Phonocardiogram (PCG) signals are valuable for intraoperative hemodynamic monitoring, but poor-quality signals can affect the reliability of subsequent analysis. This study identifies six types of abnormal PCG signals that can occur during surgery and proposes novel signal quality assessment (SQA) indicators for their classification. A retrospective dataset was collected at Asan Medical Center using an esophageal stethoscope and annotated by one clinician and two engineers into seven categories, including normal, yielding 718 segments from 126 patients. To evaluate the effectiveness of the SQA indicators, we performed leave-one-subject-out cross-validation using a logistic regression model, achieving a macro-average precision, recall, and F1-score of 0.92. Beyond assessing signal acceptability, the proposed method reports signal status and guides appropriate downstream actions to support continuous monitoring.