Generative Pre-trained Transformer 4 makes cardiovascular magnetic resonance reports easy to understand

BACKGROUND

Patients are increasingly using Generative Pre-trained Transformer 4 (GPT-4) to better understand their own radiology findings.

PURPOSE

To evaluate the performance of GPT-4 in transforming cardiovascular magnetic resonance (CMR) reports into text that is comprehensible to medical laypersons.

METHODS

ChatGPT with GPT-4 architecture was used to generate three different explained versions of 20 various CMR reports (n = 60) using the same prompt: "Explain the radiology report in a language understandable to a medical layperson". Two cardiovascular radiologists evaluated understandability, factual correctness, completeness of relevant findings, and lack of potential harm, while 13 medical laypersons evaluated the understandability of the original and the GPT-4 reports on a Likert scale (1 "strongly disagree", 5 "strongly agree"). Readability was measured using the Automated Readability Index (ARI). Linear mixed-effects models (values given as median [interquartile range]) and intraclass correlation coefficient (ICC) were used for statistical analysis.

RESULTS

GPT-4 reports were generated on average in 52 s ± 13. GPT-4 reports achieved a lower ARI score (10 [9-12] vs 5 [4-6]; p < 0.001) and were subjectively easier to understand for laypersons than original reports (1 [1] vs 4 [4,5]; p < 0.001). Eighteen out of 20 (90%) standard CMR reports and 2/60 (3%) GPT-generated reports had an ARI score corresponding to the 8th grade level or higher. Radiologists' ratings of the GPT-4 reports reached high levels for correctness (5 [4, 5]), completeness (5 [5]), and lack of potential harm (5 [5]); with "strong agreement" for factual correctness in 94% (113/120) and completeness of relevant findings in 81% (97/120) of reports. Test-retest agreement for layperson understandability ratings between the three simplified reports generated from the same original report was substantial (ICC: 0.62; p < 0.001). Interrater agreement between radiologists was almost perfect for lack of potential harm (ICC: 0.93, p < 0.001) and moderate to substantial for completeness (ICC: 0.76, p < 0.001) and factual correctness (ICC: 0.55, p < 0.001).

CONCLUSION

GPT-4 can reliably transform complex CMR reports into more understandable, layperson-friendly language while largely maintaining factual correctness and completeness, and can thus help convey patient-relevant radiology information in an easy-to-understand manner.

Cardiac magnetic resonance based evaluation of aortic stiffness and epicardial fat volume in patients with hypertension, diabetes mellitus, and myocardial infarction

Background Aortic stiffness and epicardial fat relate to cardiovascular risk. Their relationship with each other and their role with hypertension, diabetes mellitus (DM), and myocardial infarction (MI) can be evaluated by cardiac magnetic resonance (CMR). Purpose To explore an association between aortic stiffness and epicardial as well as paracardial fat volume (EFV and ParaFV, respectively) in hypertensive patients and to relate the results to the presence of DM and MI. Material and Methods A total of 156 hypertensive and 20 non-hypertensive participants were examined at 1.5 Tesla. A 2D-velocity-encoded sequence was acquired to assess aortic pulse wave velocity (PWV in m/s) as a measure of aortic stiffness. A 3D-Dixon sequence was used to determine EFV and ParaFV. Results PWV correlated with EFV (R = 0.474; P < 0.001), but not with ParaFV. Fat volumes (in mL/m²) and PWV were lower in non-hypertensive controls compared to hypertensive patients. EFV and PWV were significantly higher in diabetic hypertensive patients without MI (n = 19; PWV: 10.4 ± 2.9; EFV: 92.5 ± 19.3) compared to hypertension-only patients (n = 84 [no DM or MI]; EFV: 64.8 ± 25.1, PWV: 9.0 ± 2.6; P < 0.05). Logistic regression analysis showed a significant association between the presence of a MI and a higher EFV ( P < 0.05), but not with PWV ( P = 0.060) or ParaFV ( P = 0.375). Conclusion A relationship between aortic stiffness and EFV was found in hypertensive patients. Both were increased in the presence of DM; however, only EFV was increased in the presence of MI. This may relate to the PWV lowering effect of the antihypertensive medication used by hypertensive patients and underscores the benefit of EFV assessment in this regard.

Feature-tracking myocardial strain analysis in acute myocarditis: diagnostic value and association with myocardial oedema

OBJECTIVES

To investigate the diagnostic value of cardiac magnetic resonance (CMR) feature-tracking (FT) myocardial strain analysis in patients with suspected acute myocarditis and its association with myocardial oedema.

METHODS

Forty-eight patients with suspected acute myocarditis and 35 control subjects underwent CMR. FT CMR analysis of systolic longitudinal (LS), circumferential (CS) and radial strain (RS) was performed. Additionally, the protocol allowed for the assessment of T1 and T2 relaxation times.

