AI for Health - Hernia Detection
Detecting hernia on X-ray images in collaboration with hospital radiologists.
Premium participation: Get 1-on-1 Mentoring & GuidanceTry a free mentoring session
Detecting hernia on radiology X-ray images
When a cervical disc is herniated, the spinal cord and/or nerves exiting the spine can become compressed. This can result in several symptoms for patients, such as pain that radiates from the upper arm all the way through the fingers, numbness and/ or weakness in all limbs, and in some cases difficulty with their gait. The most common treatment is the surgical removal of the disc tissue, called Anterior Cervical Discectomy and Fusion (ACDF).
The Zuyderland Medical Centre in Heerlen & Sittard-Geleen currently performs the ACDF surgery on their patients. It is however not yet known what ‘normal’ movement of the spine exactly is. The Centre has tried to quantify and analyze the differences between the Range of Motion (ROM) of the normal cervical spine, and the spine after the ACDF and prosthesis procedure. However, the process of segmenting the vertebrae has been done manually. Given a total of 52 frames per patient, this takes up an enormous amount of time for the experts, while not really performing an immediate expert task.
The neuroscientists and medical experts would like your help to investigate it, to analyze normal and abnormal motion of the neck to relate it to disease in order to optimize surgical techniques in the long term. This helps automate the process, which makes it faster and to allow larger groups of patients to be analyzed, which gives more scientific evidence.
Our challenge partner
GOAL: Develop an algorithm that automatically segments and tracks the cervical vertebrae over time
Analyze (ab)normal motion of the neck through X-ray images and the influence on the development of complaints. Based on cervical X-ray images, develop an ML model that can automatically segment and track the cervical vertebrae over time, quantifying and analyzing the differences between the Range of Motion (ROM) of the normal cervical spine, and the spine after the ACDF and prosthesis procedure.
Skills & Technologies you will (learn to) use:
- Medical image analysis
- Image segmentation to identify the state of the neck
Who are we looking for?
Anyone with an interest in artificial intelligence or machine learning can apply. We expect some experience with programming languages. Experience with supervised learning methods, convolutional neural networks, or medical image analysis is welcome, but anyone with the right motivation and ‘proof’ of understanding of the concepts discussed in this proposal can sign up!
You can join as a contributor (12 hours per week commitment for 2 months), coach (2-4 hours per week, only for experienced ML professionals) and teacher (give one relevant ML / domain masterclass).
To achieve this goal, our project will be guided by an expert from the Zuyderland Medical Centre and FruitPunch AI. The project will kick off in the second week of October and the team will consist of 15 people that will work on the project for approximately 12 hours a week for 3 months.
To use the technologies and skill required for this project, you’ll be aided by the chapter advisors: Carmen van Vilsteren, Chair Topsector Life Sciences & Health and Director of e/MTIC, Fons van der Sommen, Assistant professor computer vision & image processing at TU/e and Nicky Hekster, teacher and trainer for AI in Healthcare & Life Sciences.
Did you know
- A herniated cervical disc can cause several symptoms such as numbness and/or weakness in all limbs.
- The most common treatment for a herniated cervical disc is to remove the disc tissue by a surgery. This operation is called Anterior Cervical Discectomy and Fusion (ACDF).
- ACDF will cause fusion of two spinal vertebrae, which will of course have consequences for the range of motion (ROM). Due to this limited movement, more load will have to be carried by the spine below and above the fusion, resulting in another herniated disc in the long term.
- It is thought that the secondary problems that are caused by ACDF may be prevented by placing a motion-sparing prosthesis instead of a fusion-inducing cage.
Info Session: 11 October 2021
Presenter: Valérie Schuermans PhD, neuroscientist at Zuyderland
Application deadline: 13 October 2021
Challenge Kick-off: 15 October 2021
Midterm Presentations: 3 November 2021
Final presentation: 28 March 2022