Abstract
Purpose: Human perception and decisions are biased toward previously seen
stimuli. This phenomenon is known as serial dependence and has been extensively studied for the last decade. Recent evidence suggests that clinicians’ judgments of mammograms might also be impacted by serial dependence. However, the stimuli used in previous psychophysical experiments on this question, consisting of artificial geometric shapes and healthy tissue backgrounds, were unrealistic. We utilized realistic and controlled generative adversarial network (GAN)-generated radiographs that were produced by GAN to mimic images that clinicians typically encounter.
Approach: Mammograms from the digital database for screening mammography
were utilized to train a GAN. This pretrained GAN was then adopted to generate
a large set of authentic-looking simulated mammograms: 20 circular morph continuums, each with 147 images, for a total of 2940 images. Using these stimuli in a standard serial dependence experiment, participants viewed a random GAN-generated mammogram on each trial and subsequently matched the GAN-generated mammogram encountered using a continuous report. The characteristics of serial dependence from each continuum were analyzed.
Results: We found that serial dependence affected the perception of all naturalistic GAN-generated mammogram morph continuums. In all cases, the perceptual judgments of GAN-generated mammograms were biased toward previously encountered GAN-generated mammograms. On average, perceptual decisions had 7% categorization errors that were pulled in the direction of serial dependence.
Conclusions: Serial dependence was found even in the perception of naturalistic
GAN-generated mammograms created by a GAN. This supports the idea that serial dependence could, in principle, contribute to decision errors in medical image perception tasks
stimuli. This phenomenon is known as serial dependence and has been extensively studied for the last decade. Recent evidence suggests that clinicians’ judgments of mammograms might also be impacted by serial dependence. However, the stimuli used in previous psychophysical experiments on this question, consisting of artificial geometric shapes and healthy tissue backgrounds, were unrealistic. We utilized realistic and controlled generative adversarial network (GAN)-generated radiographs that were produced by GAN to mimic images that clinicians typically encounter.
Approach: Mammograms from the digital database for screening mammography
were utilized to train a GAN. This pretrained GAN was then adopted to generate
a large set of authentic-looking simulated mammograms: 20 circular morph continuums, each with 147 images, for a total of 2940 images. Using these stimuli in a standard serial dependence experiment, participants viewed a random GAN-generated mammogram on each trial and subsequently matched the GAN-generated mammogram encountered using a continuous report. The characteristics of serial dependence from each continuum were analyzed.
Results: We found that serial dependence affected the perception of all naturalistic GAN-generated mammogram morph continuums. In all cases, the perceptual judgments of GAN-generated mammograms were biased toward previously encountered GAN-generated mammograms. On average, perceptual decisions had 7% categorization errors that were pulled in the direction of serial dependence.
Conclusions: Serial dependence was found even in the perception of naturalistic
GAN-generated mammograms created by a GAN. This supports the idea that serial dependence could, in principle, contribute to decision errors in medical image perception tasks
| Original language | English |
|---|---|
| Article number | 045501 |
| Number of pages | 15 |
| Journal | Journal of Medical Imaging |
| Volume | 10 |
| Issue number | 4 |
| Early online date | 4 Jul 2023 |
| DOIs | |
| Publication status | Published - 4 Jul 2023 |
Bibliographical note
AcknowledgmentsThis work was supported by the National Institutes of Health (Grant No. R01CA236793).
Funding Information
The authors have identified the following funders and award numbers, either on the submission form at the time of submission or in the Acknowledgments of the manuscript. Please check this list of funding agencies and make any necessary corrections using the full and official name of the funding organization. You may also wish to edit the Acknowledgments, if needed. This information may be used to help SPIE comply with funding reporting mandates.
• National Institutes of Health; Award no. R01CA236793
Keywords
- serial dependance
- generative adversarial networks
- visual search
- radiological searching
Access to Document
- Ren_JMI_etal_Serial_Dependence_In_VoR
© The Authors. Published by SPIE under a Creative Commons Attribution 4.0 International License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI. [DOI: 10.1117/1.JMI.10.4.045501] https://creativecommons.org/licenses/by/4.0/