RESULTS

When compared with healthy controls, myocarditis patients demonstrated reduced LS, CS and RS values (LS: -19.5 ± 4.4% vs. -23.6 ± 3.1%, CS: -23.0 ± 5.8% vs. -27.4 ± 3.4%, RS: 28.9 ± 8.5% vs. 32.4 ± 7.4%; P < 0.05, respectively). LS (T1: r = 0.462, P < 0.001; T2: r = 0.436, P < 0.001) and CS (T1: r = 0.429, P < 0.001; T2: r = 0.467, P < 0.001) showed the strongest correlations with T1 and T2 relaxations times. Area under the curve of LS (0.79) was higher compared with those of CS (0.75; P = 0.478) and RS (0.62; P = 0.008).

CONCLUSIONS

FT CMR myocardial strain analysis might serve as a new tool for assessment of myocardial dysfunction in the diagnostic work-up of patients suspected of having acute myocarditis. Especially, LS and CS show a sufficient diagnostic performance and were most closely correlated with CMR parameters of myocardial oedema.

KEY POINTS

• Myocardial strain measures are considerably reduced in patients with suspected myocarditis. • Myocardial strain measures can sufficiently discriminate between diseased and healthy patients. • Myocardial strain measures show basic associations with the extent of myocardial oedema/inflammation.

Comprehensive Cardiac Magnetic Resonance for Short-Term Follow-Up in Acute Myocarditis

Background

Cardiac magnetic resonance (CMR) can detect inflammatory myocardial alterations in patients suspected of having acute myocarditis. There is limited information regarding the degree of normalization of CMR parameters during the course of the disease and the time window during which quantitative CMR should be most reasonably implemented for diagnostic work‐up.

Methods and Results

Twenty‐four patients with suspected acute myocarditis and 45 control subjects underwent CMR. Initial CMR was performed 2.6±1.9 days after admission. Myocarditis patients underwent CMR follow‐up after 2.4±0.6, 5.5±1.3, and 16.2±9.9 weeks. The CMR protocol included assessment of standard Lake Louise criteria, T1 relaxation times, extracellular volume fraction, and T2 relaxation times. Group differences between myocarditis patients and control subjects were highest in the acute stage of the disease (P<0.001 for all parameters). There was a significant and consistent decrease in all inflammatory CMR parameters over the course of the disease (P<0.01 for all parameters). Myocardial T1 and T2 relaxation times—indicative of myocardial edema—were the only single parameters showing significant differences between myocarditis patients and control subjects on 5.5±1.3‐week follow‐up (T1: 986.5±44.4 ms versus 965.1±28.1 ms, P=0.022; T2: 55.5±3.2 ms versus 52.6±2.6 ms; P=0.001).

Conclusions

In patients with acute myocarditis, CMR markers of myocardial inflammation demonstrated a rapid and continuous decrease over several follow‐up examinations. CMR diagnosis of myocarditis should therefore be attempted at an early stage of the disease. Myocardial T1 and T2 relaxation times were the only parameters of active inflammation/edema that could discriminate between myocarditis patients and control subjects even at a convalescent stage of the disease.

Incremental value of quantitative CMR including parametric mapping for the diagnosis of acute myocarditis

AIM

Cardiac magnetic resonance (CMR) can visualize inflammatory tissue changes in acute myocarditis. Several quantitative image-derived parameters have been described to enhance the diagnostic value of CMR, but no direct comparison of these techniques is available.

METHODS AND RESULTS

A total of 34 patients with suspected acute myocarditis and 50 control subjects underwent CMR. CMR protocol included quantitative assessment of T1 relaxation times using modified Look-Locker inversion recovery (MOLLI) and shortened MOLLI (ShMOLLI) acquisition schemes, extracellular volume fraction (ECV), T2 relaxation times, and longitudinal strain. Established Lake-Louise criteria (LLC) consisting of T2-weighted signal intensity ratio (T2-ratio), early gadolinium enhancement ratio (EGEr), and late gadolinium enhancement (LGE) were assessed. Receiver operating characteristics analysis was performed to compare diagnostic performance. Areas under the curve of native T1 (MOLLI: 0.95; ShMOLLI: 0.92) and T2 relaxation times (0.92) were higher compared with those of the other CMR parameters (T2-ratio: 0.71, EGEr: 0.71, LGE: 0.87, LLC: 0.90, ECV MOLLI: 0.77, ECV ShMOLLI: 0.80, longitudinal strain: 0.83). Combined with LGE, each native mapping technique outperformed the diagnostic performance of LLC (P < 0.01, respectively). A combination of native parameters (T1, T2, and longitudinal strain) significantly increased the diagnostic performance of CMR compared with LLC without need of contrast media application (0.99 vs. 0.90; P = 0.008).

CONCLUSION

In patients suspected of having acute myocarditis, diagnostic performance of CMR can be improved by implementation of quantitative CMR parameters. Especially, native mapping techniques have the potential to replace current LLC. CLINICALTRIALS.

GOV NUMBER

NCT02299856.