Modeling Risk of Radiation-Associated Malignant Progression in Familial NF2 with Machine Learning

Abstract

NF2 is a rare hereditary disorder that predisposes patients to benign tumors on the acoustic nerve called vestibular schwannomas. Radiotherapy is a non-invasive treatment that is sometimes recommended for patients who have developed problematic tumors, but it has proven very difficult to evaluate the risk that radiation promotes malignancy, and much more serious brain tumors. A large, long-term study on the after-effects of radiation therapy on patients with NF2 was recently carried out by a team based at the University of Manchester. Using data from this long-running study, we will develop a machine learning-based model that computes an individualized assessment of the risk of post-radiation malignancy. The clinical applications will be more precise guidance about treatment-related risk, allowing patients and clinicians to make personalized decisions to lower the chance of complications. This model will be packaged in an open-source plugin compatible with widely used medical imaging software, and the modelling approach will be communicated to medical physicists and neurosurgeons in peer-reviewed publications. If there is confirmed to be an increased risk of malignancy, and risk factors identified, these will be communicated to the scientific and neurosurgical communities directly. The timescale to delivering outcomes relevant to patients should be short, with clear recommendations and freely available software tools by the end of the project. This research will advance the care of patients with NF2 by ensuring that they receive the correct treatment and advice, and thus enjoy a lower risk of malignancy overall. The bioinformatic methods developed may also help to identify new genomic risk factors in future research.

Document Details

Document Type
DoD Grant Award
Publication Date
Jan 04, 2024
Source ID
HT94252310434

Entities

People

  • Chay Paterson

Organizations

  • United States Army
  • University of Manchester

Tags

Fields of Study

  • Medicine

Readers

  • Distributed Systems and Data Platform Development
  • Molecular and genetic basis of cancer.
  • Oncology

Technology Areas

  • AI & ML