Search

Blood tests play an essential role in everyday medicine and are used by doctors in several diagnostic procedures. Moreover, this data is multivariate – and often some diseases, such as COVID-19, could have different symptom manifestations and outcomes. This study proposes a method of extracting useful information from blood tests using UMAP technique - Uniform Manifold Approximation and Projection for Dimension Reduction combined with DBSCAN clustering and statistical approaches. The analysis performed here indicates several clusters of infection prevalence varying between 2%–37%, showing that our procedure is indeed capable of finding different patterns. A possible explanation is that COVID-19 is not just a respiratory infection but a systemic disease with critical hematological implications, primarily on white-cell fractions, as indicated by relevant statistical test p-values in the range of 0.03–0.1. The novel analysis procedure proposed could be adopted in other data-sets of different illnesses to help researchers to discover new patterns of data that could be used in various diseases and contexts.

Purpose CyberKnife can perform fiducial-free lung tracking, even when the target is visible from only one of the double imaging view, thanks to the Lung-Optimized-Treatment package. However, the uncertainty along the blind view (out-of-plane direction) must be taken into account in margin definition. The goal was to determine the accuracy of target localization in 1-view treatments (33% of all lung cases at Centro Diagnostico Italiano). Methods The actual tumor position, stored in logfiles from 2-view treatments, was projected using a geometric relation onto both image planes for each of 26 patients’ data. In the same plots exhale and inhale planned tumor positions, extracted from TPS and based on breath-hold TC, were novelty represented preserving metric with respect to imaging center. Actual vs. planned positions were compared through a home-made software in Matlab able to calculate also the extra-margin to estimated ITV which would have been necessary to cover the 95% of target positions for the 95% of patient population in case of 1-view treatment. The program was validated using data from 1-view and 2-view lung treatments with XLT Phantom (CIRS). Results The validation test confirmed the reliability of the technique. Expansions were equal to 5 mm and 7 mm for lesions in superior and inferior lobe respectively. The method proved to overestimate margins due to imaging sampling algorithm, data dependence on 2-view cases and the small amount of patients. However, results showed clearly that biphasic TC acquired in breath-hold way was rarely a good model for breath amplitude and breath center position in relation to spine alignment center. Conclusions This graphic comparison method is a useful tool to check localization accuracy of CyberKnife system for 1-view treatments and identify movement error components. It could be employed in clinical workflow to get patient-related information for customized margin definition.