Introductions

The use of Artificial Intelligence (AI) in nuclear pharmacy is a relatively new concept, but one that has the potential to revolutionize the field. AI has the potential to improve the accuracy and efficiency of nuclear pharmacy operations, as well as to reduce the risk of errors and improve patient safety. In this paper, we will discuss the potential applications of AI in nuclear pharmacy, the challenges associated with its implementation, and the potential benefits it could bring to the field. We will also discuss the ethical considerations that must be taken into account when using AI in nuclear pharmacy.

Nuclear pharmacy is a specialized field of pharmacy that deals with the preparation, dispensing, and use of radioactive materials for medical purposes. Nuclear pharmacists are responsible for ensuring that the correct dose of radioactive material is administered to the patient, as well as for monitoring the patient’s response to the treatment. The use of AI in nuclear pharmacy could help to improve the accuracy and efficiency of these processes, as well as to reduce the risk of errors and improve patient safety.

One potential application of AI in nuclear pharmacy is the use of computer-aided diagnosis (CAD) systems. These systems use AI algorithms to analyse patient data and provide a diagnosis. This could help to reduce the time and effort required to diagnose a patient, as well as to reduce the risk of errors. Additionally, AI could be used to automate the process of calculating the correct dose of radioactive material for a patient, which could help to reduce the risk of errors and improve patient safety.

Another potential application of AI in nuclear pharmacy is the use of robotic systems for the preparation and dispensing of radioactive materials. These systems could be used to automate the process of preparing and dispensing radioactive materials, which could help to reduce the risk of errors and improve patient safety. Additionally, robotic systems could be used to monitor the patient’s response to the treatment, which could help to reduce the risk of errors and improve patient safety.

In addition to the potential applications of AI in nuclear pharmacy, there are also a number of challenges associated with its implementation. One of the main challenges is the need for accurate data. AI algorithms require large amounts of accurate data in order to be effective, and this data must be collected and stored in a secure manner. Additionally, AI algorithms must be regularly updated in order to remain effective, which can be a time-consuming and expensive process. Finally, AI algorithms must be carefully monitored in order to ensure that they are not making mistakes or introducing bias into the system.

AI (Artificial Intelligence) has the potential to revolutionize the healthcare industry, including nuclear pharmacy. Nuclear pharmacy is a specialized field of pharmacy that deals with the preparation and dispensing of radioactive drugs for diagnostic and therapeutic purposes. The use of AI in nuclear pharmacy can improve patient outcomes, increase efficiency, and reduce costs.

Nuclear medicine is a branch of medical imaging that uses small amounts of radioactive materials, called radiopharmaceuticals, to diagnose and treat various diseases. Radiopharmaceuticals are typically administered by injection, inhalation, or ingestion and then detected by a special camera or scanner. The use of radioactive materials requires specialized training, equipment, and facilities to ensure safety and efficacy.

AI can help nuclear pharmacy in several ways. One of the primary uses of AI in nuclear pharmacy is to improve the accuracy and efficiency of radiopharmaceutical preparation. Radiopharmaceuticals are typically prepared on-demand and have a short half-life, which means that they decay rapidly. Accurate and timely preparation is crucial to ensure that patients receive the correct dose at the right time. AI can help automate and streamline the preparation process, reducing the risk of errors and delays.

Another way AI can help nuclear pharmacy is by improving patient outcomes. AI can help identify patients who are most likely to benefit from nuclear medicine procedures and optimize the dose of radiopharmaceuticals to reduce the risk of side effects. AI can also help identify patients who may be at higher risk of complications or adverse reactions, allowing healthcare providers to adjust treatment plans accordingly.

AI can also help reduce costs associated with nuclear medicine procedures. By optimizing the use of radiopharmaceuticals and reducing waste, AI can help lower the overall cost of nuclear medicine procedures. Additionally, by automating and streamlining the preparation process, AI can help reduce the need for highly trained and specialized staff, further reducing costs.

However, the use of AI in nuclear pharmacy also poses some challenges. One of the primary challenges is ensuring the safety and efficacy of radiopharmaceuticals. AI systems must be carefully validated and tested to ensure that they produce accurate and reliable results. Additionally, AI systems must be designed to account for the unique characteristics of radiopharmaceuticals, such as their short half-life and potential for radiation exposure.

