NemoClaw Knowledge Wiki

Module 2 Ethics Governance V4 AI Cleaned

23 min read

🗂️ Philosophy, Ethics & Religion · View mindmap

preface_schema: ‘1.0’ title: ‘scenarios in daily practice.’ source_type: ‘Academic’ publisher: ‘Ieee’ publishing_date: ‘Unknown’ authors: [‘Privacy Law’] available_at: ‘Unknown’ credibility_tier_value: ‘4’ credibility_tier_key: ‘institutional’ credibility_tier_label: ‘Institutional’ credibility: ‘Draft Institutional Report’ keywords: [‘data’, ‘information’, ‘health’, ‘digital’, ‘privacy’, ‘your’, ‘professional’, ‘apps’] abstract: ‘concepts, but practical tools you need for everyday allied health practice in the digital age. To Do This Week: 1. Formative Quiz (Ungraded): Complete the 15-question multiple-choice quiz covering the APPs, the NDB scheme, and the core principles of AI ethics. 2. Reflective Scenario Report (Ungraded): Post your 300-word response to the Ethical Dilemma in Data Sharing exercise in the discussion forum. 3. Discussion Post (Optional): Respond to at least one peer”s post on the scenario exercise. Do you agree with their approach? Can you offer a different perspective? 4. Review the Readings & Resources below to consolidate your learning. Key Readings & Resources Core Readings • Office of the Australian Information Commissioner (OAIC). (2022). Australian Privacy Principles quick reference. (This is the foundational guide to the APPs). • Australian Government. (2019). Australia”s Artificial Intelligence Ethics Framework. (Outlines the key principles for ethical AI in Australia). • Victorian Department of Health. (2021). Digital health capability framework for allied health professionals. (An excellent example of how these principles are applied in practice). • Australian Digital Health Agency. Website section on Data breaches and the My Health Record system. • World Health Organization. (2021). Ethics and governance of artificial intelligence for health: WHO guidance. ---‘

Page 1

. Website section on Data breaches and the My Health Record system. • World Health Organization. (2021). Ethics and governance of artificial intelligence for health: WHO guidance. ---‘

Page 1

Week 2: Data Governance, Ethics, and Legislation Total Estimated Time: 8–10 hours Module Welcome: Your Duty of Care in a Digital World Welcome to Week 2. In our first week, we explored the “what” of digital health—the core technologies transforming allied health practice. Now, we turn to the “how”: how do we use these powerful tools safely, ethically, and legally? Digital health is built on a foundation of trust. Every time a client shares their information, they are trusting you to protect it. In Australia, this duty of care is not just an ethical obligation but a legal one, governed by a complex landscape of legislation and professional standards. This week, we will equip you with the knowledge to navigate this landscape confidently. We will demystify the key laws, explore the ethical frontiers of Artificial Intelligence (AI), and provide practical skills to manage your professional digital identity. This week builds directly on your foundational knowledge and is essential for safe practice. By the end of this week, you will have a robust framework for making sound decisions about data, privacy, and technology in your everyday practice. Learning Outcomes for Week 2 Upon successful completion of this week, you will be able to:

  1. Explain and apply Australian data governance principles, privacy legislation, and cybersecurity practices in an allied health context.
  2. Critically assess the ethical implications of AI, including Generative AI (GenAI) and digital tools for documentation, ensuring bias mitigation, transparency, and consent.
  3. Implement safety-first approaches to managing your digital identity and data-sharing scenarios in daily practice. Topic 1: Foundations of Data Governance and Privacy Law At the heart of digital health governance is the Privacy Act

hes to managing your digital identity and data-sharing scenarios in daily practice. Topic 1: Foundations of Data Governance and Privacy Law At the heart of digital health governance is the Privacy Act 1988 (Cth), Australia’s principal law protecting personal information. As an allied health professional, you will frequently handle “sensitive information,” which receives the highest level of protection under the Act. What qualifies as sensitive information? Sensitive information is a specific type of personal information that is given a higher level of protection. It is defined as information or an opinion about an individual’s: • Racial or ethnic origin • Political opinions

