A novel approach for calculating food safety models and health risk assessments of potentially toxic elements (PTEs) in cow milk.

This study introduces the Milk Quality Index (MQI), a novel metric for assessing milk quality that utilizes machine learning to enhance predictive accuracy. Lead (Pb) levels (318 ± 185 mg/kg) exceeded safety limits, with chromium (Cr), aluminum (Al), and selenium (Se) also raising concerns in raw cow milk from southwestern Iran. The MQI classified 80 % of samples as 'Fair' (range: 3.95-7.03, mean: 5.46), with random forest (RF) modeling confirming Se, calcium (Ca), and magnesium (Mg) as key contributors. Health risk assessments revealed elevated noncarcinogenic (HI50th = 2.48) and carcinogenic (TCR50th = 2.39E-4) risks in children. At the same time, arsenic (As) and nickel (Ni) had the greatest impact on the HI and TCR, respectively. Approximately 96.78 % of children and 98.96 % of adults may be exposed to elevated carcinogenic risks, respectively. This approach highlights the importance of PTE monitoring in milk to enhance food safety and protect public health.