The role of artificial intelligence in paediatric neuroradiology

Abstract

Imaging plays a fundamental role in the managing childhood neurologic, neurosurgical and neuro-oncological disease. Employing multi-parametric MRI techniques, such as spectroscopy and diffusion- and perfusion-weighted imaging, to the radiophenotyping of neuroradiologic conditions is becoming increasingly prevalent, particularly with radiogenomic analyses correlating imaging characteristics with molecular biomarkers of disease. However, integration into routine clinical practice remains elusive. With modern multi-parametric MRI now providing additional data beyond anatomy, informing on histology, biology and physiology, such metric-rich information can present as information overload to the treating radiologist and, as such, information relevant to an individual case can become lost. Artificial intelligence techniques are capable of modelling the vast radiologic, biological and clinical datasets that accompany childhood neurologic disease, such that this information can become incorporated in upfront prognostic modelling systems, with artificial intelligence techniques providing a plausible approach to this solution. This review examines machine learning approaches than can be used to underpin such artificial intelligence applications, with exemplars for each machine learning approach from the world literature. Then, within the specific use case of paediatric neuro-oncology, we examine the potential future contribution for such artificial intelligence machine learning techniques to offer solutions for patient care in the form of decision support systems, potentially enabling personalised medicine within this domain of paediatric radiologic practice.

Analysis of the role of artificial intelligence in pediatric radiography

“…However, due to the physical vulnerability of children, various treatments may be harmful to them, which makes brain tumors always be a serious challenge for physicians [8]. Due to the complexity of brain tumors in children, artificial intelligence can be useful in two directions, firstly in the diagnostic direction, where the use of artificial neural networks has made it possible to use only T-2 weighted imaging to classify medulloblastoma, diffuse midline glioma, and hairy cell astrocytoma [8]. Some success has also been achieved using decision trees to classify high-and low-grade gliomas [8].…”

“…Due to the complexity of brain tumors in children, artificial intelligence can be useful in two directions, firstly in the diagnostic direction, where the use of artificial neural networks has made it possible to use only T-2 weighted imaging to classify medulloblastoma, diffuse midline glioma, and hairy cell astrocytoma [8]. Some success has also been achieved using decision trees to classify high-and low-grade gliomas [8].…”

“…The analysis of MRI textures of brain tumors is performed using an algorithm such as support vector machines, which transforms imaging data into image textures that are not visible to the human eye, and thus provides a qualitative analysis of brain tumors [8]. Texture analysis is based on whole tumor imaging data to analyze the heterogeneity within the tumor, which is an accuracy that may not be achieved by any normal biopsy [8]. Textural analysis is now 95% accurate for hairy cell astrocytoma's, medulloblastomas and ventricular meningiomas, and has been successful in diagnosing brain tumors [8].…”

“…Texture analysis is based on whole tumor imaging data to analyze the heterogeneity within the tumor, which is an accuracy that may not be achieved by any normal biopsy [8]. Textural analysis is now 95% accurate for hairy cell astrocytoma's, medulloblastomas and ventricular meningiomas, and has been successful in diagnosing brain tumors [8].…”

Machine Learning in the Classification of Pediatric Posterior Fossa Tumors: A Systematic Review

“…Prior work has shown that artificial intelligence is becoming an increasingly viable tool with the potential to improve diagnostic speed and accuracy [ 8 , 9 ]. Machine learning has already been heavily implemented in the diagnosis of brain tumors in both children and adults, with previous studies reporting algorithms that can differentiate gliomas, meningiomas, and pituitary tumors based on extracted imaging features with accuracies as high as 99% [ 10 , 11 , 12 ]. Additional work has shown the possibility of using these methods to not only differentiate between tumor types, but also to subclassify tumors by grade, stage, and even molecular features [ 12 , 13 , 14 ].…”