Combining patient pathway visualisation with prediction outcomes for chemotherapy treatments

Abstract

The Edinburgh Cancer Centre (ECC) contains NHS Lothian cancer patient data from multiple resources. However, the lack of proxy between numerous scattered resources hinders the capability to use the information collected in a useful way. ECC data is very varied and includes patient characteristics (e.g., age, weight, height, gender), information on diagnosis (e.g., stage, site, comorbidities) and treatments (e.g., surgery, chemotherapy, radiotherapy). The visualisation of a fraction of ECC data in the form of a patient timeline can aid and enhance the process of observing and identifying the overall condition of the patient, as well as understand how it may compare with cohorts of patients with similar characteristics. We have previously developed machine learning models for predicting treatment outcomes for breast cancer patient data that have undergone chemotherapy. In this paper, we describe, examine, and propose a solution to connect all these aspects and provide a bridge for several resources. This will make it easier for clinicians and other healthcare professionals to support service planning, deliver better quality of care and consequently improve service outcome within NHS Lothian.