Understanding Population Health and Readiness through Visual Analytics

The assessment of neuropsychological conditions as well as the diagnosis of many of the injuries commonly seen within the military involves the understanding, correlation, and evaluation of a large collection of clinical variables. Physicians and clinical staff are often faced with the challenging task of correlating many clinical evaluations and diagnostic tests to determine the extent of an injury, condition, or disease. For instance, in many situations clinicians are confronted with a rich set of multi-modal measurements that range from psychological evaluations, imaging studies, neuroendocrinology tests, psychiatric assessments, and physical evaluations. Then, based on the associations between those variables, clinicians must pick a specific treatment plan or determine the readiness of the service members. Currently, most clinicians use a small subset of neuropsychology measurements because the step of combining and analyzing the clinical data quickly becomes overwhelming and difficult to comprehend.

This proposal addresses the clear gap between the acquisition of clinical measurements and the diagnostic step by providing an intuitive, flexible, and customizable interactive visualization framework. The proposed system will help clinicians obtain new insights about the underlying conditions of the patient and will provide a software tool that MHS can distribute to researchers or clinicians so they can visually study the associations between individual clinical assessment techniques. The set of tools proposed will be particularly useful for: (i) identifying and categorizing patterns or clusters of symptoms across multiple disciplines that can be used to better determine the readiness of individual service members; (ii) categorizing individual patients within a particular cluster of symptoms to better understand the population as well as the effectiveness of different treatment techniques, (iii) summarizing large amount of data within a single application, thus reducing the frequency in which additional evaluations are requested because specific measurements or data could not be found in AHLTA.

Try it out