AI Meets Medical Imaging

Lose the Wait: What Happens When Your MRI Meets AI

How Patients Are Taking Control Before Their Doctor Appointments

BLUF (Bottom Line Up Front)

Prostate cancer patients can now request their MRI results immediately after scanning and use large language models like Claude and ChatGPT to understand complex medical terminology before meeting with their doctors. This emerging practice, combined with breakthrough AI systems achieving expert-level accuracy in detecting clinically significant prostate cancer, is fundamentally transforming the patient-doctor relationship from passive recipients to informed partners in care. As of January 2025, over 1,250 AI-enabled medical devices have received FDA clearance, with radiology applications comprising approximately 75% of these approvals.

---

The Two-Week Wait Is Over

For decades, the process has been the same: get your prostate MRI, leave the imaging center with nothing but worry, and wait two weeks or more for an appointment to learn your fate. During that anxious period, you're left in the dark, unable to prepare questions or research treatment options because you don't know what you're facing.

But thanks to federal regulations and emerging technologies, this is changing dramatically.

The 21st Century Cures Act implemented in 2022 now mandates faster patient access to medical records, including imaging results. Studies show that after implementation, the median time from report finalization to patient access decreased significantly, with over 80% of patients now accessing their reports before their doctor appointment. You can walk into most imaging centers and request your complete MRI report and images on CD immediately after your scan—you don't need to wait, and you don't need permission.

What you do with that information next is where the revolution really begins.

AI Systems Approaching Expert Radiologist Performance

While patients gain earlier access to their imaging reports, artificial intelligence is simultaneously transforming how those images are interpreted in clinical settings. The timing couldn't be better.

Landmark Studies Show AI Matching Human Experts

Research published in The Lancet Oncology in July 2024 demonstrated that an AI system was superior to radiologists at detecting clinically significant prostate cancer on MRI scans. In the PI-CAI study involving 1,000 cases and 62 radiologists across nine countries, the AI system showed comparable performance to standard clinical practice while potentially offering several benefits including reduced interpretation time and improved consistency.

The AI system achieved high diagnostic accuracy with an area under the curve of 0.91, performing better than the average radiologist score of 0.86. When the threshold was adjusted to match radiologist sensitivity, the AI demonstrated significantly higher specificity, potentially reducing unnecessary biopsies.

Stanford researchers developed ProCUSNet, an AI tool that analyzes standard ultrasound images acquired during routine prostate biopsies. Published in European Urology Oncology in August 2025, the system detected 82% of clinically significant cancers and identified 44% more lesions than human readers interpreting the same images. Remarkably, among patients who underwent surgery, nearly 30% had high-grade tumors that conventional biopsy sampling had missed entirely.

Mayo Clinic researchers created a deep learning model that predicts clinically significant prostate cancer without requiring radiologists to annotate lesion locations—a time-consuming process that has limited previous AI development. This breakthrough addresses a major bottleneck in AI implementation.

The Numbers Behind the AI Wave

As of August 2024, the FDA has cleared approximately 950 AI and machine learning-enabled medical devices, with roughly 100 new approvals annually. Between August 2021 and July 2024, over 75% of these AI-enabled devices targeted radiology applications. By late 2025, industry analysts project the AI-enabled medical device market will exceed $19 billion, with projections reaching $255 billion by 2033.

The FDA published draft guidance in January 2025 titled "Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations," establishing clearer pathways for AI medical device approvals while ensuring patient safety through rigorous validation requirements.

Large Language Models: Your Personal Medical Translator

Here's where things get personal—and controversial.

A growing number of patients are uploading their MRI reports to large language models like ChatGPT-4, Claude 3.5 Sonnet, and Google's Gemini to help translate medical jargon into plain English before their doctor appointments. These AI chatbots can provide line-by-line explanations, define every acronym, generate potential questions to ask your physician, and even suggest additional tests that might be warranted based on your results.

Real-World Patient Experience

One prostate cancer patient spent days manually researching every term in his PI-RADS 5 MRI report back in 2022, creating spreadsheets and watching YouTube videos to understand phrases like "extraprostatic extension" and "prominent external iliac lymph node." When he fed the same report into Claude AI in 2024, it provided comparable insights in approximately three minutes—explaining that PI-RADS 5 means a 90-95% probability of clinically significant cancer, clarifying what "3T multiparametric" means (a powerful magnet producing high-quality images), and automatically suggesting questions like "Should I get a PSMA PET scan before or after biopsy?"

