ABSTRACT
OBJECTIVE: To extract texture features from vocal cord leukoplakia (VCL) images and establish a VCL risk stratification prediction model using machine learning (ML) techniques. METHODS: A total of 462 patients with pathologically confirmed VCL were retrospectively collected and divided into low-risk and high-risk groups. We use a 5-fold cross validation method to ensure the generalization ability of the model built using the included dataset and avoid overfitting. Totally 504 texture features were extracted from each laryngoscope image. After feature selection, 10 ML classifiers were utilized to construct the model. The SHapley Additive exPlanations (SHAP) was employed for feature analysis. To evaluate the model, accuracy, sensitivity, specificity, and the area under the receiver operating characteristic (ROC) curve (AUC) were utilized. In addition, the model was transformed into an online application for public use and further tested in an independent dataset with 52 cases of VCL. RESULTS: A total of 12 features were finally selected, random forest (RF) achieved the best model performance, the mean accuracy, sensitivity, specificity, and AUC of the 5-fold cross validation were 92.2 ± 4.1%, 95.6 ± 4.0%, 85.8 ± 5.8%, and 90.7 ± 4.9%, respectively. The result is much higher than the clinicians (AUC between 63.1% and 75.2%). The SHAP algorithm ranks the importance of 12 texture features to the model. The test results of the additional independent datasets were 92.3%, 95.7%, 90.0%, and 93.3%, respectively. CONCLUSION: The proposed VCL risk stratification prediction model, which has been developed into a public online prediction platform, may be applied in practical clinical work. LEVEL OF EVIDENCE: 3 Laryngoscope, 2024.
ABSTRACT
Arginase-1 catalyzes the conversion of arginine to ornithine and urea, which is the final step of the urea cycle used to remove excess ammonia from the body. Arginase-1 deficiency leads to hyperargininemia in mice and man with severe lethal consequences in the former and progressive neurological impairment to varying degrees in the latter. In a tamoxifen-induced arginase-1 deficient mouse model, mice succumb to the enzyme deficiency within 2 weeks after inducing the knockout and retain <2 % enzyme in the liver. Standard clinical care regimens for arginase-1 deficiency (low-protein diet, the nitrogen-scavenging drug sodium phenylbutyrate, ornithine supplementation) either failed to extend lifespan (ornithine) or only minimally prolonged lifespan (maximum 8 days with low-protein diet and drug). A conditional, tamoxifen-inducible arginase-1 transgenic mouse strain expressing the enzyme from the Rosa26 locus modestly extended lifespan of neonatal mice, but not that of 4-week old mice, when crossed to the inducible arginase-1 knockout mouse strain. Delivery of an arginase-1/enhanced green fluorescent fusion construct by adeno-associated viral delivery (rh10 serotype with a strong cytomegalovirus-chicken ß-actin hybrid promoter) rescued about 30% of male mice with lifespan prolongation to at least 6 months, extensive hepatic expression and restoration of significant enzyme activity in liver. In contrast, a vector of the AAV8 serotype driven by the thyroxine-binding globulin promoter led to weaker liver expression and did not rescue arginase-1 deficient mice to any great extent. Since the induced arginase-1 deficient mouse model displays a much more severe phenotype when compared to human arginase-1 deficiency, these studies reveal that it may be feasible with gene therapy strategies to correct the various manifestations of the disorder and they provide optimism for future clinical studies.
