[Soil Cadmium Prediction and Health Risk Assessment of an Oasis on the Eastern Edge of the Tarim Basin Based on Feature Optimization and Machine Learning].

Soil heavy metal pollution poses a serious threat to food security, human health, and soil ecosystems. Based on 644 soil samples collected from a typical oasis located at the eastern margin of the Tarim Basin, a series of models, namely, multiple linear regression （LR）, neural network （BP）, random forest （RF）, support vector machine （SVM）, and radial basis function （RBF）, were built to predict the soil heavy metal content. The optimal prediction result was obtained and utilized to analyze the spatial distribution features of heavy metal contamination and relevant health risks. The outcomes demonstrated that： ① The average Cd content in the study area was 0.14 mg·kg-1, which was 1.17 times the soil background value of Xinjiang, making it the primary factor of soil heavy metal contamination in the area. Additionally, the carcinogenicity risk coefficients of Cd for both adults and children were less than 10-4, indicating that there were no significant long-term health risks for humans in the area. ② The estimation accuracies of the five inversion models were compared, and the validation set of the RF model had an R2 value of 0.763 7, which was the highest among the five models. Additionally, the RMSE, MAE, and MBE of the RF model were the smallest among the five models. Therefore, the predicted values of the RF model were most consistent with the measured values of the soil Cd content. The predicted map of soil Cd distribution derived from the RF model coincided best with the interpolation map. ③ The RF model outperformed the other four models in predicting health risks associated with the soil Cd element for both adults and children, resulting in better prediction results. Comparatively, the predicted values of the LR model in the validation set varied greatly, leading to unreliable results. It was demonstrated that the RF was the best model for predicting soil Cd content and evaluating health risks in the study area, considering its superior generalization capability and anti-overfitting ability.

Comparison of four markers of hepatic fibrosis and hepatic function indices in patients with liver cirrhosis and hepatoma.

BACKGROUND: The present study aimed to compare four hepatic fibrosis markers [i.e., hyaluronic acid (HA), laminin (LN), procollagen III N-terminal peptide (PIIINP), and collagen type IV (CIV)] and 16 hepatic function indices in patients with liver cirrhosis of varying etiology. METHODS: The hepatic function indices and hepatic fibrosis markers were measured in 108 patients with liver cirrhosis and hepatoma using an automatic biochemical analyzer and luminescent immune analyzer. Twenty healthy controls were enrolled to compare the differences between liver cirrhosis and hepatoma of varying etiology and to analyze the correlations between the hepatic function indices and fibrosis markers. RESULTS: There was no correlation between alanine aminotransferase (ALT), total protein (TP), alkaline phosphatase (ALP), or the four markers of hepatic fibrosis in liver cirrhosis caused by hepatitis B (P>0.05). Aspartate aminotransferase (AST) was positively correlated with HA (r=0.428, P=0.007), LN (r=0.458, P=0.004), and CIV (r=0.374, P=0.021). Total bilirubin (TBIL) and direct bilirubin (DBIL) were positively correlated with LN (TBIL: r=0.480, P=0.002; DBIL: r=0.457, P=0.004), PIIINP (TBIL: r=0.380, P=0.017; DBIL: r=0.406, P=0.011), and CIV (TBIL: r=0.415, P=0.010; DBIL: r=0.400, P=0.013). Total bile acid (TBA) and γ-glutamyltranspeptidase (GGT) were positively correlated with PIIINP (TBA: r=0.363, P=0.025; GGT: r=0.353, P=0.029) and CIV (TBA: r=0.419, P=0.009; GGT: r=0.335, P=0.040). Leucine aminopeptidase (LAP) was positively correlated with LN (r=0.482, P=0.002). Cholinesterase (CHE) (HA: r=-0.452, P=0.004, LN: r=-0.336, P=0.039; PIIINP: r=-0.468, P=0.003; CIV: r=-0.485, P=0.002), prealbumin (PA) (HA: r=-0.575, P=0.000, LN: r=-0.413, P=0.010; PIIINP: r=-0.344, P=0.035; CIV: r=-0.371, P=0.022), albumin (ALB) (HA: r=-0.541, P=0.000, LN: r=-0.373, P=0.021; PIIINP: r=-0.353, P=0.030; CIV: r=-0.415, P=0.010), and superoxide dismutase (SOD) (HA: r=-0.334, P=0.040, LN: r=-0.347, P=0.033; PIIINP: r=-0.487, P=0.002; CIV: r=-0.536, P=0.001) were negatively correlated with the four markers of hepatic fibrosis. There was no correlation between ALT, AST, TBIL, TP, ALP, GGT, or the four hepatic fibrosis markers in hepatoma caused by hepatitis B (P>0.05). Meanwhile, DBIL and TBA were positively correlated with CIV (DBIL: r=0.519, P=0.023; TBA: r=0.563, P=0.012), while CHE (r=-0.604, P=0.006), ALB (r=-0.564, P=0.012), and SOD (r=-0.489, P=0.034) were negatively correlated with CIV. Moreover, PA was negatively correlated with LN (r=-0.510, P=0.026) and CIV (r=-0.696, P=0.001). CONCLUSIONS: The concentrations of the serological indices differed significantly based on the specific liver cirrhosis etiology. There was a strong correlation between the hepatic function indices and four hepatic fibrosis markers. Thus, the detection of these markers might improve the diagnosis and treatment of hepatoma.