In the medical domain, experts are often required to make diagnoses by viewing medical images. How did they gain their visual expertise? How can we best train others to gain a similar level of expertise as quickly as possible? Some of the work in the lab on this topic has focused on comparing and optimising breast screening protocols. Other work has investigated to what degree perceptual training can be used to train people to recognise fractures in X-ray images and, more recently, to diagnose fatty liver disease in liver ultrasound images. Additionally, we are interested in comparing human expertise against computers and investigating to what degree deep convolution neural networks can be used as models of human perceptual learning.
Selected Journal Articles
- Adams, M, Chen, W, Holcdorf, D, McCusker, MW, Howe, PDL, & Gaillard, F. (2019). Computer vs human: Deep learning versus perceptual training for the detection of neck of femur fractures. Journal of Medical Imaging and Radiation Oncology, 63(1), 27-32. [PDF]
- Chen, W, HolcDorf, D, McCusker, MW, Gaillard, F, Howe, PDL (2017). Perceptual training to improve hip fracture identification in conventional radiographs. PLOS ONE, 12(12), e0189192. [PDF]
- Chen W & Howe PDL (2016). Comparing breast screening protocols: Inserting catch trials does not improve sensitivity over double screening. PLOS ONE 11(10):12. [PDF]