Radiotherapy (RT) is a major component in the treatment of most head and neck cancer (HNC) cases. Duringirradiation, sensitive regions such as the salivary glands can sustain injury, resulting in xerostomia (dry mouth).This side effect is common and can significantly reduce quality of life during and post-treatment. The focus ofthis application is prediction during treatment planning of whether patients will suffer high-grade xerostomia(NCI CTCAE Grade 2-3) at the time of their first post-treatment follow-up visit, typically 3-6 months after RT(prevalence is approximately 40%). Predictions will enable clinicians to carry out treatment planning withimproved knowledge of the likelihood of high-grade xerostomia development and allow better-informed and moretimely anticipation of consequences such as eating difficulty.In this Phase 1 project, Oncospace Inc. will develop a Classification and Regression Tree (CART) predictionmodel using over 1200 complete HNC patient records. Associations between high-grade xerostomia and a widerange of dosimetric, clinical and demographic features will be automatically discovered and the features with thestrongest associations will populate the nodes of a decision tree. The terminal leaf nodes will each contain theprobability of high-grade xerostomia for the subset of patients in that node. In addition, leaf nodes will be assignedbinary class labels designating a high- or low risk of high-grade xerostomia. This type of model providestransparency and interpretability, which are beneficial for clinical acceptance and for demonstration of safety toregulatory agencies. The software will be built using the Microsoft Azure cloud architecture and be deployed viaa Software as a Service (SaaS) model.There are three distinct aims of this project: 1. Populate Oncospace Inc.'s Microsoft Azure CosmosDB database with data licensed from Johns Hopkins University, including steps such as patient de-identification, data curation, and additional dataset featureengineering 2. Perform CART modeling and test model accuracy, using separate training and test datasets and a variety of performance metrics, including sensitivity, specificity, AUC, and F1-score. 3. Design a clinically acceptable risk classification strategy and a user interface (UI) to communicate model results. Expert input from a team of UI consultants and three radiation oncologists will be an integral part of the development, testing, and evaluation processes.The successful completion of these aims will demonstrate the clinical and commercial feasibility of a xerostomiaprediction model for HNC. Further development in Phase 2 will include deeper model personalization viaincorporation of advanced image features (radiomics), as well as validation of model generalizability andcommercial viability via the curation and use in model building of data from other institutions.Oncospace, formed in 2018, is uniquely positioned to carry out this work as the team includes the creators of thePinnacle radiation therapy planning system, Tomotherapy radiation treatment delivery system, and HealthMyneQuantitative Imaging Decision Support platform. Oncospace has close clinical collaboration with Johns HopkinsUniversity (JHU) for clinical feedback, validation and initial deployment. Oncospace has licensed three patentsand subscription to complete patient treatment records for over 6,000 radiation oncology patients from JHU. Thecompany has won the Microsoft Innovation Acceleration Award for its innovative platform to deliver AI-enabledhealthcare solutions to the radiation oncology community.Page 1 of 1 Narrative
During standard-of-care radiotherapy for head and neck cancer, the salivary glands often sustain radiation-
induced injury leading to xerostomia (dry mouth). Machine learning-based predictions can be used to help create
treatment plans that minimize the chance of severe xerostomia, and to better anticipate and manage its
occurrence. The approach described in this application leverages a large database of head and neck cancer
patients, including dosimetric, clinical and demographic features, to make well-informed predictions. Acceleration ; Aftercare ; After Care ; After-Treatment ; post treatment ; Anatomy ; Anatomic ; Anatomic Sites ; Anatomic structures ; Anatomical Sciences ; Architecture ; Engineering / Architecture ; Award ; Classification ; Systematics ; Decision Trees ; Disease ; Disorder ; Eating ; Food Intake ; Engineering ; Feedback ; Judgment ; Medical Records ; Medicine ; Morbidity - disease rate ; Morbidity ; Legal patent ; Patents ; Patients ; Probability ; Quality of life ; QOL ; Radiation therapy ; Radiotherapeutics ; Radiotherapy ; radiation treatment ; radio-therapy ; treatment with radiation ; Records ; Risk ; ROC Curve ; ROC Analyses ; receiver operating characteristic analyses ; receiver operating characteristic curve ; Safety ; Salivary Glands ; Salivary Glands Head and Neck ; Sensitivity and Specificity ; Computer software ; Software ; Technology ; Testing ; Time ; Universities ; Work ; Xerostomia ; Asialia ; Hyposalivation ; Mouth Dryness ; aptyalism ; dry mouth ; Healthcare ; health care ; Data Set ; Dataset ; Guidelines ; Injury ; injuries ; base ; Organ ; Label ; improved ; Site ; Clinical ; Phase ; Randomized Clinical Trials ; Ensure ; Evaluation ; Training ; Databases ; Data Bases ; data base ; Licensing ; Plant Leaves ; leaf ; Radiation Oncology ; Head and Neck Cancer ; Malignant Head and Neck Neoplasm ; head/neck cancer ; malignant head and neck tumor ; Collaborations ; machine learned ; Machine Learning ; Knowledge ; Event ; System ; Best Practice Analysis ; Benchmarking ; Visit ; Performance ; treatment planning ; HIPAA ; Kennedy Kassebaum Act ; PL 104-191 ; PL104-191 ; Public Law 104-191 ; United States Health Insurance Portability and Accountability Act ; Health Insurance Portability and Accountability Act ; Toxicities ; Toxic effect ; Agreement ; Reporting ; Position ; Positioning Attribute ; Malignant Neck Neoplasm ; malignant neck tumor ; Neck Cancer ; Head Cancer ; Radiation ; Modeling ; Cancer Radiotherapy ; cancer radiation therapy ; Cancer Treatment ; Malignant Neoplasm Therapy ; Malignant Neoplasm Treatment ; anti-cancer therapy ; anticancer therapy ; cancer-directed therapy ; cancer therapy ; Institution ; irradiation ; Gland ; preventing ; prevent ; Dose ; Data ; Economic Burden ; Radiation Oncologist ; Common Terminology Criteria for Adverse Events ; CTCAE ; Common Toxicity Criteria ; Community Clinical Oncology Program ; CCOP ; Community Oncology ; Small Business Innovation Research Grant ; SBIR ; Small Business Innovation Research ; Validation ; Process ; follow-up ; Active Follow-up ; active followup ; follow up ; followed up ; followup ; Development ; developmental ; Image ; imaging ; predictive modeling ; computer based prediction ; prediction model ; design ; designing ; Outcome ; head and neck cancer patient ; HNC patient ; Population ; Prevalence ; innovation ; innovate ; innovative ; clinical application ; clinical applicability ; inclusion criteria ; commercialization ; high risk ; standard of care ; product development ; cloud based ; quantitative imaging ; model building ; radiation-induced injury ; irradiation-induced injury ; Data Science ; individual patient ; patient subsets ; patient subgroups ; patient subpopulations ; patient subtypes ; software as a service ; radiomics ; clinical decision support ; side effect ; classification trees ; regression trees ; data curation ;
