English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Multiple objective assessment of image-quality-based studies have reportedthat several deep-learning-based denoising methods show limited performance onsignal-detection tasks. Our goal was to investigate the reasons for thislimited performance. To achieve this goal, we conducted a task-basedcharacterization of a DL-based denoising approach for individual signalproperties. We conducted this study in the context of evaluating a DL-basedapproach for denoising SPECT images. The training data consisted of signals ofdifferent sizes and shapes within a clustered-lumpy background, imaged with a2D parallel-hole-collimator SPECT system. The projections were generated atnormal and 20% low count level, both of which were reconstructed using an OSEMalgorithm. A CNN-based denoiser was trained to process the low-count images.The performance of this CNN was characterized for five different signal sizesand four different SBR by designing each evaluation as an SKE/BKSsignal-detection task. Performance on this task was evaluated using ananthropomorphic CHO. As in previous studies, we observed that the DL-baseddenoising method did not improve performance on signal-detection tasks.Evaluation using the idea of observer-study-based characterization demonstratedthat the DL-based denoising approach did not improve performance on thesignal-detection task for any of the signal types. Overall, these resultsprovide new insights on the performance of the DL-based denoising approach as afunction of signal size and contrast. More generally, the observer study-basedcharacterization provides a mechanism to evaluate the sensitivity of the methodto specific object properties and may be explored as analogous tocharacterizations such as modulation transfer function for linear systems.Finally, this work underscores the need for objective task-based evaluation ofDL-based denoising approaches.

Investigating the limited performance of a deep-learning-based SPECT denoising approach: An observer-study-based characterization