Another challenge is ensuring the privacy and security of patient data. AI systems require large amounts of data to train and improve their algorithms. This data often includes sensitive patient information, such as medical histories and imaging results. Healthcare providers must take steps to ensure that patient data is protected and used only for its intended purposes.

Despite these challenges, the use of AI in nuclear pharmacy has the potential to transform the field and improve patient outcomes. As AI technology continues to evolve, it is likely that we will see more advanced and sophisticated applications in nuclear pharmacy and other areas of healthcare.

OPPORTUNITIES

Quantitative Imaging and Process Improvement

Nuclear medicine is evolving toward even better image quality and more accurate and precise quantification in the precision medicine era, most recently in the paradigm of theragnostic. 

Diagnostic Imaging

AI techniques in the “patient-to-image subdomain” improve acquisition and models in the “image-to-patient subdomain” enable improved decision-making for interventions on patients Image generation considerations are elaborated in supplement opportunities

However, examples include improved image reconstruction from raw data (list mode, sinogram), data corrections including attenuation, scatter and motion, and post-reconstruction image enhancement, among others. These enhancements could impact PET and SPECT in clinical use today. Multi-time-point acquisitions and PET/MR may see improved feasibility. Specific opportunities in image analysis are elaborated on in the supplement. A few examples include image registration, organ and lesion segmentation, biomarker measurements and multi-omics integration, and kinetic modelling.

Emerging Nuclear Imaging Approaches

New developments are also emerging such as Total Body PET (TB-PET) (18) which presents unique data and computational challenges. Another potential use of AI is to separate multi-channel data from single-session multi-isotope dynamic PET imaging. This pragmatic advancement could be valuable to extract greater phenotyping information in the evaluation of tumour heterogeneity.

Radiopharmaceutical Therapies (RPTs)

There are several areas where AI is expected to significantly impact RPTs:      

AI-Driven Theragnostic Drug Discovery and Labelling.

The use of AI for molecular discovery has been explored to select the most promising leads to design suitable theragnostic for the target in question. For example, machine learning models could be trained using parameters from past theragnostic successes and failures (e.g., logP, Kd, BP) to establish which best predict a given outcome (e.g., specific binding, blood-brain-barrier penetration, tumour-to-muscle ratio). New AI approaches are revolutionizing our understanding of protein-ligand interactions.

Precision Dosimetry.

The field of radiopharmaceutical dosimetry is progressing rapidly. After the administration of radiopharmaceuticals, dynamical and complex pharmacokinetics result in time-variable biodistribution. Interaction of ionizing particles arising from the injected agent with the target and normal tissue results in energy deposition. Quantification of this deposited energy and its biological effect is the essence of dosimetry with opportunities to link the deposited energy to its biological effect on diseased and normal tissues.

Predictive Dosimetry and Digital Twins.

Existing models can perform dosimetry before (e.g., I-131 MIBG) or the following treatment. Personalized RPTs require predictive dosimetry for optimal dose prescription where AI can play a role. Pre-therapy (static or dynamic) PET scans could model radiopharmaceutical pharmacokinetics and absorbed doses in tumours and normal organs. Furthermore, it is possible to additionally utilize intra-therapy scans (e.g., single-timepoint SPECT in the first cycle of RPTs) to better anticipate and adjust doses in subsequent cycles.

Clinical Workflow:

Increase Throughput While Maintaining Excellence  AI may impact operations in nuclear medicine, such as patient scheduling and resource utilization predictive maintenance of devices to minimize unexpected downtimes, monitoring quality control measurement results to discover hidden patterns and indicate the potential for improvement, and monitoring the performance of devices in real-time to capture errors and detect aberrancies.

AI ECOSYSTEM

Actualization of Opportunities and Contextualization of Challenges

While early nuclear medicine AI systems are already emerging, many opportunities remain in which the continuous propagation of AI technology could augment our precision patient care and practice efficiencies. The environment where AI development, evaluation n, implementation, and dissemination occurs needs a sustainable ecosystem to enable progress, while appropriately mitigating concerns of stakeholders. The total life development and cycle of AI systems, from concept to appropriation of training d prototyping, production testing, validation and evaluation, implementation/deployment, and post– data, model deployment surveillance, occurs within a framework which we call the enhancing the tr “AI Ecosystem”

CHALLENGES FOR DEVELOPMENT, VALIDATION, DEPLOYMENT, AND IMPLEMENTATION Development of AI Applications/Medical Devices

Five challenges that should be addressed include data, optimal network architecture, measurement and communication of uncertainty, identification of clinically impactful use cases, and improvements in team science approaches (supp Development Challenges).