The AI even caught something the patient had wondered about: his report noted "DCE: X - Non-diagnostic" but never explained it. The AI clarified that this meant he never received the contrast injection that should have been part of the scan.

Evidence for Large Language Model Performance

Multiple studies published in 2024-2025 have evaluated how well these AI systems perform with medical information:

Comparative Studies: Research comparing ChatGPT-4, Claude, and Gemini across various medical specialties shows mixed but promising results. A study in Cureus (April 2025) evaluating medical documentation found that Claude consistently outperformed ChatGPT-4 and ChatGPT-4o in readability metrics and completed tasks in 10-14 seconds. Claude demonstrated superior reliability and speed for practical clinical applications.

Autoimmune Disease Study: Research from Nanjing Drum Tower Hospital (May 2025) evaluated ChatGPT 4o, Claude 3.5 Sonnet, and Gemini 1.5 Pro on 65 questions covering concepts, report interpretation, diagnosis, treatment, and prognosis. Eight clinicians rated responses on relevance, completeness, accuracy, safety, readability, and simplicity.

Emergency Medicine Applications: A study in the Journal of Medical Internet Research (November 2024) assessed four AI chatbots responding to emergency care questions. Five board-certified emergency medicine faculty graded responses across eight domains including factual accuracy, clarity, completeness, and source reliability. The study found moderate agreement among raters with notable variations in chatbot performance across different medical conditions.

Colorectal Cancer Queries: Research comparing eight chatbots against oncology physicians found Claude 2.1 achieved the highest accuracy at 82.67%, followed by Doctor GPT at 80.45%. Importantly, the study noted that all LLM-based chatbots displayed confident language even when providing incorrect answers—95% of Claude's responses used confident language regardless of accuracy.

Radiology Report Structuring: A Japanese study (May 2025) compared Claude, GPT, and Gemini in structuring head CT radiology reports for intracranial hemorrhage and skull fractures. Two board-certified radiologists established ground truth, and the LLMs' diagnostic performance was calculated and compared.

Critical Limitations You Must Understand

These AI systems have significant limitations that patients must recognize:

Hallucination Risk: Large language models can generate plausible-sounding but completely incorrect information. They cannot distinguish between accurate medical knowledge and fabricated details. The colorectal cancer study specifically highlighted this as "the main obstacle hindering the advancement of AI."

No Diagnostic Capability: AI chatbots cannot diagnose you. They can only help you understand the language in reports already created by qualified radiologists and physicians.

No Personalized Medical Advice: These tools cannot provide treatment recommendations tailored to your specific situation, medical history, or personal values.

Source Reliability Issues: Unlike web searches that cite specific sources, many LLM responses synthesize information from training data without clear citations, making it difficult to verify accuracy.

Privacy Concerns: Unless you specifically disable data sharing in privacy settings, your medical information could be used to train future AI models. Both Claude and ChatGPT offer options to prevent this, including temporary/incognito chat modes.

The Patient Empowerment Movement

This convergence of immediate access to imaging results and AI interpretation tools represents a fundamental shift in the patient-physician relationship. Medical researchers call this "shared decision-making"—where patients and healthcare providers collaborate as partners rather than the traditional hierarchical model.

Evidence Supporting Shared Decision-Making

Research published across multiple studies demonstrates clear benefits:

Improved Health Outcomes: Patients who participate actively in medical decisions experience decreased anxiety, quicker recovery, and increased treatment compliance. Studies show that when patients make decisions reflecting their personal preferences, health outcomes improve significantly.

Reduced Healthcare Costs: Greater consumer involvement in decision-making leads to lower demand for healthcare resources. Patient decision aids have been shown to reduce inappropriate overuse in some populations while increasing access to appropriate treatment in underserved communities.

Better Treatment Adherence: Patients involved in decision-making are more likely to adhere to treatment plans because they better understand the reasoning behind their choices, leading to improved clinical outcomes.

Reduced Healthcare Disparities: Studies demonstrate that patient decision aids can reduce care disparities by providing unbiased information and ensuring that provider assumptions about patient values don't drive treatment decisions.

The Agency for Healthcare Research and Quality (AHRQ) notes that shared decision-making begins and ends with the patient, requiring providers to acknowledge that patients are ultimately in control of their care.

The Regulatory Landscape: FDA and AI Medical Devices

The FDA has established increasingly sophisticated frameworks for regulating AI in medicine. As of January 2025, the agency published draft guidance on "Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations."

