Subgroup Discovery for Test Selection: A Novel Approach and Its Application to Breast Cancer Diagnosis

Abstract

Abstract. We propose a new approach to test selection based on the discovery of subgroups of patients sharing the same optimal test, and present its application to breast cancer diagnosis. Subgroups are defined in terms of background information about the patient. We automatically determine the best t subgroups a patient belongs to, and decide for the test proposed by their majority. We introduce the concept of prediction quality to measure how accurate the test outcome is regarding the disease status. The quality of a subgroup is then the best mean prediction quality of its members (choosing the same test for all). Incorporating the quality computation in the search heuristic enables a significant reduction of the search space. In experiments on breast cancer diagnosis data we showed that it is faster than the baseline algorithm APRIORI-SD while preserving its accuracy.

Evolutionary fuzzy rule extraction for subgroup discovery in a psychiatric emergency department

“…The use of these techniques has been applied to develop a framework using data mining technologies that make it possible to automatically analyse huge clinical data sets and to discover patterns behind them (Masuda et al 2002), to reanalyse a published study of which variables predicted psychiatrists' decisions to hospitalise suicide attempters, who were assessed in the emergency department (Baca-García et al 2006), to provide an infrastructure for the highest quality research in psychiatric disorders, particularly in schizophrenia and schizo affective disturbances (Kielan et al 2004), or to obtain rules to guide doctors towards a good relation with their patients to improve the results or the psychiatric treatments (Aguilar- Ruiz et al 2004). SD has been successfully applied in different medical domains, including the detection of patient groups with risk of atherosclerotic coronary heart disease (Gamberger and Lavrac 2003), breast cancer diagnosis (López et al 2009;Mueller et al 2009), brain ischaemia data analysis Kralj et al 2007), profiling examiners for sonographic examinations (Atzmueller et al 2005), identification of interesting diagnostic patterns to supplement a medical documentation and consultation system (Atzmueller et al 2004), or scrutinizing blood glucose management guidelines (Nannings et al 2009). …”

Subgroup Discovery Algorithms: A Survey and Empirical Evaluation

“…Subgroup discovery has addressed different real-world problems in the bio-informatics domain. In [37,40], relevant features are extracted by implementing the SD subgroup discovery algorithm for the detection of different cancer types. [41] employs an SD approach SD4TS (Subgroup Discovery for Test Selection) for breast cancer diagnosis.…”

Anytime discovery of a diverse set of patterns with Monte Carlo tree search

“…Beam search performs a level-wise exploration of the search space: A beam of a given size (or dynamic size for recent work) is built from the root of the search space. This beam only keeps the most promising subgroups to extend at each level [ [36,44,53]]. The redundancy issue due to the beam search is tackled with the pattern skyline paradigm by [54], and with a ROC-based beam search variant for SD by [42].…”