• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site
National Research University Higher School of EconomicsStudent ThesesMachine Learning for Childhood Acute Lymphoblastic Leukemia Treatment Optimization in Subgroups

Student
Title
Supervisor
Faculty
Educational Programme
Final Grade
Year of Graduation
Natalia Korepanova
Machine Learning for Childhood Acute Lymphoblastic Leukemia Treatment Optimization in Subgroups
Data Science
(Master’s programme)
2016
Acute lymphoblastic leukemia (ALL) is one of the most frequent oncological diseases in childhood and adolescence. The development of treatment protocols for childhood ALL results in recovery of the most part of all patients. However, new attempts of treatment modification do not allow one to improve survival. All these protocols suppose the minimum treatment personalization, while it may become a key to improvement of survival and development of effective therapy.

This paper is devoted to the problem of identification of subgroups of patients where the difference in efficiency between compared treatment strategies is significant. It is considered that we have data on patients’ features and outcome of the therapy under the assumption of the random choice of treatment for every patient from the set of compared treatments. Usually, such data can be collected in randomized controlled clinical trials.

An overview of existing machine learning approaches to this problem is presented. Apart from that a new approach to subgroup identification and generation of hypotheses about the difference of treatments efficiency is proposed. This approach is based on using pattern structures for data description, generating only closed descriptions of subgroups, and finding subgroups with significant difference in efficiency by applying a version of Close-by-One algorithm.

The proposed approach is applied to the data of randomized controlled trials on treating childhood ALL. The results are compared to the results of application of other approaches to the same data. It is concluded that the proposed approach enables one to receive more subgroups with significant difference in efficiency of treatments. A measure of stability of the obtained subgroup descriptions is proposed. This measure is based on the distance between concepts in the lattice of pattern concepts.

Student Theses at HSE must be completed in accordance with the University Rules and Regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses