The learning curve in bladder MRI using VI-RADS assessment score during an interactive dedicated training program.

OBJECTIVE: The purpose of the study was to evaluate the effect of an interactive training program on the learning curve of radiology residents for bladder MRI interpretation using the VI-RADS score. METHODS: Three radiology residents with minimal experience in bladder MRI served as readers. They blindly evaluated 200 studies divided into 4 subsets of 50 cases over a 3-month period. After 2 months, the first subset was reassessed, resulting in a total of 250 evaluations. An interactive training program was provided and included educational lessons and case-based practice. The learning curve was constructed by plotting mean agreement as the ratio of correct evaluations per batch. Inter-reader agreement and diagnostic performance analysis were performed with kappa statistics and ROC analysis. RESULTS: As for the VI-RADS scoring agreement, the kappa differences between pre-training and post-training evaluation of the same group of cases were 0.555 to 0.852 for reader 1, 0.522 to 0.695 for reader 2, and 0.481 to 0.794 for reader 3. Using VI-RADS ≥ 3 as cut-off for muscle invasion, sensitivity ranged from 84 to 89% and specificity from 91 to 94%, while the AUCs from 0.89 (95% CI:0.84, 0.94) to 0.90 (95% CI:0.86, 0.95). Mean evaluation time decreased from 5.21 ± 1.12 to 3.52 ± 0.69 min in subsets 1 and 5. Mean grade of confidence improved from 3.31 ± 0.93 to 4.21 ± 0.69, in subsets 1 and 5. CONCLUSION: An interactive dedicated education program on bladder MRI and the VI-RADS score led to a significant increase in readers' diagnostic performance over time, with a general improvement observed after 100-150 cases. KEY POINTS: • After the first educational lesson and 100 cases were interpreted, the concordance on VI-RADS scoring between the residents and the experienced radiologist was significantly higher. • An increase in the grade of confidence was experienced after 100 cases. • We found a decrease in the evaluation time after 150 cases.

Comprehensive Perspective for Lung Cancer Characterisation Based on AI Solutions Using CT Images.

Lung cancer is still the leading cause of cancer death in the world. For this reason, novel approaches for early and more accurate diagnosis are needed. Computer-aided decision (CAD) can be an interesting option for a noninvasive tumour characterisation based on thoracic computed tomography (CT) image analysis. Until now, radiomics have been focused on tumour features analysis, and have not considered the information on other lung structures that can have relevant features for tumour genotype classification, especially for epidermal growth factor receptor (EGFR), which is the mutation with the most successful targeted therapies. With this perspective paper, we aim to explore a comprehensive analysis of the need to combine the information from tumours with other lung structures for the next generation of CADs, which could create a high impact on targeted therapies and personalised medicine. The forthcoming artificial intelligence (AI)-based approaches for lung cancer assessment should be able to make a holistic analysis, capturing information from pathological processes involved in cancer development. The powerful and interpretable AI models allow us to identify novel biomarkers of cancer development, contributing to new insights about the pathological processes, and making a more accurate diagnosis to help in the treatment plan selection.