健保局為達到醫療資源的合理分配及醫療成本妥善控管等目的，而在其醫療支付制度設計有配套之專業審查程序。目前的專業審查作業均以耗費大量專業人力與時間的人工作業方式進行，實有自動化的必要。本研究將模糊理論、遺傳演算法和類神經網路適當整合，提出一個混合型軟式計算系統來協助判斷「健保醫令申報」、「疾病分類編碼」與「醫療申報費用」間的異常關聯，以達成健保醫療費用專業審查自動化之目的。所提出系統經以真實案例驗證後顯示它確實可提昇審查作業之效率，並能協助醫療院所改善疾病分類編碼品質，進而提昇醫療服務品質。

“Peer review” is an important screening and managerial procedure that the National Health Insurance Bureau (NHIB) used to help control the healthcare spending in a reasonable level. Currently, peer review is still a labor-intensive and costly task for all its participants because none of them could afford the high developing cost for those automatic peer review solutions. In this research, we proposed a hybrid soft computing system by integrating artificial neural network, genetic algorithm and fuzzy system as a low-cost and easy-to-develop alternative to build intelligent peer review systems for monitoring healthcare insurance claim automatically. Implementation and evaluation results indicate that the proposed hybrid model is not only a practical approach with high accuracy but also is ready to be used to enhance the quality of healthcare management.