AI application and ethics

Question 1

Recognitze the 5 characteristics of the big data as they apply to brainhealth

Answer 1

• Volume: data from electronic health records, imaging, health apps, genetic sequencing
• variety: many kinds of data with different structures
• velocity: data vary in speed of collection/generation
• value depends on question being asked.

Question 2

distinguish between artifical intelligence, supervised machine learning, unsupervised machinelearning, and artificial intelligence

Answer 2

Artifical Intelligence: Artifical systems that appear to think like humans.

Machine learning: Systems that can learn from experience or data without direct human programming
- Supervised: modals are trained on known, labelled data. Reuqires huge volume of data and human labour
- Unsupervised: Models learn from unlabelled data. Requires huge processing power.

Question 3

demonstrate applications of AI to brain health with examples

Answer 3

1. Risk prediction
   - Goal: predict AD diagnosis using brain scans
   - Method: Train ML model, using MRI data to predict AD using neural activity.
2. clinical decision making
   - Goal: surgically remove epileptogenic brain region to treat seizures using intracranial EEG.
   - Method: Proposed ML model uses unlabelled feature of the raw ieeg output to identify seizure origin.
3. Neurotech
   - Goal: contol limb prothesis with neural activity
   - Method: Train ML model on the mapping between neural activity and limb movement.
4. Brain Modelling
   - Goal: understand how rat brains represent space
   - Method: Trained ML model to “navigate space” with training behaviour +neural activity.
   - Result: model developed representations resembling real rat cortex.

Question 4

Analyze ethical issues at the intersection of AI and brain health using terms like culpability, moral accountability, bias, and privacy

Answer 4

1. Culpability:
   - Responsibility based on ontention knowledge, or control
   - AI introduces potential harms no one person could predict or prevent
2. Moral accountability:
   - duty to explain to ones reasons and actions to others
   - AI may be unexplainable to their users
3. Bias and discrimination:
   - Groups that are under-represented may recieve lower quality care.
4. Privacy:
   - AI output can be sensitive, such as future disease risk, personal preferences, emotional states.