention of remaining in place. Due to their direct and long-term contact with the body, implantable devices are subject to rigorous standards and requirements to ensure the health and safety of patients. Not only implantable, but also wearable technologies will play a central role. The ubiquity of wearable computing elements is on the rise, albeit hindered by inherent limitations in resolution and bandwidth imposed by current digital technology paradigms. This evolution now intersects with biological frameworks and processes, paving the way for a transformative era in healthcare.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Continuous monitoring of vital parameters
- Citizen-centred care and 6P medicine (Preventive, Predictive, Participatory, Personalized, Psychosocial, Platforms)
- Development of new families of wearable sensors (smart rings, smart glasses)
Requirements
- Massification of wearable and implantable technologies
- Availability and affordability
- Miniaturized batteries
- Data analysis
- Data management models
- Organisation of information campaigns
Risks
- Psychological stress, agitation and/or hypochondria due to control and technological dependence on continuous measurement
- Over-diagnosis and obsession
- Invasiveness of drug delivery for implantable devices
- Possibility of self-determination and the choice to refuse continuous control
- Medicalisation of existence
- Lack of common platforms for data collection and analysis
Impacts
- Quasi-normal lifesyle for individuals requiring medications and continuous monitoring
Example of potential fields of application
- Remote patient monitoring
- Glucose monitoring
- Heart rate monitoring
- Connected contact lenses
- Parkinson’s disease monitoring
- Depression and mood monitoring
- Ingestible sensors connected inhalers
Neurotechnology
Neurotechnology aims to comprehend, enhance, and restore functionality to the nervous system. A neurotech device is a technological device designed to continuously modulate and stimulate neural impulses. While a segment of neurotechnology is dedicated to investigative endeavors, examining mental health disorders or sleep patterns, its use is progressively expanding to modulate brain function or nervous system activity for therapeutic or rehabilitative objectives. Neurological conditions such as Parkinson’s disease, Alzheimer’s disease, major depression, and brain injuries stand to benefit significantly from neurotechnological interventions. The diffusion of such devices holds immense promise in healthcare, particularly as demographic shifts lead to an ageing population and increased life expectancy, consequently heightening the demand for efficacious treatments for neurological ailments.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- Influence on the brain or nervous system for therapeutic or rehabilitation purposes
- Conditions such as Parkinson’s, Alzheimer’s, major depression and brain injuries could find relief thanks to neurotechnology
- Monitoring and treating mental health conditions
- Better pain management
- Cognitive improvement
- Less side effects due to the use/abuse of psychotropic drugs
Requirements
- Creation of neural network schemes
- Access to the operation of the individual
Risks
- Ethical admissibility of heteronomous intervention in the cognitive process
- Unauthorised access to sensitive information stored in the brain
Impacts
- The demand for effective treatments for neurological disorders will increase as the population ages and life expectancy increases and neurotechnology can help to control neurological disorders
Example of potential fields of application
- Brain imaging
- Neurostimulation
- Neuro-devices
Next Generation Vaccines
The foundation of next-generation vaccines lies in the identification of specific proteins derived from the infectious agent, which evokes an immune response akin to that elicited by the intact pathogen. Cancer vaccines represent a subset of immunotherapeutic interventions aimed at instructing the immune system to recognise and eradicate malignant cells by presenting antigens characteristic of cancer cells. Currently, vaccines stand as the foremost efficacious means for averting widespread transmission of infectious diseases. Global endeavors are concentrated on combatting and ultimately eradicating COVID-19, with dedicated efforts toward vaccine development initiated upon its declaration as a pandemic.
Phases
prediction prevention diagnosis treatment rehabilitation
Opportunities
- They are able to self-replicate with a very low dosage and prolonged efficacy
- It teaches the organism to produce more mRNA
- Opportunity to vaccinate new target groups
Requirements
- Pre-clinical testing that includes in vitro studies and studies in animal models through which the mechanism of action is defined
- Education of the population on correct health choices (also about vaccines)
- Establish a legislative framework of new compulsory vaccinations
Risks
- Refusal to undergo vaccinations
Impacts
- Through the massification of people subjected to prevention it enables primary prevention strategies to be undertaken to avoid the spread of contagious diseases
Example of potential fields of application
- Human papillomavirus vaccines
- Antigen vaccines
- DNA vaccines
- Hepatitis B vaccines
- Cancer vaccines
- Dendritic cell vaccines
- Anti-idiotype vaccines