Page 2

• Membership of a political association • Religious beliefs or affiliations • Philosophical beliefs • Membership of a professional or trade association • Membership of a trade union • Sexual orientation or practices • Criminal record • Health information • Genetic information • Biometric information or templates used for identification This information is considered sensitive when it is also “personal information,” meaning it is about an individual who is identified or reasonably identifiable. Mishandling this type of data poses greater risks to individuals, including discrimination or harm. Understanding your obligations is not optional; it’s a core professional competency. The 13 Australian Privacy Principles (APPs) The cornerstone of the Privacy Act is the 13 Australian Privacy Principles (APPs). These are principles-based standards that govern the entire lifecycle of patient data—from collection to destruction. They are designed to be flexible and technology-neutral, allowing them to adapt to new tools and workflows. A breach of an APP is considered an “interference with the privacy of an individual” and can lead to significant penalties and regulatory action from the Office of the Australian Information Commissioner (OAIC). Activity: The 13 APPs in

interference with the privacy of an individual” and can lead to significant penalties and regulatory action from the Office of the Australian Information Commissioner (OAIC). Activity: The 13 APPs in Allied Health Practice (Video) This animated video provides a professional overview of how the APPs apply to a typical digital health workflow. • Style: Professional 2D animation (Vyond/Animaker style) with a clear, professional narrator. • Duration: ~3 minutes • Content: o Scene 1: Introduction. An animated Allied Health Professional (AHP) is shown at their desk, managing various types of patient data on a tablet. The narrator introduces the APPs as the key guide to protecting this sensitive data. o Scene 2: The Data Lifecycle - Collection (APPs 3, 4, 5). An animated timeline appears, highlighting the first stage: “Collect.” The narrator explains that you must only collect necessary information, with consent, and be transparent about the purpose of collection.

Page 3

o Scene 3: The Data Lifecycle - Use and Disclosure (APPs 6, 7, 8, 9). The AHP is shown in a video call with a GP, securely transferring a file. They are then shown declining a data request from an insurance company. The narrator explains that data can only be used for the primary purpose it was collected for, or for a related secondary purpose the client would expect. o Scene 4: The Data Lifecycle - Quality and Security (APPs 10, 11). An animated patient record is shown, with a “Data Quality” checkmark appearing as a typo is corrected. A shield icon then appears over the record, symbolizing security. The narrator explains the duty to keep information accurate and to protect it from misuse, loss, and unauthorized access. o Scene 5: The Data Lifecycle - Individual Rights (APPs 1, 12, 13). An animated client is shown requesting and receiving a copy of their file, then asking for a correction to their address. The narrator explains the principles of transparency and the rights of individuals to

ated client is shown requesting and receiving a copy of their file, then asking for a correction to their address. The narrator explains the principles of transparency and the rights of individuals to access and correct their information. The Notifiable Data Breaches (NDB) Scheme The NDB scheme mandates that you must notify individuals and the OAIC if a data breach occurs that is likely to result in serious harm. This could be the loss of a laptop with patient files, a database hack, or even mistakenly emailing a report to the wrong person. For breaches related to the My Health Record system, there are specific, immediate notification requirements to the Australian Digital Health Agency (ADHA). Topic 2: Your Professional Digital Identity The Victorian Department of Health’s Digital Health Capability Framework highlights “Digital Professionalism” as a core domain for all health workers. This involves actively managing your digital footprint. • Securing Credentials: Use strong, unique passwords for every system and enable two- factor authentication (2FA) wherever possible. • Recognising Threats: Be able to spot phishing emails, credential-harvesting scams, or social engineering attempts designed to steal your login details. • Maintaining Boundaries: Understand that your professional and personal digital footprints can overlap. What you post on social media can be seen by clients and employers and reflects on your professional standing. Activity: Spot the Privacy Breach (H5P Interaction) • Exercise Title: “Spot the Privacy Breach: Clinic Reception Area” • Format: H5P Interactive Image-Based Exercise. You will be presented with a dynamic image of a busy clinic reception area. • Task: Review the scene and click on the “hotspots” that identify actions, objects, or behaviors that could breach patient privacy according to the Australian Privacy Principles (APPs).

Page 4

Task: Review the scene and click on the “hotspots” that identify actions, objects, or behaviors that could breach patient privacy according to the Australian Privacy Principles (APPs).

