Visual Verification of Cancer Staging for Therapy Decision Support

Abstract

It is generally accepted practice that each cancer patient case should be discussed in a clinical expert meeting, the so‐called tumor board. A central role in finding the best therapy options for patients with solid tumors plays the Tumor, lymph Node, and Metastasis staging (TNM staging). Correctness of TNM staging has a significant impact on the therapy choice and hence on the patient's post‐therapeutic quality of life or even survival. If inconsistencies in the TNM staging occur, possible explanations and solutions must be found based on the complex patient records, which takes the costly time of (multiple) physicians. We propose a more efficient visual analysis component, which supports a physician in verifying the given TNM staging before forwarding it to the tumor board. Our component comprises a Bayesian network model of the TNM staging process. Using information from the patient records and Bayesian inference, the models computes a patient‐specific TNM staging, which is then explored and compared to the given staging by means of a graph‐based visualization. Our component is implemented in a research prototype that supports an understanding of the model computations, allows for a fast identification of important influencing factors, and facilitates a quick detection of differences between two TNM stagings. We evaluated our component with five physicians, each studying 20 cases of laryngeal cancer.

Exploring the limits of complexity: A survey of empirical studies on graph visualisation

“…Nine additional studies use graphs with more than 200 nodes. Similar to the above, some of these studies only show subparts of the network [34,82,101,133], while some evaluate tools that scale well with large networks [26,27,34,52,53,59,82,133]. Four out of these nine studies have also used networks with less than 100 nodes.…”

Der intelligente HNO-Operationssaal der Zukunft

“…In a clinical evaluation trial, this software was analyzed retrospectively with 20 patient datasets with 2 calculated BN each, applying original and manipulated TNM classifications. The results of the study could reveal that the developed visualization software allows verifying the patient's case in an appropriate timeframe and reducing the probability of inexact (non-helpful) data due to an improved transparency and verifiability [12,48]. Overall, this approach presented the technical feasibility and also the possible clinical integration of digital patient models for supporting therapy decision making in (preoperative) interdisciplinary tumor boards based on Bayesian Networks, which is also confirmed in the literature [62][63][64].…”

“…However, it is difficult to achieve this objective because of the quantity and variety of collected patient data and their fragmented storing with different media as well as the multitude of diversified therapeutic options. In the scientific experimental stage, projects of the ICCAS Leipzig address this research area by modelling decision processes and the development of systems supporting the decision making processes, patient-specific therapy process models, methods for extraction and structuring of relevant information from patient files, and standardized information models [4,[11][12][13]. In the context of working on a digital patient model for decision making (introduction into projects regarding artificial intelligence), laryngeal cancer was chosen as ENT-related example because a sufficient complexity could be expected for the model to be created.…”

The State of the Art in Enhancing Trust in Machine Learning Models with the Use of Visualizations

“…This topic class contains 2 out of 9 papers on topic analysis applications [CAA*19, KKZE20]. A common theme here is network visualization used for explaining Bayesian networks [CWS*17, VKA*18] and decision trees [vv11]. Other subtopics (which lead to research opportunities) are the exploration of behavior with regard to the decomposition of projections, showing the internal parts of ML models (and how classes are formed inside them), and the role of the user; they are all covered by our categorization presented in Section 6. Topic 3 – hyper‐parameters & reward.…”