On mining clinical pathway patterns from medical behaviors

Latent Treatment Pattern Discovery for Clinical Processes

“…Clinical process analysis (CPA) has experienced increased attention over the years due to its importance to health-care management in general and to its usefulness for capturing the actionable knowledge to administrate, automate, and schedule the best practice for individual patients in their therapy and treatment processes [7][8][9][10]. This has resulted in various successful approaches that are capable of analyzing clinical activities in CPs.…”

“…The majority of research has focused on CP pattern discovery that aims to recognize what clinical activities are essential/critical for CPs, and also where temporal orders of these activities are quantified with numerical bounds. Research in CP pattern discovery exploits the fact by automatically measure clinical activities to aid clinical experts analyze and improve CPs [7,9,10].…”

“…Although many excellent studies have been proposed for CP pattern discovery [7,10,15], they assume prior knowledge about CP [16,17], using which treatment patterns are derived in a completely supervised manner. In addition, many approaches are based on the experiences and knowledge of clinical experts [18,19].…”

“…To this end, the methods using data mining and machine learning technologies to analyze CPs based on associated careflow logs are receiving gradual attentions in medical informatics [7,10,11]. These techniques are also called process mining [20][21][22][23].…”

“…However, the diversity of clinical activities in CPs is far higher than that of common business processes. The use of traditional process mining techniques may generate spaghettilike treatment patterns that are difficult to be comprehended by clinical experts [7,10], such incomprehensible patterns are either not amenable or lack of assistance to efforts of analysis and improvement of CPs. In addition, existing process mining algorithms often produce excessive volume of patterns that may overwhelm the analysts [7].…”