Prediction with
machine learning

Background

Over the past few years, there has been an increased interest in applying machine learning (ML) techniques to medical research. ML is a particular branch of artificial intelligence, which employs a variety of statistical, probabilistic and optimization techniques that allows computers to “learn” from past examples and to detect patterns from large complex data. With the growing availability of mixed data (for instance clinical and genomic), ML methods - which have great potential for modelling complex data - have been increasingly applied.

Aims

• To apply ML for cancer prognosis and prediction. Prediction is particularly interesting as it is part of a growing trend towards personalized medicine. One of the aims of the project is to evaluate if ML methods are suitable for predicting survival probabilities on a patient level.

• To identify the potential of ML methodology to censored survival data. In particular, whether ML methods can be used to improve the accuracy of predicting survival, recurrence and mortality compared to that of the traditional Cox proportional hazards model for time to event data.

• To investigate the performance of different machine learning techniques (e.g. neural networks, random forests) relative to each other, under different amounts of censoring, and relative to the Cox proportional hazards model.

• In a survival analysis context, whereas methods have been adapted to deal with right-censored data, there is a lack of research on what happens in the event of left, interval-censoring, truncation. This aspect will be investigated.

Relevance for cancer research

Within this project, research will be carried out in order to comprehensively establish the potential of ML for survival analysis of cancer data. Methodology developed will be applied to data of the Soft Tissue and Bone Sarcoma Group (STBSG). The methodology of this research could be applicable to any kind of cancerous diseases.

Project Outcomes

• Kantidakis G, Litière S, Putter H & Fiocco M (2023).  Statistical models versus machine learning for competing risks: development and validation of prognostic models. BMC Medical Research Methodology, 23:51. doi: 10.1186/s12874-023-01866-z 

• Kantidakis G (2022, November 23). Analysis of sarcoma and non-sarcoma clinical data with statistical methods and machine learning techniques. [Doctoral Thesis]

• Kantidakis G, Hazewinkel AD & Fiocco M (2022). Neural Networks for Survival Prediction in Medicine Using Prognostic Factors: A Review and Critical Appraisal. Computational and mathematical methods in medicine, 1176060. doi: 10.1155/2022/1176060

• Kantidakis G, Litière S, Neven A, Vinches M, Judson I, Blay J-Y, Wardelmann E, Stacchiotti S, D'Ambrosio L, Marréaud S, van der Graaf WTA, Kasper B, Kasper B, Fiocco M & Gelderblom H (2022). New benchmarks to design clinical trials with advanced or metastatic liposarcoma or synovial sarcoma patients: An EORTC – Soft Tissue and Bone Sarcoma Group (STBSG) meta-analysis based on a literature review for soft-tissue sarcomas. European Journal of Cancer,  174:261-276. doi: 10.1016/j.ejca.2022.07.010

• Kantidakis G, Litière S, Gelderblom H, Fiocco M, Judson I, van der Graaf WTA, Italiano A, Marréaud S, Sleijfer S, Mechtersheimer G, Messiou C & Kasper B (2022). Prognostic Significance of Bone Metastasis in Soft Tissue Sarcoma Patients Receiving Palliative Systemic Therapy: An Explorative, Retrospective Pooled Analysis of the EORTC-Soft Tissue and Bone Sarcoma Group (STBSG) Database. Sarcoma, 5815875. doi: 10.1155/2022/5815875

• Kantidakis G, Biganzoli E, Putter H & Fiocco M (2021). A Simulation Study to Compare the Predictive Performance of Survival Neural Networks with Cox Models for Clinical Trial Data. Computational and Mathematical Methods in Medicine, 2160332. doi: 10.1155/2021/2160322

• Kantidakis G, Litière S, Neven A, Vinches M, Judson I, Schöffski P, Wardelmann E, Stacchiotti S, D'Ambrosio L, Marréaud S, van der Graaf WTA, Kasper B, Fiocco M & Gelderblom H (2021). Efficacy thresholds for clinical trials with advanced or metastatic leiomyosarcoma patients: A European Organisation for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group meta-analysis based on a literature review for soft-tissue sarcomas. European Journal of Cancer, 154:253-268. doi: 10.1016/j.ejca.2021.06.025

• Kantidakis G, Putter H,  Lancia C,  de Boer J, Braat AE & Fiocco M (2020). Survival prediction models since liver transplantation - comparisons between Cox models and machine learning techniques. BMC Medical Research Methodology,  20, 277. doi: 10.1186/s12874-020-01153-1 

Team

• Prof. dr. M. Fiocco, Mathematical Institute Leiden University, Department of Biomedical Data Sciences Leiden University Medical Center & Princess Máxima Center for Pediatric Oncology 

• Prof. dr. H. Gelderblom , Department of Medical Oncology at the Leiden University Medical Center

• G. Kantidakis , PhD candidate at LUMC – MI – EORTC 

• Dr. S. Litiere,  Head of Statistics Department EORTC