Page 4

• Hotspots & Feedback:

  1. Visible Patient Records on Desk: Feedback: This is a privacy risk. Files with personal information must be kept out of public view (APP 11: Security).
  2. Receptionist Discussing Patient Details Loudly: Feedback: Discussing health status where others can overhear violates obligations around confidential disclosure (APP 6).
  3. Unattended, Logged-in Computer: Feedback: Computers with confidential records must be locked or have screens shielded from public view (APP 11: Security).
  4. Paper Sign-In Sheet with Names Visible: Feedback: A sign-in sheet where patients can see others’ details is a privacy issue (APP 3: Collection & APP 6: Disclosure). Activity: Simulated Credential Breach (Hands-on Exercise) In this exercise, you will be given a sample email and a link to a mock login portal. Your task is to identify the phishing indicators, use URL inspection tools to test link safety, and document the steps you would take to report and mitigate the threat in a clinical setting. Topic 3: My Health Record and Consent Models My Health Record (MHR) is Australia’s national electronic health record system. While awareness is high among AHPs, survey data shows that many have concerns about privacy. Let’s address some key points. • Opt-out vs. Opt-in: The system is now opt-out for most Australians to increase coverage. However, individuals can choose to opt out or permanently delete their record at any time. • Consumer Control: Individuals have significant control. They can see who has accessed their record through an audit log and can set access controls to restrict which healthcare organizations or individual providers can view their information. • Consent Complexity: For routine clinical care, “standing consent” is applied, allowing providers invo

to restrict which healthcare organizations or individual providers can view their information. • Consent Complexity: For routine clinical care, “standing consent” is applied, allowing providers involved in a patient’s care to upload information. However, for secondary uses of data (like research or quality improvement), explicit consent is generally required. Activity: My Health Record Sandbox Access a My Health Record sandbox (or demonstration version) to practice: • Adjusting consumer access settings on a mock patient record. • Interpreting audit logs to detect potential unauthorized access attempts.

Page 5

Topic 4: The Use and Ethics of AI in Allied Health Artificial intelligence (AI), which now generally incorporates the field of machine learning (ML), has been emerging with the potential to transform allied health practice over recent years. From clinical decision support to patient rehabilitation, AI-enabled systems are augmenting care, improving efficiency, and enabling novel therapies. However, each application carries significant ethical, legal, and technical considerations. This section explores real-world AI use cases for allied health professionals (AHPs), dives deeper into generative AI (GenAI) with ambient listening systems, and examines AI-driven rehabilitation and prosthetic technologies. We will provide videos and interactive exercises to reinforce learning. First, let’s look at the contemporary landscape of AI adoption. A 2023 Queensland survey of AHPs showed relatively low levels of familiarity, but that situation is rapidly shifting: Indicator % Respondents Had not used any AI tools 80.1% (185) Reported little/no AI knowledge 87% (201) Believe AI will improve healthcare 73.6% (170 desire training and agree on benefits) Main barrier: workforce knowledge/skills 77.1% (178) This data highlights the critical need for foundational knowledge in this area to practice safely and effectively. AI in Real-World Allied Health Settings AI is a

orkforce knowledge/skills 77.1% (178) This data highlights the critical need for foundational knowledge in this area to practice safely and effectively. AI in Real-World Allied Health Settings AI is already being applied across a wide range of allied health contexts. Here are some key examples of its application. Clinical Decision Support (CDS) • Use Case: AI-powered CDS systems analyse large datasets—electronic health records (EHRs), imaging, sensor data—to flag potential risks or recommend interventions. For example, an AI model can predict pressure injury risk in community nursing by integrating patient mobility data (from wearables) with clinical notes, prompting timely preventive measures. • Machine learning algorithms are typically trained on labelled datasets and may be used for image analysis (pathology tests), or to analyse patient data from a patient record. • Ethical Considerations: Issues of lack of explainability. May be partly mitigated through use of explainable AI approaches [consider an advanced breakout here for the

Page 6

tient record. • Ethical Considerations: Issues of lack of explainability. May be partly mitigated through use of explainable AI approaches [consider an advanced breakout here for the

