Interpretability on clinical analysis from Pattern Disentanglement Insight

Anonymous

Abstract: Diagnosis of a clinical condition can help medical professionals save time in making a clinical diagnosis and prevent overlooking risks. Therefore we explore the problem using free-text medical notes recorded in electronic health records (EHR). MIMIC III is a de-identified EHR database containing observations from over 40,000 patients in critical care units. Since the text corpus is unstructured in the non-database table format, existing machine learning models may be ineffective at interpreting the results, which is often desirable for clinical diagnosis. Hence, in this paper, we proposed a text mining and pattern discovery solution to discover strong association patterns from patient discharge summaries and the code of international classification of diseases (ICD9-code). The proposed approach offers a straightforward interpretation of the underlying relation of patient characteristics in an unsupervised machine learning setting. The clustering results outperform the baseline clustering algorithm and are comparable to baseline supervised methods.