Longitudinal Spatial-Nonspatial Decision Support for Competing Outcomes in Head and Neck Cancer Therapy

点击翻译按钮获取中文摘要

Cancers that depend on the spatial location of the disease affect all ethnicities and age groups, accounting for significant mortality and therapy-related side effects. In one instance, over 50,000 new cases of head and neck squamous carcinomas are diagnosed each year in the United States, leading to large, rich repositories of patient data. For each of these cases, oncologists need to anticipate survival, oncologic, and toxicity outcomes associated with treatment strategies in order to select a treatment which balances efficacy and toxicity. However, despite the wealth of data available, in the clinic decision support for cancer treatment is rudimentary and incorporates only a handful of patient characteristics, largely due to a lack of computational methodology and tools. We propose to construct a novel statistical and computational methodology for longitudinal and personalized treatment decisions over time, with specific application to head and neck cancer therapy planning. Simultaneous incorporation of complex factors---such as radiation dose location with respect to radiosensitive organs or patient reported side effects affecting quality of life---into treatment decisions over the course of cancer therapy requires the development of novel methodology. This methodology is revolutionary in that it is the first in the field to include both imaging and nonimaging data, while taking into account large-scale biological and clinical correlates. The approach is innovative through its leverage of big data repositories and through its unique blend of computational modeling principles from bioengineering and computer science. These methods allow us to incorporate diverse data types and model competing outcomes. From a clinical perspective, this integrative approach is novel in the field of cancer therapy. The resulting clinical decision support methodology will mark a significant advance in biomedical computing because it will be able to identify, for the first time, actionable timepoints for therapy and toxicity modification, based on a patient’s characteristics and quality of life indicators. The empirically-derived treatment decision support methodology developed in this project has the potential to directly improve the standard of care and the quality of life of surviving patients with a grave, often fatal and debilitating illness.