Page 6

interested student], clinical validation, and governance to ensure “human-in-the loop oversight. Telehealth and Remote Monitoring • Use Case: AI algorithms process sensor streams (heart rate, activity, gait) for remote physiotherapy, alerting clinicians to deteriorations or non-adherence. AI chatbots triage simple queries before escalating to a clinician, increasing service capacity. • Ethics & Privacy: Data encryption, patient consent for continuous monitoring, clarity about automated vs. human interactions. Speech and Language Processing Use Case: Speech pathology practices use natural language processing (NLP) to transcribe therapy sessions, analyse phonetic errors, and provide automated feedback. ML models can detect dysarthria or aphasia patterns, guiding tailored interventions. Models: Transformer-based architectures (e.g., fine-tuned BERT or Whisper) for accurate transcription across accents. Challenges: Ensuring inclusivity—training on diverse speech samples to avoid bias against non-native speakers or regional dialects. Mental Health Support Use Case: Chatbots offering cognitive behavioural therapy (CBT) modules, sentiment analysis of patient messages to flag risk. Passive monitoring of phone usage patterns to detect mood shifts. ML Methods: Hybrid models combining sentiment classifiers (CNNs, RNNs) with rule- based alerting systems. Ethics: Safety protocols for crisis detection, clear disclaimer about non-human support, escalation pathways. Physiotherapy & Exercise Science Computer-vision apps quantify range-of-motion via smartphone; Victoria University sensors predict fall risk in older adults. Machine-learning models personalise low-back-pain rehabilitation in Swinburne PhD work. AI scribes (Lyrebird Health, CliniScribe) cut note-taking time and en

sensors predict fall risk in older adults. Machine-learning models personalise low-back-pain rehabilitation in Swinburne PhD work. AI scribes (Lyrebird Health, CliniScribe) cut note-taking time and enhanced therapeutic alliance in a mixed-methods study of 65 clinicians; 95% would recommend the tool. Occupational Therapy CPD programs, such as “Safe & Ethical Implementation of AI in OT Practice” workshops, focus on risk mitigation and prompt engineering. Tele-rehab platforms integrate GenAI to expand intervention plans at BeachBreak OT, with transparent consent protocols. Dietetics & Nutrition GenAI meal-plan generators and precision-nutrition algorithms assist Accredited Practising Dietitians; Dietitians Australia digital-health role statement highlights ethical AI deployment. A 6-week “AI for Dietitians” national program (2025) trains APDs on GPT-prompting and data privacy. Rehabilitation & Robotics

Page 7

A University of Melbourne IEEE study identified motion-features via ML to monitor post- stroke recovery, which improves robot-assisted therapy tailoring. A La Trobe scoping review of 704 studies shows AI clustering into four themes: impairment analysis, assisted intervention, prediction/imaging, and neuroscience—with a trend toward deep-learning wearables. Project AI Modality Impact UTS SPROUTS generative-AI for aphasia ChatGPT-assisted message drafting Improved functional communication, patient-reported satisfaction. AusKidTalk corpus Speech-recognition ML Builds Australian child speech dataset for remote screening. Gold Coast swallow- sound classifier Acoustic ML Aims to replace paediatric videofluoroscopy, reducing radiation. AudA Position Statement 2024 AI scribes Emphasises “co-pilot” model— mandatory clinician review. Key AI Technologies Let’s look more closely at two of the most rapidly advancing areas of AI in healthcare: ambient listening systems and advanced prosthetics. Generative AI tools can streamline documentation by converting free-

ok more closely at two of the most rapidly advancing areas of AI in healthcare: ambient listening systems and advanced prosthetics. Generative AI tools can streamline documentation by converting free-form clinician speech into structured notes. Ambient listening systems passively record sessions, use voice activity detection and diarisation to segment speakers, then generate summaries. • Workflow: A microphone array captures audio; on-device pre-processing removes noise; encrypted audio streams to an AI engine. Speaker diarisation assigns segments to the clinician vs. the patient; NLP pipelines extract key findings, goals, and action items. • Ethical Framework: Must comply with APP 3 (sensitive data) and APP 11 (security), as well as local HREC guidelines. Transparent consent is paramount—patients must opt-in and know the recording scope. • Bias and Accuracy: NLP models may misinterpret health-specific jargon or patient speech with accents/disorders. Accuracy audits require test datasets reflecting local population demographics. Mitigation involves using domain-adapted language models and continuous feedback loops where clinicians correct AI outputs, which are then used to fine-tune models.

Page 8

ation demographics. Mitigation involves using domain-adapted language models and continuous feedback loops where clinicians correct AI outputs, which are then used to fine-tune models.

