Deep learning diffusion fingerprinting to detect brain tumour response to chemotherapy

Abstract

1Artificial neural networks are being widely implemented for a range of different biomedical 2 imaging applications. Convolutional neural networks are by far the most popular type of deep 3 learning architecture, but often require very large datasets for robust training and evaluation. 4We introduce deep learning diffusion fingerprinting (DLDF), which we have used to classify 5 diffusion-weighted magnetic resonance imaging voxels in a mouse model of glioblastoma 6 (GL261 cell line), both prior to and in response to Temozolomide (TMZ) chemotherapy. We 7show that, even with limited training, DLDF can automatically segment brain tumours from 8 normal brain, can automatically distinguish between young and older (after 9 days of growth) 9 tumours and that DLDF can detect whether or not a tumour has been treated with 10 chemotherapy. Our results also suggest that DLDF can detect localised changes in the 11 underlying tumour microstructure, which are not evident using conventional measurements 12 of the apparent diffusion coefficient (ADC). Tissue category maps generated by DLDF 13 showed regions containing a mixture of normal brain and tumour cells, and in some cases 14 evidence of tumour invasion across the corpus callosum, which were broadly consistent with 15 histology. In conclusion, DLDF shows the potential for applying deep learning on a pixel-wise 16 level, which reduces the need for vast training datasets and could easily be applied to other 17 multi-dimensional imaging acquisitions. 18 19

Big Brain Data: On the Responsible Use of Brain Data from Clinical and Consumer-Directed Neurotechnological Devices

“…Most recent efforts in leveraging advanced machine learning in clinical neuroscience have been particularly fruitful in neuroimaging-based diagnosis (and/or prediction) in neurology and psychiatry. In neurology, such approaches have already been able to: detect morphological brain changes typical of Alzheimer's disease from neuroimaging [28,29], to predict brain tumor response to chemotherapy from brain images [30], or distinguish typical from atypical Parkinson's syndromes [31]. In psychiatric research, examples for leveraging machine learning are the prediction of outcomes in psychosis [32], the persistence and severity of depressive symptoms [33], and the prediction of suicidal behavior [7].…”

Current Applications and Future Promises of Machine Learning in Diffusion MRI

“…The authors in [54] proposed a Deep Learning Diffusion Fingerprinting (DLDF) method based on DNN and used to classify DW-MRI voxels. The authors showed that this model can learn even with limited training samples, and the method is capable to segment brain tumours, distinguish between young and older tumours and detect whether or not a tumour has been treated with chemotherapy.…”