In 2020, an experiment testing AI solutions for lung X-ray analysis on a multi-hospital network was conducted. The multi-hospital network linked 178 Moscow state healthcare centers, where all chest X-rays from the network were redirected to a research facility, analyzed with AI, and returned to the centers. The experiment was formulated as a public competition with monetary awards for participating industrial and research teams. The task was to perform the binary detection of abnormalities from chest X-rays. For the objective real-life evaluation, no training X-rays were provided to the participants. This paper presents one of the top-performing AI frameworks from this experiment. First, the framework used two EfficientNets, histograms of gradients, Haar feature ensembles, and local binary patterns to recognize whether an input image represents an acceptable lung X-ray sample, meaning the X-ray is not grayscale inverted, is a frontal chest X-ray, and completely captures both lung fields. Second, the framework extracted the region with lung fields and then passed them to a multi-head DenseNet, where the heads recognized the patient’s gender, age and the potential presence of abnormalities, and generated the heatmap with the abnormality regions highlighted. During one month of the experiment from 11.23.2020 to 12.25.2020, 17,888 cases have been analyzed by the framework with 11,902 cases having radiological reports with the reference diagnoses that were unequivocally parsed by the experiment organizers. The performance measured in terms of the area under receiving operator curve (AUC) was 0.77. The AUC for individual diseases ranged from 0.55 for herniation to 0.90 for pneumothorax.
Volumetric modulated arc therapy (VMAT) for lung Stereotactic body radiation therapy (SBRT) has been used effectively in combination with flattening filter-free (FFF) beams with higher dose rates. Robust optimization provides significantly more robust dose distributions to targets and organ at risk (OAR) than the planning target volume (PTV)-based optimization. The fluence of an FFF beam may influence the dosimetric outcome for lung cancer, because a secondary build-up and lateral disequilibrium may affect target coverage. This study aims to assess the differences between 6X-and 10X-FFF beams using PTV-based and robust optimizations for lung cancer patients. Lung VMAT SBRT plans incorporating PTV-based and robust optimized plans using 6X-and 10X-FFF beams were generated and evaluated for perturbation doses to the internal target volume (ITV) and lung. Materials/Methods: Ten lung cancer patients with a breath-holding were selected. All VMAT plans were contoured, optimized, and calculated based on a breath-holding CT image dataset used in the treatment planning system. Four VMAT plans were generated for each patient; namely, an optimized plan based on the planning target volume PTV margin and a second plan based on a robust optimization of the ITV with setup uncertainties, each for the 6X-and 10X-FFF beams. Both optimized plans were normalized by the percentage of the prescription dose covering 95% of the target volume (D 95%) to the PTV. The dose prescription was 42 Gy with 10.5 Gy per fraction. All optimized plans were evaluated using perturbed doses by specifying user-defined shifted values from the isocenter. The ITV D 99% and lung doses were evaluated. Results: As for the robust optimized plans comparison, the D 99% doses to the ITV exhibited statistically significant differences for the 6X-and 10X-FFF beams (p Z 0.010). The mean dose, V 20Gy , and V 5Gy to the lung for the 10X-FFF beam were increased by 9.6% (p Z 0.002), 15.0% (p Z 0.004), and 14.6% (p Z 0.004) on average, respectively. The average perturbed D 99% doses to the ITV, compared to the nominal plan, decreased by 7.8% (6X-FFF) and 6.5% (10X-FFF) for the PTV-based optimized plan, and 7.3% (6X-FFF) and 5.8% (10X-FFF) for the robust optimized plan. The standard deviation of the D 99% dose to the ITV for the PTV-based plan with the 6X-and 10X-FFF beams were 163.6 and 158.9 cGy, respectively. The standard deviations of D 99% to the ITV for the robust optimized plan with the 6X-and 10X-FFF beams were 138.9 and 128.5 cGy, respectively. Conclusion: 10X-FFF beam applied under a robust optimized plan for lung SBRT cancer treatment using a breath-holding technique is a feasible method, despite a slightly higher dose to the lung, compared to a 6X-FFF beam.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.