Machines study guide with 100 correct

HOSA Biomedical Debate 23-24 Machines study guide with 100% correct answers MYCIN (1970) - answer - Helps with diagnosing bacterial infections by analyzing patient symptoms and providing treatment recommendations

- Limitations: lack flexibility and inability to adapt to new info

Neural Networks (1980's-1990's) - answer - Uses interconnected nodes to analyze data and identify patterns -Quick medical reference: Patient data to predict possibilities of medical conditions

-Limitations: availability and quality of data , often incomplete or inaccurate

Support Vector Machines - answer - Classification and regression analysis

• Diagnosis of breast cancer, detection of Alzheimer's disease, prediction of patient outcomes.

Deep Learning (DL) - answer - Uses artificial neural networks to analyze large datasets and identify complex patterns in data and make accurate predictions

• For medical imaging, drug discovery, and personalized medicine

Medical Imaging - answer - Diagnosis and treatment of many medical conditions

• Can improve accuracy and efficiency of medical imaging by automating image analysis and providing

more accurate diagnoses

• Can analyze large data sets of medical records and genetic data to get effective and personalized

treatments.

• AI potential: Drug discovery (AI identifies drug candidates faster and accurately than traditional

method)

- Advances: able to identify subtle patterns and features that humans miss

• Limitations: ability of radiologists to interpret complex images 1 / 2

Rule Based System - answer - Operates on predefined rules and decision-making algorithms to analyze data & generate recommendations/decisions

• Can be utilized to analyze electrocardiography (ECG) data for diagnosing heart disease
• Advantages: transparency and interoperability, adaptable and can be easily updated with new

knowledge

• Limitations: Heavily rely on accuracy and completeness of predefined rules, Struggle to handle

uncertainty or ambiguity Robotic Process Automation - answer - Automates rules based tasks by mimicking human interactions with computer systems and applications

• Perform tasks that require little to no cognitive decision making
• Used to automate the process of verifying patient insurance eligibility
• Programmed to extract patient data from health records, interact w/ insurance portals, and perform

verification steps

• Advantages: Can work with existing It infrastructure w/o needing extensive change or integrations, can

work 24/7 without interruptions, and can interact with apps and systems through same user interfaces as humans

• Limitations: not good for tasks with complex decision making, unstructured data, or needing high level

cognitive abilities Machine Learning (ML) - answer - Uses training algorithms on large datasets to identify patterns and make predictions

• Used for diagnostic imaging, drug discovery, and personalized treatment recommendations

Natural Language Processing - answer - Analyze electronic health records, patient notes, other text based data to identify potential trends and patterns Robotics - answer - Surgical robots use advanced sensors and imaging systems to provide surgeons with real time feedback during procedures

• Patient monitoring and rehabilitation

Expert Systems - answer Replicate decision making capabilities of human experts

• / 2