Internship – AI Applied to CT scans M/F

at Median Technologies
Published February 2, 2023
Location Valbonne, France
Category Deep Learning  
Job Type Internship  

Description

Company presentation

Since 2002, Median Technologies has been expanding the boundaries of the identification, interpretation, analysis and reporting of imaging data in the medical world. Our core activity is to develop advanced imaging software solutions and platforms for clinical drug development in oncology, diagnostic support, and cancer patient care. Our software solutions improve the management of cancer patients by helping to better identify pathologies, develop and select patient-specific therapies (precision medicine).

The company employs a highly-qualified team and leverages its scientific, technical, medical, and regulatory expertise to develop innovative medical imaging analysis software based on Artificial Intelligence, cloud computing and big data. We are driven by our core values that are essential to us. These values define who we are, what we do, the way we do it, and what we, as Median, aspire to:
• Leading innovation with purpose
• Committing to quality in all we do
• Supporting our customers in achieving their goals
• Always remembering to put the patient first

Today, we are a team of more than 220 people spread worldwide in the US, Europe and China. Our company is growing in a fulfilling international and multicultural environment.

Job description

In the context of extended screening programs, the use of automated tools allowing radiologists to read medical images faster will become more and more relied on. Median Technologies has been working for years on AI applied to medical imaging. In particular, our
Lung Cancer Screening product aims at helping clinicians spot early lesions in the lung that can possibly constitute early-stage cancers based on CT scans.

This product actually involves many different components, whether it be detecting nodules in CT scans or determining how malignant they may be. In-depth characterization of detected nodules can help improve such performance and comply with industry standards, with criteria such as nodule calcification, nodule texture and so on.

The goal of this internship is to develop and test methods that efficiently learn in-depth characterization of detected nodules in order to improve the malignancy assessment. Our interest is in particular focused on methods based on hierarchical models.

Assignments

Your internship will consist in achieving the following tasks:

  1. Bibliography: You will first write a short bibliography on research works introducing models for nodule in-depth description and characterization, with a particular focus on hierarchical models;
  2. Model Design and Implementation: You will then design and implement a model predicting in-depth nodule characterization to improve our current malignancy assessment. You will use Git to share your work with the team;
  3. Model Test and Reporting: Your final goal will be to test your model on in-house data, and write a detailed report on how learning nodule descriptors actually improves the malignancy assessment.

Profile required

Skills required:

You are expected to have experience with the following libraries/technologies: at least one Deep Learning framework based on Python (preferably PyTorch); Git; Pandas; NumPy.

You are also expected to be comfortable with reading scientific papers. Having already published one of several papers would be appreciated. You are expected to be comfortable with the following concepts: Deep Learning; Linear Algebra; Statistics and Probabilities; Multivariate Calculus; Gradient Descent and its variants. Experience with 3D medical image processing and/or image segmentation would be appreciated.

Education: in the process of obtaining a Master’s Degree or a Computer Science Degree in any field related to Deep Learning.
Qualities: Good communication in English, spoken and written. French would be a plus. Curiosity, open-mindedness, determination and persistence, patience

Drop files here browse files ...