Pressure ulcers with Artificial Intelligence

// student

// Okategoriserade

The field of Artificial Intelligence (AI) has increased extensively during the last decade. Several ongoing projects already utilize the power of AI to solve different problems. However, computational calculations bring additional cost since it often needs a lot of data in order to train the networks, which requires a powerful GPU. Thus, it is interesting to investigate how networks can be optimized.

This thesis can be done in our office in Gothenburg, Stockholm or Sundsvall

High level description
AI within healthcare is a topic that is increasing. Pressure ulcers are common problems for older people who are hospitalized. This problem is also common at home. A pressure ulcer is caused by sustained pressure being placed on a particular part of the body. This pressure interrupts the blood supply to the affected area of the skin.

Therefore, it would be of interest to investigate and implement a neural network or a machine learning algorithm that could decide if or when a pressure ulcer is occurring. You will also investigate how you can generate a dataset that can be used for your algorithms. For instance, body temperature, current time, age, gender, length, and weight are only examples of features that can be used in order to generate a dataset. It is up to you which features you decide to use. It is also up to you what kind of AI method you choose.

Project Description
Investigate and implement a suitable solution that can predict if or when a pressure ulcer occurs. You need to evaluate algorithms and networks and provide the accuracy of the different models. How often do they make correct decisions? An interesting topic would also be to investigate the minimum size of training data to get the desired result.

Who are we looking for?
We are looking for someone interested in and knowledgeable of AI and machine learning technologies. This work is suitable for those of you who are studying computer engineering or equivalent at the bachelor's or master's level.

Purpose
The aim of this project is to investigate and implement suitable machine and deep learning technologies in order to predict if or when pressure ulcers occur. More information will be given during the interviews.

Important information to you who will apply for a thesis project

  • The Thesis is to be done during the spring term of 2023
  • If you would like to co-write your thesis, both of you need to apply separately and conduct the tests. Just make sure to inform us in your application that you would like to write in pairs and who your thesis partner is.
  • We will update you regarding the process via e-mail so make sure to stay tuned

Related posts

// Okategoriserade

Pressure ulcers with Artificial Intelligence

The field of Artificial Intelligence (AI) has increased extensively during the last decade. Several ongoing projects already utilize the power of AI to solve different problems. However, computational calculations bring additional cost since it often needs a lot of data in order to train the networks, which requires a powerful GPU. Thus, it is interesting to investigate how networks can be optimized.

Read more
// Okategoriserade

Pressure ulcers with Artificial Intelligence

The field of Artificial Intelligence (AI) has increased extensively during the last decade. Several ongoing projects already utilize the power of AI to solve different problems. However, computational calculations bring additional cost since it often needs a lot of data in order to train the networks, which requires a powerful GPU. Thus, it is interesting to investigate how networks can be optimized.

Read more
// Okategoriserade

Pressure ulcers with Artificial Intelligence

The field of Artificial Intelligence (AI) has increased extensively during the last decade. Several ongoing projects already utilize the power of AI to solve different problems. However, computational calculations bring additional cost since it often needs a lot of data in order to train the networks, which requires a powerful GPU. Thus, it is interesting to investigate how networks can be optimized.

Read more