Page 8

Let’s see how this works: Activity: Video Script - Ambient AI in Practice • Style: Mixed live-action (clinician-patient) plus animated overlays to visualise AI processing. • Duration: ~3 minutes. • Script Outline: o Scene 1 (30s): (Live-action) In a clinic room, an OT explains the microphone setup to a patient. Narration: “Today’s session will be captured by our AI assistant. Don’t worry—everything is secure and used only for your care plan.” o Scene 2 (45s): (Animation) An audio waveform enters a cloud, with labelled boxes showing noise filtering, diarisation, and NLP extraction. Narration: “The ambient listening system filters out background noise, distinguishes speakers, and extracts goals, challenges, and action items in real time.” o Scene 3 (60s): (Live-action) The clinician reviews a dashboard post-session with key goals highlighted and recommended exercises auto-populated. Narration: “Within minutes, your therapist receives structured notes and tailored exercise plans, ensuring no detail is missed and freeing them to focus on you.” o Scene 4 (45s): (Animation) An ethical checklist overlay appears: consent, upload source, local processing, audit logs. Narration: “Your privacy is protected by encryption, local processing where possible, and you can review or delete recordings at any time.” o Scene 5 (30s): (Live-action) Therapist and patient high-five. Narration: “AI is here to assist, not replace. Your therapist remains fully in control.” Activity: H5P Exercise - Interactive Session Simulation • Objective: Practice identifying AI transcript errors and correcting them. • Activity: Embed H5P Dialogue Cards. • Present a snippet of an AI-generated transcript with two errors (e.g., a mis-heard medical term, an incorrect speaker tag). • Prompt the l

orrecting them. • Activity: Embed H5P Dialogue Cards. • Present a snippet of an AI-generated transcript with two errors (e.g., a mis-heard medical term, an incorrect speaker tag). • Prompt the learner to click and edit the text inline. • Provide immediate feedback by highlighting the corrected terms and explaining the rationale. • Follow-up: A multiple-choice quiz on ethical consent points (e.g., “Which APP covers the security of recorded audio?”). AI enhancement of rehabilitation by adapting patient exercises in real time and powering advanced prosthetics that respond to neural or muscular signals. • Adaptive Exercise Prescription o Use Case: Computer vision systems (using deep convolutional neural networks) track patient movements during physiotherapy via camera or wearable sensors.

Page 9

The system scores technique, counts repetitions, and adjusts difficulty based on performance. o Implementation: Edge computing on in-clinic devices ensures low latency; models are trained on anonymised motion capture datasets. o Ethics: Informed consent for video capture, secure handling of biometric data, clear role delineation—the AI suggests, but the clinician prescribes modifications. • Brain-Computer Interface (BCI) Prosthetics o Technology: Machine learning models decode electroencephalography (EEG) or electromyography (EMG) signals to control prosthetic limbs. Reinforcement learning adapts control mappings to user preferences over sessions. o Clinical Workflow: AHPs work with biomedical engineers to calibrate signal thresholds, train models on each user’s signals, and integrate sensory feedback (e.g., haptic motors) to close the loop. o Ethical & Safety Considerations: Continuous monitoring for signal drift, fail-safe modes if AI misinterprets signals, user training and autonomy, and transparent discussion about limitations and risks. Activity: Video Script - AI-Driven Prosthetics • Style: Documentary-style clinician interviews, mixed with anim

s, user training and autonomy, and transparent discussion about limitations and risks. Activity: Video Script - AI-Driven Prosthetics • Style: Documentary-style clinician interviews, mixed with animation of neural signal processing. • Duration: ~4 minutes. • Script Highlights: o A clinician introduces a patient: “Meet Sarah. After losing her arm, she’s learning to move her prosthetic hand using her thoughts.” o Animation of brain signals captured by EEG, processed by ML in real time, and translated into finger movements. o Sarah demonstrates picking up a cup; the clinician explains how reinforcement learning adapts the system to her control patterns. o Ethical note (Clinician voiceover): “We prioritise safety with built-in thresholds. If signals are ambiguous, the system pauses to prevent unintended movements.” Activity: H5P Exercise - Ethics in AI Rehabilitation • Drag and Drop: Match ethical principles to system features: o Autonomy → User control override button on prosthetic. o Non-maleficence → Built-in safety cutoff when signal confidence < 60%. o Transparency → Dashboard showing signal-to-action mapping. • Reflection prompt: “How would you explain the AI’s role to a patient worried about ‘losing control’ of their prosthetic?” (Text entry with peer feedback).

