ABSTRACT
Artificial intelligence (AI) tools may assist breast screening mammography programs, but limited evidence supports their generalizability to new settings. This retrospective study used a 3-year dataset (April 1, 2016-March 31, 2019) from a U.K. regional screening program. The performance of a commercially available breast screening AI algorithm was assessed with a prespecified and site-specific decision threshold to evaluate whether its performance was transferable to a new clinical site. The dataset consisted of women (aged approximately 50-70 years) who attended routine screening, excluding self-referrals, those with complex physical requirements, those who had undergone a previous mastectomy, and those who underwent screening that had technical recalls or did not have the four standard image views. In total, 55 916 screening attendees (mean age, 60 years ± 6 [SD]) met the inclusion criteria. The prespecified threshold resulted in high recall rates (48.3%, 21 929 of 45 444), which reduced to 13.0% (5896 of 45 444) following threshold calibration, closer to the observed service level (5.0%, 2774 of 55 916). Recall rates also increased approximately threefold following a software upgrade on the mammography equipment, requiring per-software version thresholds. Using software-specific thresholds, the AI algorithm would have recalled 277 of 303 (91.4%) screen-detected cancers and 47 of 138 (34.1%) interval cancers. AI performance and thresholds should be validated for new clinical settings before deployment, while quality assurance systems should monitor AI performance for consistency. Keywords: Breast, Screening, Mammography, Computer Applications-Detection/Diagnosis, Neoplasms-Primary, Technology Assessment Supplemental material is available for this article. © RSNA, 2023.
ABSTRACT
OBJECTIVES: Rituximab is effective in inducing remission in ANCA-associated vasculitis (AAV), with randomized evidence to support its use as four infusions of 375 mg/m(2) (the conventional lymphoma dosing schedule). As B cell depletion (BCD) appears to occur very rapidly after the first dose, we questioned the need for repeat dosing and adopted a standard single-dose protocol of 375 mg/m(2) to treat active AAV. METHODS: All consecutive cases with newly diagnosed or relapsing AAV for whom conventional immunosuppression was contraindicated or ineffective were enrolled. All were rituximab naive. Circulating CD19(+) B cells and clinical and serological markers of disease activity were recorded at regular intervals. Complete remission (CR) was defined as the absence of clinical features of AAV with a prednisolone dose <10 mg/day. RESULTS: Nineteen patients were included, 17 (89%) with generalized disease and 2 (11%) with severe disease (creatinine level >500 µM). Eight (42%) were on additional immunosuppression at the time of rituximab treatment. Satisfactory BCD (<0.005 cells/µl) was achieved in 89% of patients after a median of 13 days. Three-month BCD probability was 89%. Median time to CR following a single dose of rituximab was 38 days and the 3-month probability of CR was 80%. Median time to B cell repopulation was 9.2 months and to disease relapse/redose was 27 months. Use of this single-dose protocol saved an estimated £4533/patient (US$7103; 5276) compared with a 4 × 375 mg/m(2) dosing schedule. CONCLUSION: Our single-centre experience suggests that a single dose of rituximab of 375 mg/m(2) is a reasonable and more cost-effective therapy for inducing remission in patients with AAV.
