Objectives: To demonstrate that radiologists, with the help of artificial intelligence (AI), are able to better classify screening mammograms into the correct breast imaging reporting and data system (BI-RADS) category, and as a secondary objective, to explore the impact of AI on cancer detection and mammogram interpretation time. Methods: A multi-reader, multi-case study with cross-over design, was performed, including 314 mammograms. Twelve radiologists interpreted the examinations in two sessions delayed by a 4 weeks wash-out period with and without AI support. For each breast of each mammogram, they had to mark the most suspicious lesion (if any) and assign it with a forced BI-RADS category and a level of suspicion or “continuous BI-RADS 100”. Cohen’s kappa correlation coefficient evaluating the inter-observer agreement for BI-RADS category per breast, and the area under the receiver operating characteristic curve (AUC), were used as metrics and analyzed. Results: On average, the quadratic kappa coefficient increased significantly when using AI for all readers [κ = 0.549, 95% CI (0.528–0.571) without AI and κ = 0.626, 95% CI (0.607–0.6455) with AI]. AUC was significantly improved when using AI (0.74 vs 0.77, p = 0.004). Reading time was not significantly affected for all readers (106 s without AI and vs 102 s with AI; p = 0.754). Conclusions: When using AI, radiologists were able to better assign mammograms with the correct BI-RADS category without slowing down the interpretation time.
Uterus measurements are useful for assessing both the treatment and follow-ups of gynaecological patients. The aim of our study was to develop a deep learning (DL) tool for fully automated measurement of the three-dimensional size of the uterus on magnetic resonance imaging (MRI). In this single-centre retrospective study, 900 cases were included to train, validate, and test a VGG-16/VGG-11 convolutional neural network (CNN). The ground truth was manual measurement. The performance of the model was evaluated using the objective key point similarity (OKS), the mean difference in millimetres, and coefficient of determination R2. The OKS of our model was 0.92 (validation) and 0.96 (test). The average deviation and R2 coefficient between the AI measurements and the manual ones were, respectively, 3.9 mm and 0.93 for two-point length, 3.7 mm and 0.94 for three-point length, 2.6 mm and 0.93 for width, 4.2 mm and 0.75 for thickness. The inter-radiologist variability was 1.4 mm. A three-dimensional automated measurement was obtained in 1.6 s. In conclusion, our model was able to locate the uterus on MRIs and place measurement points on it to obtain its three-dimensional measurement with a very good correlation compared to manual measurements.
Objectives: Few studies have evaluated and compared artificial intelligence (AI) models for breast cancer detection on the same database. The purpose is to compare the performances of 3 artificial intelligence algorithms on the same dataset at a free operative threshold and to evaluate how the performance is impacted by the chosen threshold. Materials: In this retrospective single-center study, a dataset of 314 2D screening mammograms performed between 2012 and 2020 was established with a prevalence of 19.6% of histologically proven cancer. Three AI constructors using CE (European conformity) marking algorithms have agreed to take part in the study by submitting the cancer-enriched cohort to their algorithm. They chose a free operative threshold to distinguish benign and malignant mammograms. The most suspicious lesion was marked and correctly located by the AI algorithm. Statistical parameters were analyzed and compared between the algorithms. Results: Regarding both sensitivity and specificity, at the chosen threshold, AI 1 had the best compromise between sensitivity (74%) and specificity (79%). AI 2 had a statistically lower sensitivity (52%; p<0.05) with a higher specificity (98.4%) than the other AI algorithms. AI 3 had a nonsignificant sensitivity difference (69.9%) but a significantly lower specificity (45.3%; p<0.001) than the other two. The performances varied depending on the chosen threshold; when the AI 2 threshold was lowered, the sensitivity increased (69.9%), while a higher specificity (86.1%) was maintained. Conclusion: Two AI algorithms stood out in terms of performance when the threshold was optimized, resulting in an acceptable sensitivity and specificity.