Evaluation (Verification of Performance)

Theories on appropriate evaluation of AI software are a broad and active area of current investigation. Establishing clear and consistent guidelines for performance profiling remains challenging. Most current verification studies evaluate AI methods based on metrics that are agnostic to performance in clinical tasks. While such evaluation may help demonstrate promise, there is an important need for further testing on specific clinical tasks before the algorithms can be implemented. Failure mode profiling is among the most important challenges (supp Evaluation Challenges). 

Ethical, Regulatory, and Legal Ambiguities

Major ethical concerns include informed consent for data usage, replication of historical bias and unfairness embedded in training data, unintended consequences of AI device agency, inherent opaqueness of some algorithms, concerns about the impact of AI on healthcare disparities, and trustworthiness (supp Ethical Challenges). AI in nuclear medicine has limited legal precedent.

Implementation of Clinical AI Solutions & Post-Deployment Monitoring

The lack of an AI-Platform integrating AI applications in nuclear medicine workflow in among the most critical challenges of implementation Barriers of dissemination can be categorized at an individual level (healthcare providers), at the institutional level (organization culture), and at the societal level Deployment is not the end of the implementation process. (Supp Implementation Challenges).

TRUST AND TRUSTWORTHINESS

In medicine, trust is the essence, not a pleasance.

Successful solutions to the abovementioned challenges for the sustainability of AI ecosystems in medicine. necessary but not sufficiently developed and validated AI devices with supportive regulatory context, appropriate reimbursement and successful primary implementation may still fail if physicians, patients, and society lose trust due to lack of transparency and other critical elements of trustworthiness such as perceived inattention to health disparity or racial injustice. In a recent survey, Martinho et al. found significant perceived mistrust among healthcare providers in regard to AI systems and the AI industry while realizing the importance and benefits of this new technology. Responders also emphasized the importance of ethical utilization, and the need for physicians in the loop interactions with AI systems, among other factors. There is a need for a comprehensive analysis of the AI ecosystem to define and clarify the core elements of trustworthiness in order to realize the benefits of AI in clinical practice.

STRATEGIES FOR SUCCESS

Part 1: SNMMI Initiatives In July 2022 SNMMI created the AI Task Force to strategically assess the emergence of AI in nuclear medicine (supp SNMMI initiatives). Areas of important focus were designated working groups, such as the AI & Dosimetry working group for predictive dosimetry and treatment planning.

PART 2: SNMMI Action Plan The Task Force recommends the establishment of the AI Centre of Excellence (AICE) to facilitate a sustainable AI ecosystem (supp SNMMI action plan). A Nuclear Medicine Imaging Archive (NMIA) will address the need for meaningful data access. The Trustworthy AI in Medicine and Society Coalition (TAIMS coalition) will address the need for an AI Bill of Rights.

Part 3: SNMMI Recommendations for the future are also provided in the supplement (supp SNMMI Recommendations). 

In conclusion,

AI has the potential to revolutionize the field of nuclear pharmacy. AI algorithms could be used to improve the accuracy and efficiency of nuclear pharmacy operations, as well as to reduce the risk of errors and improve patient safety. However, there are a number of challenges associated with the implementation of AI in nuclear pharmacy, as well as a number of ethical considerations that must be taken into account. Despite these challenges, the potential benefits of AI in nuclear pharmacy make it an exciting and promising field of research.

We can all benefit from efforts to ensure fairness, inclusion, and lack of bias in the entire lifecycle of AI algorithms in different settings. There are three levels of facilitation that can support and enable the appropriate environment for trustworthy AI. First, our community must establish guidelines, such as those referenced in this article, to promote the facilitate trustworthy AI through the natural development of trustworthy AI. Second, we can AI Centre of Excellence (AICE ). Third, we can make trustworthy AI occur through active engagement and communicative actions. By encouraging the establishment of trustworthy AI in nuclear medicine, SNMMI aims to decrease health disparity, increase health system efficiency, and contribute to the improved overall health of society using AI applications in the practice of nuclear medicine.