Shukla‐Dave, Amita

Abstract Innovation in medical imaging artificial intelligence (AI)/machine learning (ML) demands extensive data collection, algorithmic advancements, and rigorous performance assessments encompassing aspects such as generalizability, uncertainty, bias, fairness, trustworthiness, and interpretability. Achieving widespread integration of AI/ML algorithms into diverse clinical tasks will demand a steadfast commitment to overcoming issues in model design, development, and performance assessment. The complexities of AI/ML clinical translation present substantial challenges, requiring engagement with relevant stakeholders, assessment of cost-effectiveness for user and patient benefit, timely dissemination of information relevant to robust functioning throughout the AI/ML lifecycle, consideration of regulatory compliance, and feedback loops for real-world performance evidence. This commentary addresses several hurdles for the development and adoption of AI/ML technologies in medical imaging. Comprehensive attention to these underlying and often subtle factors is critical not only for tackling the challenges but also for exploring novel opportunities for the advancement of AI in radiology.

Abstract The adoption of artificial intelligence (AI) tools in medicine poses challenges to existing clinical workflows. This commentary discusses the necessity of context-specific quality assurance (QA), emphasizing the need for robust QA measures with quality control (QC) procedures that encompass (1) acceptance testing (AT) before clinical use, (2) continuous QC monitoring, and (3) adequate user training. The discussion also covers essential components of AT and QA, illustrated with real-world examples. We also highlight what we see as the shared responsibility of manufacturers or vendors, regulators, healthcare systems, medical physicists, and clinicians to enact appropriate testing and oversight to ensure a safe and equitable transformation of medicine through AI.

AbstractPurpose has often been proposed as a quantitative imaging biomarker for diagnosis, prognosis, and treatment response assessment for various tumors. None of the many software tools for quantification are standardized. The ISMRM Open Science Initiative for Perfusion Imaging–Dynamic Contrast‐Enhanced (OSIPI‐DCE) challenge was designed to benchmark methods to better help the efforts to standardize measurement.MethodsA framework was created to evaluate values produced by DCE‐MRI analysis pipelines to enable benchmarking. The perfusion MRI community was invited to apply their pipelines for quantification in glioblastoma from clinical and synthetic patients. Submissions were required to include the entrants' values, the applied software, and a standard operating procedure. These were evaluated using the proposed score defined with accuracy, repeatability, and reproducibility components.ResultsAcross the 10 received submissions, the score ranged from 28% to 78% with a 59% median. The accuracy, repeatability, and reproducibility scores ranged from 0.54 to 0.92, 0.64 to 0.86, and 0.65 to 1.00, respectively (0–1 = lowest–highest). Manual arterial input function selection markedly affected the reproducibility and showed greater variability in analysis than automated methods. Furthermore, provision of a detailed standard operating procedure was critical for higher reproducibility.ConclusionsThis study reports results from the OSIPI‐DCE challenge and highlights the high inter‐software variability within estimation, providing a framework for ongoing benchmarking against the scores presented. Through this challenge, the participating teams were ranked based on the performance of their software tools in the particular setting of this challenge. In a real‐world clinical setting, many of these tools may perform differently with different benchmarking methodology.

Abstract Background Patients with human papillomavirus–related oropharyngeal cancers have excellent outcomes but experience clinically significant toxicities when treated with standard chemoradiotherapy (70 Gy). We hypothesized that functional imaging could identify patients who could be safely deescalated to 30 Gy of radiotherapy. Methods In 19 patients, pre- and intratreatment dynamic fluorine-18-labeled fluoromisonidazole positron emission tomography (PET) was used to assess tumor hypoxia. Patients without hypoxia at baseline or intratreatment received 30 Gy; patients with persistent hypoxia received 70 Gy. Neck dissection was performed at 4 months in deescalated patients to assess pathologic response. Magnetic resonance imaging (weekly), circulating plasma cell-free DNA, RNA-sequencing, and whole-genome sequencing (WGS) were performed to identify potential molecular determinants of response. Samples from an independent prospective study were obtained to reproduce molecular findings. All statistical tests were 2-sided. Results Fifteen of 19 patients had no hypoxia on baseline PET or resolution on intratreatment PET and were deescalated to 30 Gy. Of these 15 patients, 11 had a pathologic complete response. Two-year locoregional control and overall survival were 94.4% (95% confidence interval = 84.4% to 100%) and 94.7% (95% confidence interval = 85.2% to 100%), respectively. No acute grade 3 radiation–related toxicities were observed. Microenvironmental features on serial imaging correlated better with pathologic response than tumor burden metrics or circulating plasma cell-free DNA. A WGS-based DNA repair defect was associated with response (P = .02) and was reproduced in an independent cohort (P = .03). Conclusions Deescalation of radiotherapy to 30 Gy on the basis of intratreatment hypoxia imaging was feasible, safe, and associated with minimal toxicity. A DNA repair defect identified by WGS was predictive of response. Intratherapy personalization of chemoradiotherapy may facilitate marked deescalation of radiotherapy.

Level of Evidence : 5 Technical Efficacy : Stage 5 J. Magn. Reson. Imaging 2019.