As artificial intelligence (AI) becomes a part of healthcare, clinicians from multiple specialties are calling attention to the need for ethical guidance and guardrails to protect everyone.1,2
“Clinicians need to be proactive to ensure these systems elevate the care we provide rather than undermining it,” says Mina S. Makary, MD, director of the interventional radiology inpatient service at The Ohio State University Wexner Medical Center.
Rather than reacting to these risks, clinicians implementing AI algorithms should ensure policies and safeguards are in place and routinely assess them. “They can also be advocates for such protections through interactions with industries and organizations, as well as mentorship of trainees,” adds Makary, co-author of a paper on ethics of AI in interventional radiology.3
Pegah Khosravi, PhD, is an assistant professor in the department of biological sciences at New York City College of Technology. Khosravi, an AI scientist who has closely collaborated with radiologists and medical experts, was the lead author of a recently published paper on ethical issues in neuroradiology.4 Medical Ethics Advisor (MEA) interviewed Khosravi about that paper and other related topics. (Editor’s Note: This transcript has been lightly edited for clarity.)
MEA: What are some central ethical concerns regarding AI in neuroradiology?
Khosravi: As we bear witness to the swift strides of AI technologies, our duty becomes clear: to meticulously dissect the ethical implications, navigational quandaries, and latent perils embedded within the realm of AI in medical imaging.
MEA: How can clinicians in hospitals and research settings help?
Khosravi: Amid the burgeoning landscape of AI-assisted medical imaging, clinicians stand as pivotal figures. The ethical considerations draped around the use of AI-augmented diagnostic tools hold paramount importance. Clinicians bear the solemn responsibility of traversing these intricacies by preserving their clinical acumen while harnessing AI as a supplementary adjunct.
MEA: What are ethical approaches for data management?
Khosravi: Clinicians in hospital and research settings wield the potential to catalyze profound transformation by adopting a foresighted stance toward data management. Recognizing data ownership as a cardinal precept, hospitals can affirm that patient data are fundamentally the domain of the individuals it represents.
When data harvested for the betterment of patient well-being are harnessed ethically for the advancement of science, the scope of benefit broadens beyond the individual to encompass the scientific collective and society at large.
Rather than discarding seemingly tangential data, the emphasis lies on secure and systematic preservation. An example could be the demographic information of patients. While this information might not directly relate to a specific medical diagnosis or treatment, it plays a crucial role in identifying potential biases or disparities in AI algorithms. Demographic data, such as age, gender, ethnicity, and socioeconomic background, can reveal important insights into how certain groups might be disproportionately affected by the AI’s predictions. By preserving and analyzing these data, we can better understand and address any unintended biases in AI systems and ensure equitable healthcare outcomes for all patients.
In the context of neuroradiology, another example of seemingly tangential data that should not be discarded, even if it appears to have low quality or limited utility with current technology, could be the raw or minimally processed data obtained from advanced imaging techniques. For instance, consider functional MRI (fMRI) data that capture brain activity by measuring changes in blood flow.
Advancements in AI and data processing techniques might allow for the extraction of valuable insights from seemingly noisy or low-quality fMRI signals. Discarding such data prematurely, due to its current limitations, could lead to missed opportunities for future breakthroughs.
MEA: How can researchers and clinicians safeguard patient data?
Khosravi: Researchers are responsible for developing and implementing data management protocols that prioritize data security, integrity, and accessibility. They establish robust storage systems, encryption measures, and access controls to prevent unauthorized use or breaches.
Clinicians, on the other hand, contribute by providing valuable insights into the clinical context and relevance of the data. They help identify which data points are critical for medical decision-making and research, while also advocating for patient privacy and consent.
MEA: What are concerns about professional autonomy?
Khosravi: The role of AI in clinical decision-making raises concerns about the autonomy of healthcare professionals. Ethical considerations involve determining the appropriate balance between AI assistance and clinical judgment.
MEA: What about informed consent?
Khosravi: Patients have the right to know if AI technologies are being used to aid in their diagnosis or treatment. Ensuring that patients provide informed consent for AI involvement is an important ethical consideration.
REFERENCES
1. Cacciamani GE, Chen A, Gill IS, Hung AJ. Artificial intelligence and urology: Ethical considerations for urologists and patients. Nat Rev Urol 2023; Jul 31. doi: 10.1038/s41585-023-00796-1. [Online ahead of print].
2. Kaya Bicer E, Fangerau H, Sur H. Artificial intelligence use in orthopedics: An ethical point of view. EFORT Open Rev 2023;8:592-596.
3. Rockwell HD, Cyphers ED, Makary MS, Keller EJ. Ethical considerations for artificial intelligence in interventional radiology: Balancing innovation and patient care. Semin Intervent Radiol 2023;40:323-326.
4. Khosravi P, Schweitzer M. Artificial intelligence in neuroradiology: A scoping review of some ethical challenges. Front Radiol 2023;3:1149461.
Rather than reacting to associated risks, clinicians implementing AI algorithms should ensure policies and safeguards are in place and routinely assess them. They also can be advocates for such protections through interactions with industries and organizations, as well as mentorship of trainees.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content