Tuber formosanum is a valuable edible and medicinal fungus, yet systematic research on the efficient and green extraction of its polyphenolic compounds remains limited. This study aims to establish an optimized ultrasound-assisted extraction (UAE) process for polyphenols from T. formosanum. UAE was conducted using a KS-700XDS power-adjustable ultrasonic bath at 50 kHz. Initially, UAE was validated as the most efficient method among several conventional techniques, achieving a 14.7% higher yield than heating reflux extraction while consuming 7.5-fold less energy. Subsequently, Lasso regression identified ultrasonic power, time, temperature, and ethanol concentration as the critical variables influencing total polyphenol content (TPC). Through response surface methodology (RSM), the optimal extraction conditions were determined: 95 min, 210 W, 70 ℃, and 60% ethanol. Under these parameters, the TPC reached 2.72 mg/g, representing a 16.24% improvement over pre-optimization levels. To elucidate the contribution patterns of each factor, SHAP analysis was employed, revealing that ultrasonic power was the dominant factor, while time and temperature exhibited unimodal optimal contribution trends, corroborating the RSM findings. Antioxidant assays confirmed that the extract obtained under optimized conditions demonstrated significantly enhanced DPPH and ABTS radical scavenging activity and ferric reducing power compared to unoptimized extracts. This study not only establishes an efficient UAE protocol for T. formosanum polyphenols but also provides a novel methodological framework integrating green extraction technology with interpretable machine learning for the sustainable valorization of bioactive compounds from rare fungal resources.
Keywords: Antioxidant; Polyphenols; SHAP analysis; Tuber formosanum; Ultrasound-assisted extraction.
Copyright © 2026 The Author(s). Published by Elsevier B.V. All rights reserved.