Researchers have made a significant breakthrough in the treatment of urinary acid stones by utilizing a combination of radiomics and machine learning. By accurately predicting pure uric acid (pUA) urinary stones, this innovative approach allows for effective and non-invasive treatment using oral chemolysis.
The study, conducted between 2019 and 2021, focused on categorizing urinary stones into two groups: pure uric acid (pUA) and other compositions (Others) using non-contrast-enhanced computed tomography scans (NCCTs). By analyzing composition data from 576 samples obtained from 302 cases, researchers were able to identify 118 cases with pUA urinary stones.
Utilizing radiomics features extracted from NCCT scans, the research team employed the Least Absolute Shrinkage and Selection Operator to identify the ten most crucial characteristics for predicting pUA urinary stones. These features were then used to develop a Support Vector Machine model, which demonstrated high accuracy, sensitivity, specificity, and precision in predicting pUA stones on the testing dataset.
In conclusion, the machine learning system trained on radiomics features from NCCT scans has shown promise in reliably predicting pUA urinary stones. This research holds potential for aiding healthcare providers in selecting appropriate treatment strategies for stone disease, ultimately improving patient outcomes and quality of care.
