Disparities in Care Among Patients With Cardiac Implantable Electronic Devices Undergoing MRI

Machine Learning and Bias in Medical Imaging: Opportunities and Challenges

Bias in health care has been well documented and results in disparate and worsened outcomes for at-risk groups. Medical imaging plays a critical role in facilitating patient diagnoses but involves multiple sources of bias including factors related to access to imaging modalities, acquisition of images, and assessment (ie, interpretation) of imaging data. Machine learning (ML) applied to diagnostic imaging has demonstrated the potential to improve the quality of imaging-based diagnosis and the precision of measuring imaging-based traits. Algorithms can leverage subtle information not visible to the human eye to detect underdiagnosed conditions or derive new disease phenotypes by linking imaging features with clinical outcomes, all while mitigating cognitive bias in interpretation. Importantly, however, the application of ML to diagnostic imaging has the potential to either reduce or propagate bias. Understanding the potential gain as well as the potential risks requires an understanding of how and what ML models learn. Common risks of propagating bias can arise from unbalanced training, suboptimal architecture design or selection, and uneven application of models. Notwithstanding these risks, ML may yet be applied to improve gain from imaging across all 3A's (access, acquisition, and assessment) for all patients. In this review, we present a framework for understanding the balance of opportunities and challenges for minimizing bias in medical imaging, how ML may improve current approaches to imaging, and what specific design considerations should be made as part of efforts to maximize the quality of health care for all.

Use of Screening CT Colonography by Age and Race: A Study of Potential Access Barriers Related to Medicare Noncoverage Based on Data From the ACR's National CT Colonography Registry

OBJECTIVE

The primary objectives of this investigation were to evaluate the use of screening CT colonography (CTC) examinations by age comparing individuals of Medicare-eligible age to younger cohorts and to determine if the association between use of CTC and Medicare-eligible age varies by race. Although the Affordable Care Act requires commercial insurance coverage of screening CTC, Medicare does not cover screening CTC.

MATERIALS AND METHODS

Using the ACR's CTC registry, the distribution of procedures by age was evaluated using a negative binomial model with patient age (to capture overall trend), indicator of Medicare-eligible age (to capture immediate changes in trend at age 65), and their interaction (to capture gradual changes after age 65) as independent variables. The association between the number of screening CTCs and age was compared by racial identity.

RESULTS

The CTC registry contained data on 12,648 screening examinations. Between ages 52 and 64, the number of screening examinations increased; each additional age year was associated with a 5.3% (P < .001) increase in the number of screenings. However, after age 65, the number of screening examinations decreased by -6.9% per additional year of age above 65 compared with the trend between ages 52 and 64 (P < .001). The modal age group for CTC use was 65 to 69 years in white and 55 to 59 in black individuals.

CONCLUSION

After age 65, the number of screening CTC examinations decreased, likely due, at least in part, to lack of Medicare coverage. Medicare noncoverage may have a disproportionate impact on black patients and other racial minorities.

Cardiac MRI for Patients With Cardiac Implantable Electronic Devices

OBJECTIVE. Patients with cardiac implantable electronic devices (CIEDs) require cardiac MRI (CMRI) for a variety of reasons. The purpose of this study is to review and evaluate the value and safety of CMRI for patients with in situ CIEDs. CONCLUSION. Late gadolinium enhancement CMRI is the reference standard for assessing myocardial viability in patients with ventricular tachycardia before ablation of arrhythmogenic substrates. The use of late gadolinium enhancement CMRI for patients with CIEDs is safe as long as an imaging protocol is in place and precaution measures are taken.

Performing MRI on patients with MRI-conditional and non-conditional cardiac implantable electronic devices: an update for radiologists

Pacemakers and implantable cardioverter defibrillators are commonly encountered in clinical practice, and entails special consideration when magnetic resonance imaging (MRI) is required. It is estimated that 50-75% of patients with cardiac implantable electronic devices (CIED) will have an indication for MRI during their lifetime. Radiologists may want to recommend MRI or may be consulted about the need to perform MRI in a patient with a CIED, at which point they may need to approve or at least provide guidance as to whether MRI may be performed safely. Even in situations where final clearance will not be provided by the radiologist, he or she can provide valuable information by reviewing radiographs and determining (a) whether a device is MRI-conditional and MRI may ultimately be permitted, (b) is not MRI-conditional and MRI using the standard workflow will therefore not be approved, or (c) when additional information will clearly be required. CIED identification and verification of leads can be accomplished through review of the medical record and/or evaluation of a chest radiograph. In patients with MRI-conditional CIEDs (as well as with legacy CIEDs in those institutions that perform MRI of these patients), specific imaging protocols must be adhered to in order to prevent death or injury to the patient or damage to the device. In this update, we provide details regarding the above topics and provide an algorithm for integrating this information into a clinical workflow to efficiently triage patients with CIEDs who are being considered for MRI.