AI In Imaging 5.3.24

Question 1

Artificial intelligence definition

Answer 1

“Artificial intelligence is the science of making Ines do things that would require intelligence, if done by men” minsky 1950’s

Consist of implementing a number of techniques aimed at enabling machines to imitate a real form of intelligence

Within a generation, the problem of AI will be substantially solved - Minsky ( he wasn’t correct with this statement)

Question 2

Machine learning

Answer 2

1) machine learning identifies patterns, using statistical learning and computers by unearthing boundaries In data sets. you can use it to make predictions

2) One method for making predictions is called decision trees which uses a series of if- then statements to identify boundaries and define patterns in the data

3) Overfitting happens when some boundaries are based on distinctions that don’t make a difference. You can see if a model over fits by having test data flow through the model.

Question 3

Real world machine learning examples

Answer 3

Speech recognition
Fraud detection
Automated stock tracking
Robotic process automation (rpa)
Recommendation engines
Online chatbots - customer service
Computer vision

Question 4

Deep learning definition

Answer 4

“Scalable machine learning”

Deep learning models can be taught to perform classification tasks and recognise patterns in photos, Text, audios and other various data. it also is used to automate tasks that would normally need human intelligence such as describing images or transcribing audio files.

Question 5

Neural network definition

Answer 5

A computer system modelled on the human brain and nervous system.

Question 6

Examples of neural network

Answer 6

• Chat GPT

Question 7

Al uses in acquiring and processing images

Answer 7

Al assisted image, processing platform that provides diagnostic quality images from the very first x-ray

Fast planning facilities, organ based setting of scan and recon ranges, aiming for safer faster and more standardised workflow at the scanner

Question 8

Benefits of AI in imaging

Answer 8

+ Consistent image quality - consistent brightness and contrast despite variations and exam conditions
+ exceptional detail - achieve outstanding clarity and anatomical detail across image types
+ adaptive noise reduction - reduces noise while minimising the effect on fine details
+ metal implant handling - clear bone- metal interface without halo artifact

Question 9

Risks of AI - data collection

Answer 9

Bias added unintentionally from data collection

Unseen variations

Change in equipment

Small change in input = big change in output ( leaf digram)