Artificial Intelligence as Supporting Reader in Breast Screening: A Novel Workflow to Preserve Quality and Reduce Workload

OBJECTIVE

To evaluate the effectiveness of a new strategy for using artificial intelligence (AI) as supporting reader for the detection of breast cancer in mammography-based double reading screening practice.

METHODS

Large-scale multi-site, multi-vendor data were used to retrospectively evaluate a new paradigm of AI-supported reading. Here, the AI served as the second reader only if it agrees with the recall/no-recall decision of the first human reader. Otherwise, a second human reader made an assessment followed by the standard clinical workflow. The data included 280 594 cases from 180 542 female participants screened for breast cancer at seven screening sites in two countries and using equipment from four hardware vendors. The statistical analysis included non-inferiority and superiority testing of cancer screening performance and evaluation of the reduction in workload, measured as arbitration rate and number of cases requiring second human reading.

RESULTS

Artificial intelligence as a supporting reader was found to be superior or noninferior on all screening metrics compared with human double reading while reducing the number of cases requiring second human reading by up to 87% (245 395/280 594). Compared with AI as an independent reader, the number of cases referred to arbitration was reduced from 13% (35 199/280 594) to 2% (5056/280 594).

CONCLUSION

The simulation indicates that the proposed workflow retains screening performance of human double reading while substantially reducing the workload. Further research should study the impact on the second human reader because they would only assess cases in which the AI prediction and first human reader disagree.

Coronavirus Disease 2019 (COVID-19) in Italy: Double Reading of Chest CT Examination

To assess the performance of the second reading of chest compute tomography (CT) examinations by expert radiologists in patients with discordance between the reverse transcription real-time fluorescence polymerase chain reaction (RT-PCR) test for COVID-19 viral pneumonia and the CT report. Three hundred and seventy-eight patients were included in this retrospective study (121 women and 257 men; 71 years median age, with a range of 29-93 years) and subjected to RT-PCR tests for suspicious COVID-19 infection. All patients were subjected to CT examination in order to evaluate the pulmonary disease involvement by COVID-19. CT images were reviewed first by two radiologists who identified COVID-19 typical CT patterns and then reanalyzed by another two radiologists using a CT structured report for COVID-19 diagnosis. Weighted к values were used to evaluate the inter-reader agreement. The median temporal window between RT-PCRs execution and CT scan was zero days with a range of (-9,11) days. The RT-PCR test was positive in 328/378 (86.8%). Discordance between RT-PCR and CT findings for viral pneumonia was revealed in 60 cases. The second reading changed the CT diagnosis in 16/60 (26.7%) cases contributing to an increase the concordance with the RT-PCR. Among these 60 cases, eight were false negative with positive RT-PCR, and 36 were false positive with negative RT-PCR. Sensitivity, specificity, positive predictive value and negative predictive value of CT were respectively of 97.3%, 53.8%, 89.0%, and 88.4%. Double reading of CT scans and expert second readers could increase the diagnostic confidence of radiological interpretation in COVID-19 patients.

Retrospective comparison between single reading plus an artificial intelligence algorithm and two-view digital tomosynthesis with double reading in breast screening

OBJECTIVE

To examine the breast cancer detection rate by single reading of an experienced radiologist supported by an artificial intelligence (AI) system, and compare with two-dimensional full-field digital mammography (2D-FFDM) double reading.

MATERIALS AND METHODS

Images (3D-tomosynthesis) of 161 biopsy-proven cancers were re-read by the AI algorithm and compared to the results of first human reader, second human reader and consensus following double reading in screening. Detection was assessed in subgroups by tumour type, breast density and grade, and at two operating points, referred to as a lower and a higher sensitivity threshold.

RESULTS

The AI algorithm method gave similar results to double-reading 2D-FFDM, and the detection rate was significantly higher compared to single-reading 2D-FFDM. At the lower sensitivity threshold, the algorithm was significantly more sensitive than reader A (97.5% vs. 89.4%, p = 0.02), non-significantly more sensitive than reader B (97.5% vs. 94.4%, p = 0.2) and non-significantly less sensitive than the consensus from double reading (97.5% vs. 99.4%, p = 0.2). At the higher sensitivity threshold, the algorithm was significantly more sensitive than reader A (99.4% vs. 89.4%, p < 0.001) and reader B (99.4% vs. 94.4%, p = 0.02) and identical to the consensus sensitivity (99.7% in both cases, p = 1.0). There were no significant differences in the detection capability of the AI system by tumour type, grading and density.

CONCLUSION

In this proof of principle study, we show that sensitivity using single reading with a suitable AI algorithm is non-inferior to that of standard of care using 2D mammography with double reading, when tomosynthesis is the primary screening examination.

Benefit of double-reading cytology smears as a triage strategy among high-risk human papillomavirus-positive women in Mexico

BACKGROUND

The goal of this study was to determine whether the detection of histologically confirmed cases of cervical high-grade squamous intraepithelial lesions or worse (HSIL+) can be increased by having each liquid-based cytology (LBC) slide read by 2 cytotechnologists as part of routine screening.