Page 10

ignal-to-action mapping. • Reflection prompt: “How would you explain the AI’s role to a patient worried about ‘losing control’ of their prosthetic?” (Text entry with peer feedback).

Page 10

Benefits & Early Outcomes The integration of AI in allied health, while still emerging, is already demonstrating significant benefits across multiple domains. Benefit Domain Evidence Administrative efficiency Note-taking time reduced; out-of-hours documentation fell significantly with AI scribes. Clinical decision support AI tools improved triage accuracy and personalised exercise dosing in physio trials. Patient engagement ChatGPT-mediated goal-setting kept stroke clients engaged online. Equity & access Remote ML speech-screening addresses regional paediatric waitlists. Economic potential 30% workload reduction in admin possible across allied health, per government discussion papers. Topic 5: Applied Ethics in Data Sharing (Scenarios) Data sharing is crucial for coordinated care, but it creates complex ethical and legal questions. Let’s apply these concepts to common scenarios. Activity: The Speech Pathologist’s Dilemma (Forum Post) • Scenario: You are a speech pathologist treating a 7-year-old child for a speech delay. The parents are separated. The father, who is not the primary carer but shares parental responsibility, emails you directly asking for a copy of the child’s latest progress report. Your service’s policy requires written consent from the primary carer (the mother) for all external disclosures, and she is the only one who signed the initial consent form. • Task: In the discussion forum, write a 300-word response outlining how you would handle this situation. Address the following:

  1. Immediate Actions: What are the first steps you would take?
  2. Legal & Ethical Considerations: Which APPs are relevant? What ethical principles are at play (e.g., duty of care to the child, fairness to both parents)?
  3. Communication: How would you professionally commun

Ethical Considerations: Which APPs are relevant? What ethical principles are at play (e.g., duty of care to the child, fairness to both parents)? 3. Communication: How would you professionally communicate your decision and the required process to the father? 4. Documentation: What would you document in the client’s record?

Page 11

Activity: Cross-Service Data Sharing (Group Case Study) • Scenario: You are an occupational therapist working across two community health services. A client with a complex history requests you share their My Health Record summary with a private provider, but they have not provided formal written consent. Your internal policy requires written consent for all external disclosures. • Task: In your group, apply the ADHA Data Sharing Guidelines and APPs 6-8 to determine:

  1. Legality: Can you share the information digitally? Under what exceptions?
  2. Ethics: What bioethical principles (autonomy, justice) guide your decision?
  3. Practical Steps: How do you document and implement the consent process?
  4. Communication: Draft a script explaining the client’s rights and the limits of digital sharing. Week 2 Wrap-up & To-Do List You’ve covered some complex but critical ground this week. You can now see how data governance, privacy legislation, and ethical frameworks are not abstract concepts, but practical tools you need for everyday allied health practice in the digital age. To Do This Week:
  5. Formative Quiz (Ungraded): Complete the 15-question multiple-choice quiz covering the APPs, the NDB scheme, and the core principles of AI ethics.
  6. Reflective Scenario Report (Ungraded): Post your 300-word response to the Ethical Dilemma in Data Sharing exercise in the discussion forum.
  7. Discussion Post (Optional): Respond to at least one peer’s post on the scenario exercise. Do you agree with their approach? Can you offer a different perspective?
  8. Review the Readings & Resources below to consolidate your learning. Key Readings & Resources

post on the scenario exercise. Do you agree with their approach? Can you offer a different perspective? 4. Review the Readings & Resources below to consolidate your learning. Key Readings & Resources Core Readings • Office of the Australian Information Commissioner (OAIC). (2022). Australian Privacy Principles quick reference. (This is the foundational guide to the APPs). • Australian Government. (2019). Australia’s Artificial Intelligence Ethics Framework. (Outlines the key principles for ethical AI in Australia). • Victorian Department of Health. (2021). Digital health capability framework for allied health professionals. (An excellent example of how these principles are applied in practice). • Australian Digital Health Agency. Website section on Data breaches and the My Health Record system. • World Health Organization. (2021). Ethics and governance of artificial intelligence for health: WHO guidance.

Page 12

• Nashwan, A. J., et al. (2024). Ethical framework for AI in healthcare research. Journal of • Therapeutic Goods Administration (TGA). Information on How the TGA regulates software-based medical devices.

Graph View

Backlinks