Breast Cancer
Conditions
Brief summary
The purpose of this pilot is to provide credible performance estimate information in order to conduct a subsequent pivotal reader study.
Detailed description
This clinical research is a retrospective, pilot, multi-reader, multi-case (MRMC) study with an enriched sample of 100 breast screening or diagnostic cases which were selected from the library of mammograms collected under Fujifilm protocol FMSU2013-004A (all subjects previously provided written informed consent agreeing their image data and supporting documentation could be used for future research and investigations). Approximately 6 qualified radiologists will independently perform two reads on all (approximately 100) cases. Each reader will read each case both as a FFDM (Full Field Digital Mammography) read, and a DBT (Digital Breast Tomosynthesis) plus S-View (Synthesized View) read on the ASPIRE Bellus II workstation.
Interventions
DEVICEDBT plus S-View
DBT plus S-View images
DEVICEFFDM alone
FFDM alone images
Sponsors
Fujifilm Medical Systems USA, Inc.
Study design
Observational model
CASE_CROSSOVER
Time perspective
RETROSPECTIVE
Eligibility
Sex/Gender
FEMALE
Age
18 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
* Female subjects participating in FMSU2013-004A protocol with known clinical status
Exclusion criteria
* Female subjects that did not have known clinical status in FMSU2013-004A
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Compare Per Subject Average Area Under the ROC (Receiver Operating Characteristic) Curve (AUC) | 2 visits with 4 week washout period | Compare per subject average area under the ROC curve (AUC) for DBT (Digital Breast Tomosynthesis) plus S-View (Synthesized View) versus FFDM (Full Field Digital Mammography), based on per-subject POM (probability of malignancy) scores and requiring correct lesion localization. The primary endpoint was non-inferior per-subject average AUC. Statistician estimated AUCs for each reader in each review condition (FFDM, DBT plus S-View) based on per-subject POM scores requiring correct lesion localization. Statistician also calculated for each reader the non-parametric (trapezoidal) AUC for the FFDM read, the DBT plus S-View read, and the difference between them. These AUC's, differences between them, and the associated variance-covariance matrix were obtained using the method of DeLong, et al. |
Countries
United States
Participant flow
Pre-assignment details
This was a retrospective, multi-reader multi-case reader study. Six (6) radiologist readers read all 100 cases (half FFDM and half DBT plus S-View reads) at each of 2 visits which were separated by a memory washout period of approximately 4 weeks. Images were previously acquired under Fujifilm protocol FMSU2013-004A.
Participants by arm
| Arm | Count |
|---|---|
| DBT Plus S-View and FFDM Alone Readers will read half of the DBT plus S-View images followed by half of the FFDM only images at each of 2 sessions. There was a four week memory washout period between the 2 sessions. | 100 |
| Total | 100 |
Baseline characteristics
| Characteristic | DBT Plus S-View and FFDM Alone |
|---|---|
| Age, Continuous | 58 years |
| Race (NIH/OMB) American Indian or Alaska Native | 0 Participants |
| Race (NIH/OMB) Asian | 2 Participants |
| Race (NIH/OMB) Black or African American | 6 Participants |
| Race (NIH/OMB) More than one race | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 1 Participants |
| Race (NIH/OMB) White | 91 Participants |
| Region of Enrollment United States | 100 participants |
| Sex: Female, Male Female | 100 Participants |
| Sex: Female, Male Male | 0 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 100 | 0 / 100 |
| other Total, other adverse events | 0 / 100 | 0 / 100 |
| serious Total, serious adverse events | 0 / 100 | 0 / 100 |
Outcome results
Primary
Compare Per Subject Average Area Under the ROC (Receiver Operating Characteristic) Curve (AUC)
Compare per subject average area under the ROC curve (AUC) for DBT (Digital Breast Tomosynthesis) plus S-View (Synthesized View) versus FFDM (Full Field Digital Mammography), based on per-subject POM (probability of malignancy) scores and requiring correct lesion localization. The primary endpoint was non-inferior per-subject average AUC. Statistician estimated AUCs for each reader in each review condition (FFDM, DBT plus S-View) based on per-subject POM scores requiring correct lesion localization. Statistician also calculated for each reader the non-parametric (trapezoidal) AUC for the FFDM read, the DBT plus S-View read, and the difference between them. These AUC's, differences between them, and the associated variance-covariance matrix were obtained using the method of DeLong, et al.
Time frame: 2 visits with 4 week washout period
Population: Retrospective images used from Fujifilm protocol FMSU2013-004A. All readers read the same 100 images in their own randomized order during 2 sessions (they read half of each modality at each session which were separated by a 4 week washout period).
| Arm | Measure | Value (MEAN) |
|---|---|---|
| DBT Plus S-View | Compare Per Subject Average Area Under the ROC (Receiver Operating Characteristic) Curve (AUC) | 0.819 Probability |
| FFDM Alone | Compare Per Subject Average Area Under the ROC (Receiver Operating Characteristic) Curve (AUC) | 0.796 Probability |
Comparison: The primary endpoint for this study was a non-inferior per-subject average area under the receiver operating characteristic (ROC) curve (AUC) requiring correct lesion localization for DBT (digital breast tomosynthesis) plus S-View (synthesized view) versus FFDM (full field digital mammography). AUCs for each reader were estimated in each review condition based on per-subject probability of malignancy (POM) scores requiring correct lesion localization.p-value: <0.01t-test, 2 sided