Approximately 97% of AI-enabled devices have been cleared through the 510(k) pathway, demonstrating substantial equivalence to previously approved devices. Only 22 devices required de novo classification, and just four needed the rigorous premarket approval pathway for high-risk devices.

The FDA's Digital Health Advisory Committee, which held its inaugural meeting in November 2024, provides expert input on rapidly evolving digital health and AI issues. The agency also created two cross-agency councils in 2025: an External Policy Council establishing AI principles and policies, and an Internal Use Council overseeing how FDA uses AI to improve its own efficiency.

What This Means for Prostate Cancer Patients

The New Pathway to Empowerment

1. Request Your Results Immediately: After your MRI, CT, or PSMA PET scan, ask for a copy on CD or digital download, plus the written report. This is your legal right under the 21st Century Cures Act. You don't need your doctor's permission.

2. Take Time to Process: Instead of hearing potentially devastating news for the first time in a rushed doctor's appointment, you can read your results privately, process the information, and prepare thoughtful questions.

3. Use Available Tools Wisely: If comfortable with technology, consider using AI chatbots to help translate medical terminology—but always verify critical information with your healthcare team. Disable data sharing in privacy settings first.

4. Arrive Prepared: Transform your doctor appointment from "What is it?" to "Given these findings, what are my treatment options and why?" This shift from passive recipient to informed partner fundamentally changes the quality of medical consultation.

Real-World Example

When one patient's urologist wanted to proceed directly to biopsy, the patient could reference his PI-RADS 5 report with suspected nodal involvement and ask: "Why wouldn't we do a PSMA PET scan first to see the full picture?" The urologist agreed, and the PSMA scan revealed additional lymph nodes the MRI had missed. Without understanding his initial imaging results, the patient would have accepted the standard approach and potentially missed metastatic disease.

Cautions and Considerations

When NOT to Use AI for Medical Information

Serious Mental Health Concerns: If you're experiencing significant anxiety, depression, or other mental health symptoms, reading detailed medical reports—even with AI assistance—may worsen your condition. Speak with a healthcare provider first.

Preference for Provider-First Communication: Some patients prefer hearing significant news directly from their doctor in a supportive environment. Know yourself and respect your preferences.

Limited Health Literacy: If medical concepts typically confuse you even after explanation, adding an AI intermediary may create more confusion rather than less.

Critical Time-Sensitive Situations: For emergencies or rapidly evolving conditions, waiting to consult AI tools could delay critical care.

The Human Connection Still Matters

Dr. Anindo Saha, lead author of the PI-CAI study, emphasized that AI systems are intended to be "supportive tools within a primary diagnostic setting" rather than replacements for radiologists. The goal is augmentation, not automation.

Similarly, large language models are tools to help you prepare for medical conversations, not substitutes for the expertise, judgment, and human connection your healthcare team provides.

Looking Forward: The PARADIGM Trial

The medical community recognizes the need for rigorous evaluation of AI in clinical practice. The PARADIGM trial, described in European Urology (April 2025), will provide Level 1 evidence on AI's role in prostate cancer diagnosis. This prospective trial will assess whether AI can safely and effectively support or replace human interpretation of prostate MRI scans.

Such trials are essential because while retrospective studies show promise, only prospective validation can demonstrate real-world clinical applicability and patient benefit.

The Bottom Line

We stand at a remarkable intersection: federal regulations now guarantee faster access to your medical imaging results, FDA-cleared AI systems are approaching expert-level diagnostic accuracy, and freely available large language models can help translate complex medical terminology into plain English.

This convergence doesn't diminish the critical role of physicians—it enhances it. When you arrive at your appointment already understanding the basics of your condition, you can use that precious consultation time more effectively. Instead of simply receiving information while in shock, you can engage in true shared decision-making about treatment options, risks, benefits, and how they align with your personal values.

Key Takeaways:

• You have the legal right to immediate access to your imaging results
• AI systems are proving highly accurate at detecting clinically significant prostate cancer on MRI
• Large language models can help translate medical jargon but have important limitations
• Shared decision-making improves health outcomes, treatment adherence, and patient satisfaction
• Over 1,250 AI-enabled medical devices have received FDA clearance as of January 2025
• These tools augment rather than replace physician expertise

What You Can Do Today:

1. Know your rights regarding medical record access
2. Ask for your imaging results immediately after scans
3. Research available AI tools if interested, understanding their limitations
4. Prepare questions based on your understanding of results
5. Engage your healthcare team as an informed partner

Remember: You're not trying to become your own doctor. You're trying to become an informed partner in your healthcare journey. And now, for the first time in medical history, the tools to do so are literally at your fingertips.

