Applications and limitations of machine learning in radiation oncology

Abstract

Machine learning approaches to problem-solving are growing rapidly within healthcare, and radiation oncology is no exception. With the burgeoning interest in machine learning comes the significant risk of misaligned expectations as to what it can and cannot accomplish. This paper evaluates the role of machine learning and the problems it solves within the context of current clinical challenges in radiation oncology. The role of learning algorithms within the workflow for external beam radiation therapy are surveyed, considering simulation imaging, multimodal fusion, image segmentation, treatment planning, quality assurance, and treatment delivery and adaptation. For each aspect, the clinical challenges faced, the learning algorithms proposed, and the successes and limitations of various approaches are analyzed. It is observed that machine learning has largely thrived on reproducibly mimicking conventional human-driven solutions with more efficiency and consistency. On the other hand, since algorithms are generally trained using expert opinion as ground truth, machine learning is of limited utility where problems or ground truths are not well-defined, or if suitable measures of correctness are not available. As a result, machines may excel at replicating, automating and standardizing human behaviour on manual chores, meanwhile the conceptual clinical challenges relating to definition, evaluation, and judgement remain in the realm of human intelligence and insight.

Artificial Intelligence and the Medical Radiation Profession: How Our Advocacy Must Inform Future Practice

“…Several applications for the use of AI are emerging along the radiation therapy workflow and across the multidisciplinary radiation oncology practice. Given the ability of AI to perform decision-based, repetitive tasks, research has begun in several areas along the patient journey [3]. Contemporary methods of automated segmentation of tumours and organs at risk still require human oversight and editing.…”

“…This is one such area where machine learning may be used to improve the current workflow. The use of convolutional neural networks to improve auto-segmentation has shown impressive results when outlining the organs at risk, yet found the volume segmentation of contouring different shaped tumours more difficult [3]. Inverse treatment planning in radiation therapy uses a computer to develop a treatment plan based on an objective with constraints.…”

“…The enormous quantity of imaging and associated data produced by the medical radiation sciences has sparked the interest of some of the largest technology companies in the world. AI firms are seeking to access patient data for the purpose of training deep learning algorithms to automate tasks such as image classification and segmentation [2][3][4]. Even the human trait of empathy is explored as a potential, reproducible quality to enhance the patients' experience via the use of ''chatbots'' [5].…”

Overview of artificial intelligence-based applications in radiotherapy: Recommendations for implementation and quality assurance

“…The radiotherapy (RT) workflow is a complex process consisting of several time-consuming steps that have an impact on treatment quality and hence patient outcome. Artificial intelligence (AI) has been proposed as a tool to increase quality, standardization and acceleration of these steps leading to a more safe and accurate radiation administration by automation and optimization of workflows [1][2][3]. Especially with the introduction of adaptive radiotherapy (ART), a streamlined workflow is mandatory in clinical routine.…”

Auto-segmentations by convolutional neural network in cervical and anorectal cancer with clinical structure sets as the ground truth

“…Although guidelines exist, manual segmentations are subjective and time consuming, and an automated approach would make it possible to increase reproducibility, improve clinical workflow, and improve cancer care locally [1,2] and globally [3]. According to reviews on radiation oncology and automated segmentation based on artificial intelligence (AI), there are indeed clinical benefits, but there are also challenges including inaccurate or incomplete auto segmentation due to software performance or unrecognized anatomical variations [1,2,4,5].…”