Description:

INV-21122

 

Background

Pressure ulcers are a localized injury to the skin or underlying tissue usually over a bony prominence as a result of pressure. Pressure ulcers have been reported as the third most costly and deadly complication after cancer and cardiovascular diseases. In the United States, each year about 2.5 million hospitalizations happen due to pressure ulcers and 60,000 deaths are reported because of this complication. Such statistics will be increasing at an alarming rate in the following years due to the aging population and orthopedic comorbidities. Various computer vision applications have been developed to observe and model in-bed pose and contact pressure. However, the in-bed pose estimation by computer vision applications has its own specialized aspects and comes with specific challenges, including the notable differences in lighting conditions throughout the day and having pose distribution different from the common human surveillance viewpoint. Therefore, novel pose estimation frameworks that can accurately predict in-bed pressure and mitigate the occurrence of pressure ulcers are needed.

 

Technology Overview

Researchers at Northeastern designed a contact-less pressure eye (PEye) approach that employs vision signals to infer high-resolution contact pressure between the human body and its lying surface. This invention includes a camera on the ceiling with a bird’s eye view and software, based on a pre‑trained model, to estimate the contact pressure. PEye avoids the sophisticated mechanical modeling of the contact pressure and solves the problem in an end-to-end manner. PEye’s network is developed with multi-stage dual encoding; it shares decoding structure with both visual and non-visual physical inputs and enforces physical law during supervision. Imaging can be in both formats of RGB and IR imaging. The use of IR imaging helps PEye to stay effective consistently even under complete darkness, which enables long-term in-bed behavior monitoring throughout the day with varying illumination states.

 

Benefits

  • Contactless
  • Easy to maintain
  • Low cost
  • Functional with different modalities

 

Applications

  • Patient monitoring for pressure ulcers

 

Opportunity

  • Research collaboration
  • Developmental partner
  • Licensing
Patent Information:
Category(s):
AI
Software
For Information, Contact:
Myron Kassaraba
Director of Commercialization
Northeastern University
m.kassaraba@northeastern.edu
Inventors:
Shuangjun Liu
Sarah Ostadabbas
Keywords:
bedsores
computer vision and deep learning
Medical/Health
pressure ulcer