Stage-oe-small.jpg

Thema4685

Aus Aifbportal
Version vom 15. Oktober 2020, 08:45 Uhr von An7651 (Diskussion | Beiträge)
(Unterschied) ← Nächstältere Version | Aktuelle Version (Unterschied) | Nächstjüngere Version → (Unterschied)
Wechseln zu:Navigation, Suche



Artificial Intelligence: Federated Learning for Healthcare




Informationen zur Arbeit

Abschlussarbeitstyp: Bachelor, Master
Betreuer: Konstantin Pandl
Forschungsgruppe: Critical Information Infrastructures

Archivierungsnummer: 4685
Abschlussarbeitsstatus: Offen
Beginn: 15. Oktober 2020
Abgabe: unbekannt

Weitere Informationen

Background:

The deployment of machine learning (ML) in healthcare promises great potential, for example, to treat patients more accurately and faster. However, training ML models requires large amounts of data. Traditionally, training data is transferred to a central server, where the ML model is then trained. However, the healthcare industry does not allow such a setting where patient data is highly private. Therefore, a new form of ML is emerging: ML models are trained at the edge (i.e., in hospitals), and several ML models (i.e., from several hospitals) are then merged, thus increasing the ML model accuracy. This is also known as federated learning. In the thesis, you can work on several aspects of federated learning for health care (e.g., data infrastructure based on blockchain, actual ML). Your thesis can have a technical or sociotechnical focus. With your thesis, you contribute toward the deployment of trustworthy AI in healthcare!


Objectives:

Possible topics include, but are not limited to:

  • Evaluation of federated learning algorithms on different health care data sets
  • Integration of federated learning and blockchain
  • Research on the adoption of federated learning

This is an umbrella topic since topics of interest change rapidly. A specific topic will be selected during a first meeting.


Introductory literature:

  • Kairouz, P., McMahan, H. B., Avent, B., Bellet, A., Bennis, M., Bhagoji, A. N., ... & d'Oliveira, R. G. (2019). Advances and open problems in federated learning. arXiv preprint arXiv:1912.04977.
  • Li, Tian, et al. "Federated learning: Challenges, methods, and future directions." IEEE Signal Processing Magazine 37.3 (2020): 50-60.
  • Vaid, A., Jaladanki, S. K., Xu, J., Teng, S., Kumar, A., Lee, S., ... & Johnson, K. W. (2020). Federated Learning of Electronic Health Records Improves Mortality Prediction in Patients Hospitalized with COVID-19. medRxiv.
  • Liu, B., Yan, B., Zhou, Y., Yang, Y., & Zhang, Y. (2020). Experiments of federated learning for covid-19 chest x-ray images. arXiv preprint arXiv:2007.05592.
  • Sheller, M. J., Reina, G. A., Edwards, B., Martin, J., & Bakas, S. (2018, September). Multi-institutional deep learning modeling without sharing patient data: A feasibility study on brain tumor segmentation. In International MICCAI Brainlesion Workshop (pp. 92-104). Springer, Cham.