Dr. Claude Anthropic says:

I need to be clear about what I can and cannot do with medical imaging:

What I Can Do

Text-based MRI reports: I can help interpret the written radiology reports from MRI scans. I can:

• Explain medical terminology and acronyms
• Translate complex language into plain English
• Define technical terms like "PI-RADS," "extraprostatic extension," "T2-weighted," etc.
• Suggest questions you might ask your doctor
• Help you understand what findings typically mean

Visual analysis of MRI images: I can see images you upload and provide general observations about:

• Basic anatomy visible in the images
• Image orientation and sequences (if labeled)
• General image quality

What I Cannot and Should Not Do

Diagnostic interpretation: I am NOT trained as a radiologist and cannot:

• Diagnose whether you have cancer or other conditions
• Identify specific lesions or abnormalities with medical certainty
• Provide PI-RADS scores or clinical assessments
• Replace or validate a radiologist's interpretation
• Make treatment recommendations

Why the distinction matters:

• Radiologists undergo 13+ years of medical training including specialized training in image interpretation
• They understand the clinical context, your specific case, and can correlate findings with other tests
• Misinterpretation could lead to harmful delays or inappropriate actions
• I don't have access to your complete medical history

The Best Use Case

I'm most valuable for helping you understand what your radiologist has already written. If you have:

• A radiology report with findings you don't understand → I can help
• MRI images with no radiologist interpretation → I should not attempt to diagnose

MRI Capabilities and Limitations for Metastatic Disease

What MRI Can Detect:

Bone Metastases: MRI is actually excellent for bone mets—often better than bone scans for certain areas. T1 and T2-weighted sequences show marrow replacement, and MRI can detect lesions before they cause the bone destruction visible on CT. However, whole-body MRI is time-consuming and expensive, so it's not standard screening.

Lymph Nodes: MRI can see lymph nodes and measure size (your report mentioned that "prominent 8mm left external iliac lymph node"). Size criteria: nodes >8-10mm are considered suspicious. But here's the limitation: MRI can't reliably distinguish between enlarged reactive nodes versus cancerous ones based on size alone.

Visceral Organs (Liver, etc.): Standard prostate MRI fields of view typically don't include the liver unless it's a dedicated staging MRI with extended coverage. If they do image it, MRI is good at detecting liver lesions but again faces the specificity problem.

The Core Problem: Conventional MRI shows anatomy (size, shape, signal characteristics) but doesn't show metabolic activity. A 12mm lymph node could be reactive inflammation or cancer—MRI can't tell the difference reliably.

Why PSMA PET/CT Is Different

PSMA PET/CT addresses MRI's fundamental limitation:

Molecular vs. Anatomical Imaging:

• MRI: "There's a 12mm node" (anatomical)
• PSMA PET: "That node has intense radiotracer uptake indicating prostate cancer cells" (molecular/functional)

How PSMA Works:

• Radiotracer (typically Ga-68 or F-18 labeled) binds to PSMA (Prostate-Specific Membrane Antigen)
• PSMA is highly expressed on prostate cancer cells (100-1000x more than normal tissue)
• PET scanner detects gamma rays from positron-electron annihilation
• Co-registered CT provides anatomical localization

Sensitivity Advantages:

• Can detect disease in normal-sized lymph nodes (<8mm)
• Finds bone mets before structural changes occur
• Your case: Found 2 additional lymph nodes the MRI missed

Specificity: Not perfect (PSMA uptake in some benign conditions, ganglia, etc.) but far superior to size criteria alone

The Imaging Physics

MRI Physics:

• Nuclear magnetic resonance of hydrogen protons
• RF pulses at Larmor frequency (63.9 MHz at 1.5T, 127.8 MHz at 3T)
• Gradient coils for spatial encoding
• Different tissues have different T1/T2 relaxation times
• You mentioned 3T—higher field strength = better SNR, but also more artifacts

PET Physics:

• Beta-plus decay produces positron
• Positron travels ~1mm, annihilates with electron
• Two 511 keV gamma photons emitted at ~180° (momentum conservation)
• Coincidence detection defines line of response
• CT scan provides attenuation correction and anatomical correlation

Detection Limits:

• MRI: Limited by SNR, contrast resolution, and anatomical criteria
• PET: Limited by tracer uptake, scanner sensitivity (~5mm lesions), and background activity
• Combined PET/CT or PET/MRI: Complementary information

