Researchers developed a brief, image-based rating tool that measures two key dimensions of food evaluation—liking and perceived healthiness—using standardized food photographs. Testing in adults with type 2 diabetes undergoing medically supervised dietary interventions, the tool demonstrated internal coherence and sensitivity to changes over time, with meaningful associations to expert ratings and physiological measures. This fills a critical gap in nutrition research, where most studies focus solely on food intake outcomes rather than the underlying psychological drivers of dietary behavior change. The tool's ability to capture coordinated within-person changes in food preferences could revolutionize how we understand and optimize nutritional interventions. Unlike traditional dietary assessment methods that rely on self-reported consumption, this approach directly measures the evaluative processes that precede food choices. The smartphone-compatible design makes it scalable for large studies and real-world clinical applications. However, this remains a preprint awaiting peer review, and the pilot tested only diabetes patients, limiting broader generalizability. The methodology represents an incremental but valuable advance, offering researchers and clinicians a practical way to track how interventions reshape food attitudes—a key predictor of long-term dietary adherence and metabolic health outcomes.