Format

Send to

Choose Destination
Sci Rep. 2019 Jul 11;9(1):10063. doi: 10.1038/s41598-019-46296-4.

Integration of machine learning and mechanistic models accurately predicts variation in cell density of glioblastoma using multiparametric MRI.

Author information

1
School of Computing, Informatics, and Decision Systems Engineering, Arizona State University, 699 S Mill Ave, Tempe, AZ, 85281, USA.
2
Precision NeuroTherapeutics (PNT) Lab, Mayo Clinic Arizona, 5777 E Mayo Blvd, Phoenix, Arizona, 85054, USA. hawkins-daarud.andrea@mayo.edu.
3
Department of Radiology, Mayo Clinic Arizona, 5777 E Mayo Blvd, Phoenix, Arizona, 85054, USA.
4
Precision NeuroTherapeutics (PNT) Lab, Mayo Clinic Arizona, 5777 E Mayo Blvd, Phoenix, Arizona, 85054, USA.
5
Department of Pathology, Barrow Neurological Institute, Phoenix, Arizona, USA.
6
Department of Neurosurgery, Barrow Neurological Institute, Phoenix, Arizona, USA.
7
Department of Biostatistics and Bioinformatics, Moffitt Cancer Center, Tampa, Florida, 33612, USA.
8
Department of Neurosurgery, Mayo Clinic Arizona, 5777 E Mayo Blvd, Phoenix, Arizona, 85054, USA.

Abstract

Glioblastoma (GBM) is a heterogeneous and lethal brain cancer. These tumors are followed using magnetic resonance imaging (MRI), which is unable to precisely identify tumor cell invasion, impairing effective surgery and radiation planning. We present a novel hybrid model, based on multiparametric intensities, which combines machine learning (ML) with a mechanistic model of tumor growth to provide spatially resolved tumor cell density predictions. The ML component is an imaging data-driven graph-based semi-supervised learning model and we use the Proliferation-Invasion (PI) mechanistic tumor growth model. We thus refer to the hybrid model as the ML-PI model. The hybrid model was trained using 82 image-localized biopsies from 18 primary GBM patients with pre-operative MRI using a leave-one-patient-out cross validation framework. A Relief algorithm was developed to quantify relative contributions from the data sources. The ML-PI model statistically significantly outperformed (p < 0.001) both individual models, ML and PI, achieving a mean absolute predicted error (MAPE) of 0.106 ± 0.125 versus 0.199 ± 0.186 (ML) and 0.227 ± 0.215 (PI), respectively. Associated Pearson correlation coefficients for ML-PI, ML, and PI were 0.838, 0.518, and 0.437, respectively. The Relief algorithm showed the PI model had the greatest contribution to the result, emphasizing the importance of the hybrid model in achieving the high accuracy.

Supplemental Content

Full text links

Icon for Nature Publishing Group Icon for PubMed Central
Loading ...
Support Center