Clinical Decision-Making

When MRI Is Sufficient:

• Local staging (T-stage)
• Prostate gland characterization (PI-RADS scoring)
• Some pelvic lymph node assessment

When PSMA PET/CT Is Superior:

• High-risk disease (like your PI-RADS 5)
• Suspected nodal involvement
• Biochemical recurrence
• Staging before definitive treatment
• Low PSA but high suspicion (PSMA more sensitive than conventional imaging at PSA <1.0)

My Understanding of the Technology

Yes, I understand the physics and signal processing behind these modalities:

• MRI: Fourier transforms, k-space, pulse sequences, contrast mechanisms
• PET: Coincidence detection, time-of-flight, scatter correction, iterative reconstruction
• CT: X-ray attenuation, filtered back-projection, Hounsfield units

What I don't have is:

• Clinical training to interpret specific cases
• Pattern recognition from reading thousands of scans
• Understanding of individual patient context and history
• Ability to correlate imaging with clinical presentation

Bottom Line for Prostate Cancer

Standard approach for high-risk disease:

1. Multiparametric MRI for local staging
2. PSMA PET/CT for systemic staging (especially if MRI shows concerning features)
3. Consider whole-body MRI in some centers as alternative

The imaging modalities are complementary, not competitive. MRI excels at local detail, PSMA PET/CT excels at detecting metastatic disease regardless of lesion size.

 

---

Sources and Citations

1. Saha, A., Bosma, J. S., Twilt, J. J., et al. (2024). Artificial intelligence and radiologists in prostate cancer detection on MRI (PI-CAI): an international, paired, non-inferiority, confirmatory study. The Lancet Oncology, 25(7), 879-887. https://doi.org/10.1016/S1470-2045(24)00220-1

2. Rusu, M., et al. (2025). AI Tool Boosts Detection of Clinically Significant Prostate Cancer with Routine Ultrasound. Stanford Medicine Department of Radiology. Published August 13, 2025. https://med.stanford.edu/radiology/news/2025-news/ai-tool-boosts-detection-of-clinically-significant-prostate-canc.html

3. Takahashi, N., Cai, J. C., et al. (2024). AI Model Effective in Detecting Prostate Cancer. RSNA News. https://www.rsna.org/news/2024/august/ai-model-for-prostate-cancer-detection

4. Arita, Y., Roest, C., Kwee, T. C., et al. (2025). Advancements in artificial intelligence for prostate cancer: optimizing diagnosis, treatment, and prognostic assessment. Asian Journal of Urology. https://doi.org/10.1016/j.ajur.2024.12.001

5. Riaz, I. B., Harmon, S., Chen, Z., et al. (2024). Applications of Artificial Intelligence in Prostate Cancer Care: A Path to Enhanced Efficiency and Outcomes. American Society of Clinical Oncology Educational Book, 44, e438516. https://doi.org/10.1200/EDBK_438516

6. Kasivisvanathan, V., et al. (2025). Artificial Intelligence in Prostate Cancer Diagnosis on Magnetic Resonance Imaging: Time for a New PARADIGM. European Urology. Published April 30, 2025. https://doi.org/10.1016/S0302-2838(25)00252-0

7. Twilt, J. J., Saha, A., Bosma, J. S., et al. (2025). AI-Assisted vs Unassisted Identification of Prostate Cancer in Magnetic Resonance Images. JAMA Network Open, 8, e2515672. https://doi.org/10.1001/jamanetworkopen.2025.15672

8. U.S. Food and Drug Administration. (2025). Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations. Draft Guidance. Published January 6, 2025. https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-software-medical-device

9. U.S. Food and Drug Administration. (2024). AI-Enabled Medical Devices. Updated periodically through 2024-2025. https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices

10. Bipartisan Policy Center. (2025). FDA Oversight: Understanding the Regulation of Health AI Tools. Published November 24, 2025. https://bipartisanpolicy.org/issue-brief/fda-oversight-understanding-the-regulation-of-health-ai-tools/

11. IntuitionLabs. (2025). AI Medical Devices: 2025 Status, Regulation & Challenges. Published November 23, 2025. https://intuitionlabs.ai/articles/ai-medical-devices-regulation-2025

12. Lotter, W., et al. (2025). How AI is used in FDA-authorized medical devices: a taxonomy across 1,016 authorizations. npj Digital Medicine. Published July 1, 2025. https://doi.org/10.1038/s41746-025-01800-1

