Role

We are seeking a motivated student currently enrolled in a Dutch university with a strong interest in deep learning and machine learning, preferably with a focus on computer vision applications. The ideal candidate will have experience in scientific programming, with Python proficiency as a requirement, and be familiar with deep learning frameworks such as PyTorch or TensorFlow. We value individuals who are eager to learn, possess a results-oriented attitude, and are fluent in English, as effective communication will be key to collaboration in our team. This role is a great opportunity for those looking to deepen their expertise in cutting-edge AI applications and contribute to impactful projects.

Description

CAD4TB

Delft developed Computer Aided Detection for TB (CAD4TB) to assist clinicians to detect tuberculosis (TB) in TB-screening projects and in hospital or clinics, automatically finding the abnormal chest X-ray (CXR) images to focus on, instead of reading all images. 

CAD4TB is a soft- and hardware product that takes a single frontal chest radiograph as input, in the form of a DICOM image, and produces several outputs: 

The output includes a quality assessment of the input image, a general abnormality score and heat map indicating possible areas of abnormalities, a TB score between 0 and 100 related to the likelihood the subject on the X-ray has TB and the Cardiac Thoracic Ratio  

CAD4TB is used on a daily basis in over 200 sites around the world, making it the most widely used Artificial Intelligence (AI) software tool in TB screening. One of the unique features of CAD4TB is the ability to run offline on a low-cost device (CAD4TBbox) making it relatively easy to implement and use. 

BabyChecker

BabyChecker is a smartphone-based software designed to make pregnancies safer in low-resource settings where a clinician and standard ultrasound is a scarce resource. We created BabyChecker to support the detection of risky pregnancies by community health workers and increase timely referrals.   

BabyChecker algorithms receives as an input sweeps from an Obstetric 6 sweep protocol to produce 3 different outcomes:  

  • Gestational Age 
  • Fetal position 
  • Placenta Location 
  • Multiple Pregnancies 

BabyChecker is currently in active development and being used in Africa and Central America. At the moment, we are among the few companies that already have a product with AI working and deployed in developing countries. The fact that it can run offline and, in the phone, makes it accessible to places where the need for this technology is greatest. 

LUS4TB

In recent years, the interest in using ultrasound to detect lung abnormalities or diseases such as pleural effusions, pneumonia, pneumothorax, pulmonary embolism, and pulmonary contusions has grown. Additionally, ultrasound devices have become highly portable and can be connected to laptops, tablets, or even smartphones. These point-of-care ultrasound (POCUS) devices have the advantages of radiation-free imaging, low costs, real-time images, and good portability for remote locations or critically ill patients.

Multiple studies have shown that lung US (LUS) can be used to accurately diagnose pneumonia in adults and even outperforms CXR in terms of sensitivity. However, in these studies, LUS was performed and interpreted by highly skilled sonographers or trained physicians, who are not widely available in low-resource settings. Therefore, we are investigating the development of a mobile application with a standardized acquisition protocol that guides a novice user in ultrasound acquisition and uses AI to analyze the acquired data.

Education background & skills

  • Must be enrolled in a Dutch university
  • Interest in deep learning/machine learning, preferably in the field of computer vision
  • Experience in scientific programming (Python is mandatory)
  • Experience in deep learning frameworks, PyTorch or TensorFlow.
  • Eager to learn and a results-oriented attitude
  • Fluent in English is a must

Why us?

The opportunity to work in an international, multi-cultural and ambitious team. We offer a dynamic work environment in which everyone is encouraged to learn, grow and take ownership. Delft is growing rapidly and with that we foster innovation and personal growth. As an intern at Delft you will be part of a team that is motivated by doing good and having a positive impact in the world. You will work side by side with professionals from all over the world, not only learning about your specific craft but also what it’s like working for an innovative company.

Additionally, will you receive a company laptop and internship fee.

Send your resume and cover letter to hr@delft.care. Questions regarding this position can be sent to the same address.