Artificial intelligence and radiomics: fundamentals, applications, and challenges in immunotherapy

Abstract

Immunotherapy offers the potential for durable clinical benefit but calls into question the association between tumor size and outcome that currently forms the basis for imaging-guided treatment. Artificial intelligence (AI) and radiomics allow for discovery of novel patterns in medical images that can increase radiology’s role in management of patients with cancer, although methodological issues in the literature limit its clinical application. Using keywords related to immunotherapy and radiomics, we performed a literature review of MEDLINE, CENTRAL, and Embase from database inception through February 2022. We removed all duplicates, non-English language reports, abstracts, reviews, editorials, perspectives, case reports, book chapters, and non-relevant studies. From the remaining articles, the following information was extracted: publication information, sample size, primary tumor site, imaging modality, primary and secondary study objectives, data collection strategy (retrospective vs prospective, single center vs multicenter), radiomic signature validation strategy, signature performance, and metrics for calculation of a Radiomics Quality Score (RQS). We identified 351 studies, of which 87 were unique reports relevant to our research question. The median (IQR) of cohort sizes was 101 (57–180). Primary stated goals for radiomics model development were prognostication (n=29, 33.3%), treatment response prediction (n=24, 27.6%), and characterization of tumor phenotype (n=14, 16.1%) or immune environment (n=13, 14.9%). Most studies were retrospective (n=75, 86.2%) and recruited patients from a single center (n=57, 65.5%). For studies with available information on model testing, most (n=54, 65.9%) used a validation set or better. Performance metrics were generally highest for radiomics signatures predicting treatment response or tumor phenotype, as opposed to immune environment and overall prognosis. Out of a possible maximum of 36 points, the median (IQR) of RQS was 12 (10–16). While a rapidly increasing number of promising results offer proof of concept that AI and radiomics could drive precision medicine approaches for a wide range of indications, standardizing the data collection as well as optimizing the methodological quality and rigor are necessary before these results can be translated into clinical practice.

Integration of longitudinal deep-radiomics and clinical data improves the prediction of durable benefits to anti-PD-1/PD-L1 immunotherapy in advanced NSCLC patients

“…Disease progression was defined based on the patient’s general clinical status and iRECIST criteria derived from the imaging evaluation. Patients with durable clinical benefit that had a PFS longer than 6 (PFS6) or 9 (PFS9) months were denominated as responders, while the others as nonresponders [ 22 ]. Patients with censored data 6 or 9 months after treatment were excluded from the analysis.…”

Advances in artificial intelligence to predict cancer immunotherapy efficacy

“…In general, the majority of studies’ Radiomics Quality Scores (RQS) ranged from 11 to 20 out of a possible maximum score of 36 points ( 36 ). It indicates that AI-based imaging-omics analysis can delve into the spatiotemporal heterogeneity of tumors and plays an important role in predicting immunotherapy response, biomarker expression, and patient prognosis, particularly in the absence of histopathological specimens.…”

Artificial Intelligence in CT and MR Imaging for Oncological Applications

“…AI methods have shown the potential to stratify patients based on risk factors as well as provide automated measurements of tumor volume via tumor segmentation [10,15,17]. Many studies have been published on machine learning tools for computer-aided or AIassisted clinical tasks [8,9,11,18]. However, most of these tools are not yet ready for clinical deployment.…”

“…The primary driver behind the emergence of artificial intelligence (AI) in medical imaging has been the desire for greater efficacy and efficiency in clinical care [8,9]. The topics of data sampling and deep learning (DL) strategies, including levels of learning supervision (transfer learning, multi-task learning, domain adaptation, and federated and continuous learning systems), are well covered in previously published reviews [10,11]. The importance of proper data collection and standardization methods, the appropriate choice of the reference standard in relation to the task at hand, the identification of suitable training approaches, the correct selection of performance metrics, the requirements of an efficient user interface, clinical workflows, and timely quality assurance of AI tools cannot be emphasized enough [12,13].…”