METHODS

Over 36,212 women aged 30 to 64 years participated in the Forwarding Research for Improved Detection and Access for Cervical Cancer Screening and Triage (FRIDA) Study in Mexico between 2013 and 2016. For each participant, 2 cervical samples were collected at the same clinic visit, one to test for high-risk human papillomavirus (hrHPV) and the other for LBC, which was used to triage those with a hrHPV positive result. LBC slides were evaluated by 7 cytotechnologists, with each slide read independently by 2 blinded cytotechnologists. All women with atypical cells of undetermined significance or a worse result were referred to colposcopy for further evaluation and diagnosis. Three pathologists evaluated the biopsy specimens to confirm the final HSIL+ diagnosis. The HSIL+ detection rates for the single versus double reading were estimated and compared.

RESULTS

A total of 3,914 women with a positive hrHPV result were triaged with LBC. The first and second cytology readings resulted in 43 HSIL+ cases detected; the double-reading strategy detected 9 additional HSIL+ cases, resulting in a total of 52 HSIL+ cases. The HSIL+ detection rate increased from 10.99/1000 with a single reading to 13.29/1000 with the double-reading strategy (P = .004).

CONCLUSION

A 20.9% increase in HSIL+ cases detected was achieved with a double reading of the LBC slides in this sample of hrHPV-positive women.

Evaluating the performance of a deep learning-based computer-aided diagnosis (DL-CAD) system for detecting and characterizing lung nodules: Comparison with the performance of double reading by radiologists

Background

The study was conducted to evaluate the performance of a state‐of‐the‐art commercial deep learning‐based computer‐aided diagnosis (DL‐CAD) system for detecting and characterizing pulmonary nodules.

Methods

Pulmonary nodules in 346 healthy subjects (male: female = 221:125, mean age 51 years) from a lung cancer screening program conducted from March to November 2017 were screened using a DL‐CAD system and double reading independently, and their performance in nodule detection and characterization were evaluated. An expert panel combined the results of the DL‐CAD system and double reading as the reference standard.

Results

The DL‐CAD system showed a higher detection rate than double reading, regardless of nodule size (86.2% vs. 79.2%; P < 0.001): nodules ≥ 5 mm (96.5% vs. 88.0%; P = 0.008); nodules < 5 mm (84.3% vs. 77.5%; P < 0.001). However, the false positive rate (per computed tomography scan) of the DL‐CAD system (1.53, 529/346) was considerably higher than that of double reading (0.13, 44/346; P < 0.001). Regarding nodule characterization, the sensitivity and specificity of the DL‐CAD system for distinguishing solid nodules > 5 mm (90.3% and 100.0%, respectively) and ground‐glass nodules (100.0% and 96.1%, respectively) were close to that of double reading, but dropped to 55.5% and 93%, respectively, when discriminating part solid nodules.

Conclusion

Our DL‐CAD system detected significantly more nodules than double reading. In the future, false positive findings should be further reduced and characterization accuracy improved.

Is single reading with computer-aided detection (CAD) as good as double reading in mammography screening? A systematic review

BACKGROUND

In accordance with European guidelines, mammography screening comprises independent readings by two breast radiologists (double reading). CAD (computer-aided detection) has been suggested to complement or replace one of the two readers (single reading + CAD).The aim of this systematic review is to address the following question: Is the reading of mammographic x-ray images by a single breast radiologist together with CAD at least as accurate as double reading?

METHODS

The electronic literature search included the databases Pub Med, EMBASE and The Cochrane Library. Two independent reviewers assessed abstracts and full-text articles.

RESULTS

1049 abstracts were identified, of which 996 were excluded with reference to inclusion and exclusion criteria; 53 full-text articles were assessed for eligibility. Finally, four articles were included in the qualitative analysis, and one in a GRADE synthesis.

CONCLUSIONS

The scientific evidence is insufficient to determine whether the accuracy of single reading + CAD is at least equivalent to that obtained in standard practice, i.e. double reading where two breast radiologists independently read the mammographic images.

Radiologists' perceptions of computer aided detection versus double reading for mammography interpretation

RATIONALE AND OBJECTIVES

The aim of this study was to examine radiologists' use and perceptions of computer-aided detection (CAD) and double reading for screening mammography interpretation.

MATERIALS AND METHODS

A mailed survey of 257 community radiologists participating in the national Breast Cancer Surveillance Consortium was used to assess perceptions and practices related to CAD and double reading. Latent class analysis was used to classify radiologists' overall perceptions of CAD and double reading on the basis of their agreement or disagreement with specific statements about CAD and double reading.

RESULTS

Most radiologists (64%) reported using CAD for more than half the screening mammograms they interpreted, but only <5% reported double reading that much. More radiologists perceived that double reading improved cancer detection rates compared to CAD (74% vs 55% reported), whereas fewer radiologists thought that double reading decreased recall rates compared to CAD (50% vs 65% reported). Radiologists with the most favorable perceptions of CAD were more likely to think that CAD improved cancer detection rates without taking too much time compared to radiologists with the most unfavorable overall perceptions. In latent class analysis, an overall favorable perception of CAD was significantly associated with the use of CAD (81%), a higher percentage of workload in screening mammography (80%), academic affiliation (71%), and fellowship training (58%). Perceptions of double reading that were most favorable were associated with academic affiliation (98%).

CONCLUSIONS

Radiologists' perceptions were more favorable toward double reading by a second clinician than by a computer, although fewer used double reading in their own practice. The majority of radiologists perceived both CAD and double reading at least somewhat favorably, although for largely different reasons.