Search

Background: To address the numerous unmeet clinical needs, in recent years several Machine Learning models applied to medical images and clinical data have been introduced and developed. Even when they achieve encouraging results, they lack evolutionary progression, thus perpetuating their status as autonomous entities. We postulated that different algorithms which have been proposed in the literature to address the same diagnostic task, can be aggregated to enhance classification performance. We suggested a proof of concept to define an ensemble approach useful for integrating different algorithms proposed to solve the same clinical task. Methods: The proposed approach was developed starting from a public database consisting of radiomic features extracted from CT images relating to 535 patients suffering from lung cancer. Seven algorithms were trained independently by participants in the AI4MP working group on Artificial Intelligence of the Italian Association of Physics in Medicine to discriminate metastatic from non-metastatic patients. The classification scores generated by these algorithms are used to train SVM classifier. The Explainable Artificial Intelligence approach is applied to the final model. The ensemble model was validated following an 80–20 hold-out and leave-one-out scheme on the training set. Results: Compared to individual algorithms, a more accurate result was achieved. On the independent test the ensemble model achieved an accuracy of 0.78, a F1-score of 0.57 and a log-loss of 0.49. Shapley values representing the contribution of each algorithm to the final classification result of the ensemble model were calculated. This information represents an added value for the end user useful for evaluating the appropriateness of the classification result on a particular case. It also allows us to evaluate on a global level which methodological approaches of the individual algorithms are likely to have the most impact. Conclusion: Our proposal represents an innovative approach useful for integrating different algorithms that populate the literature and which lays the foundations for future evaluations in broader application scenarios. [ABSTRACT FROM AUTHOR]

In the era of artificial intelligence and precision medicine, the use of quantitative imaging methodological approaches could improve the cancer patient's therapeutic approaches. Specifically, our pilot study aims to explore whether CT texture features on both baseline and first post-treatment contrast-enhanced CT may act as a predictor of overall survival (OS) and progression-free survival (PFS) in metastatic melanoma (MM) patients treated with the PD-1 inhibitor Nivolumab. Ninety-four lesions from 32 patients treated with Nivolumab were analyzed. Manual segmentation was performed using a free-hand polygon approach by drawing a region of interest (ROI) around each target lesion (up to five lesions were selected per patient according to RECIST 1.1). Filtration-histogram-based texture analysis was employed using a commercially available research software called TexRAD (Feedback Medical Ltd, London, UK; https://fbkmed.com/texrad-landing-2/) Percentage changes in texture features were calculated to perform delta-radiomics analysis. Texture feature kurtosis at fine and medium filter scale predicted OS and PFS. A higher kurtosis is correlated with good prognosis; kurtosis values greater than 1.11 for SSF = 2 and 1.20 for SSF = 3 were indicators of higher OS (fine texture: 192 HR = 0.56, 95% CI = 0.32–0.96, p = 0.03; medium texture: HR = 0.54, 95% CI = 0.29–0.99, p = 0.04) and PFS (fine texture: HR = 0.53, 95% CI = 0.29–0.95, p = 0.03; medium texture: HR = 0.49, 209 95% CI = 0.25–0.96, p = 0.03). In delta-radiomics analysis, the entropy percentage variation correlated with OS and PFS. Increasing entropy indicates a worse outcome. An entropy variation greater than 5% was an indicator of bad prognosis. CT delta-texture analysis quantified as entropy predicted OS and PFS. Baseline CT texture quantified as kurtosis also predicted survival baseline. Further studies with larger cohorts are mandatory to confirm these promising exploratory results. [ABSTRACT FROM AUTHOR]