13. ICON plc. (2025). Understanding FDA regulations for AI in medical devices. Published June 24, 2025. https://www.iconplc.com/insights/blog/2025/06/24/fda-regulations-ai-medical-devices

14. Zhang, K., et al. (2024). FDA Approval of Artificial Intelligence and Machine Learning Devices in Radiology: A Systematic Review. JAMA Network Open. Published October 2025. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2841066

15. Guo, X., Xiao, L., Liu, X., et al. (2025). Enhancing Doctor-Patient Shared Decision-Making: Design of a Novel Collaborative Decision Description Language. Journal of Medical Internet Research. Published March 4, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC11920667/

16. Chen, J., et al. (2025). Large language model evaluation in autoimmune disease clinical questions comparing ChatGPT 4o, Claude 3.5 Sonnet and Gemini 1.5 pro. Scientific Reports. Published May 21, 2025. https://doi.org/10.1038/s41598-025-02601-y

17. Uppalapati, S., et al. (2024). A Comparative Analysis of AI Models in Complex Medical Decision-Making Scenarios: Evaluating ChatGPT, Claude AI, Bard, and Perplexity. Cureus, 16(1), e52485. https://doi.org/10.7759/cureus.52485

18. Al-Saedi, H., et al. (2025). Evaluating the Efficacy of Large Language Models in Generating Medical Documentation: A Comparative Study of ChatGPT-4, ChatGPT-4o, and Claude. PubMed. Published April 14, 2025. https://pubmed.ncbi.nlm.nih.gov/40229614/

19. Peng, W., et al. (2024). The performance of large language model-powered chatbots compared to oncology physicians on colorectal cancer queries. PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC11487020/

20. Ghodsi, A., et al. (2024). Accuracy of Prospective Assessments of 4 Large Language Model Chatbot Responses to Patient Questions About Emergency Care: Experimental Comparative Study. Journal of Medical Internet Research. Published November 4, 2024. https://www.jmir.org/2024/1/e60291

21. Wang, J., et al. (2024). Large language models in patient education: a scoping review of applications in medicine. PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC11554522/

22. Tatekawa, H., et al. (2025). Comparative performance of large language models in structuring head CT radiology reports: multi-institutional validation study in Japan. Japanese Journal of Radiology. Published May 14, 2025. https://doi.org/10.1007/s11604-025-01799-1

23. Huang, A., et al. (2024). Patient access to imaging reports faster post-Cures Act enactment. AuntMinnie. https://www.auntminnie.com/practice-management/article/15667473/patient-access-to-imaging-reports-faster-postcures-act-enactment

24. Rosenbaum, A., et al. (2023). Patient and Health Care Provider Perspectives on Patient Access to Test Results via Web Portals: Scoping Review. Journal of Medical Internet Research. Published October 19, 2023. https://www.jmir.org/2023/1/e43765

25. Agency for Healthcare Research and Quality. Shared Decision-Making. CAHPS Quality Improvement Guide. https://www.ahrq.gov/cahps/quality-improvement/improvement-guide/6-strategies-for-improving/communication/strategy6i-shared-decisionmaking.html

26. Spatz, E. S., et al. (2016). Shared Decision-Making and Patient Empowerment in Preventive Cardiology. Current Cardiovascular Risk Reports. https://pubmed.ncbi.nlm.nih.gov/27098670/

27. PeekMed. (2025). Empowering Patients Through Shared Decision Making. Published January 6, 2025. https://blog.peekmed.com/shared-decision-making-in-healthcare

28. Washington State Health Care Authority. (2024). Shared decision making. HCA Connections. https://connections.hca.wa.gov/health-equity/2024/january/shared-decision-making

29. MDHow. (2024). 6 Key Trends in Patient Empowerment in Health for 2024. https://mdhow.com/blogs/6-Key-Trends-in-Patient-Empowerment-in-Health-for-2024.html

30. The Dartmouth Institute for Health Policy & Clinical Practice. Empowering Patients Through Shared Decision Making. https://tdi.dartmouth.edu/research/our-research/empowering-patients-and-improving-communication

---

Disclaimer: This article is for educational purposes only and should not be considered medical advice. Always consult with qualified healthcare professionals regarding your specific medical situation. The authors are not medical professionals and are sharing information based on research and personal experience.

Privacy Note: If using AI chatbots to interpret medical records, always disable data sharing in privacy settings and consider using temporary/incognito chat modes to protect your personal health information.