Deep Learning Features Can Improve Radiomics-Based Prostate Cancer Aggressiveness Prediction
The study findings reveal that incorporating deep features derived from autoencoder models trained to reconstruct the full prostate gland (both zonal models show worse performance than radiomics only models), combined with radiomic features, often lead to a statistically significant increase in model performance for disease aggressiveness classification. Additionally, the results also demonstrate that the choice of feature selection is key to achieving good performance, with principal component analysis (PCA) and PCA + relief being the best approaches and that there is no clear difference between the three proposed latent representation extraction techniques.
Authors
- M. V. Lisitskaya (ORCID: https://orcid.org/0000-0002-8402-7643)
- Aycan Uysal (ORCID: https://orcid.org/0000-0001-6348-9403)
- Nickolas Papanikolaou (ORCID: https://orcid.org/0000-0003-3298-2072)
- Leonardo Vanneschi (ORCID: https://orcid.org/0000-0003-4732-3328)
- Celso Matos (ORCID: https://orcid.org/0000-0001-7383-6309)
- José Guilherme de Almeida (ORCID: https://orcid.org/0000-0002-1887-0157)
- Sara Silva (ORCID: https://orcid.org/0000-0001-8223-4799)
- Ana Rodrigues (ORCID: https://orcid.org/0009-0006-3252-7343)
- Nuno M. Rodrigues (ORCID: https://orcid.org/0000-0001-5312-8276)
Institutions
- Champalimaud Foundation (PT)
- University of Lisbon (PT)
- Universidade do Porto (PT)
- Lomonosov Moscow State University (RU)
- Gülhane Askerî Tıp Akademisi (TR)
- Sağlık Bilimleri Üniversitesi (TR)
- Universidade Nova de Lisboa (PT)
Publication Details
- Journal
- JCO Clinical Cancer Informatics
- Published
- 2024-09-01
- DOI
- https://doi.org/10.1200/cci.23.00180
- Citations
- 4
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 1.62
Funders
- European Commission
- Horizon 2020 Framework Programme
- Fundação para a Ciência e a Tecnologia