3D fusion between fluoroscopy angiograms and SPECT myocardial perfusion images to guide percutaneous coronary intervention

How AI and Robotics Will Advance Interventional Radiology: Narrative Review and Future Perspectives

The rapid advancement of artificial intelligence (AI) and robotics has led to significant progress in various medical fields including interventional radiology (IR). This review focuses on the research progress and applications of AI and robotics in IR, including deep learning (DL), machine learning (ML), and convolutional neural networks (CNNs) across specialties such as oncology, neurology, and cardiology, aiming to explore potential directions in future interventional treatments. To ensure the breadth and depth of this review, we implemented a systematic literature search strategy, selecting research published within the last five years. We conducted searches in databases such as PubMed and Google Scholar to find relevant literature. Special emphasis was placed on selecting large-scale studies to ensure the comprehensiveness and reliability of the results. This review summarizes the latest research directions and developments, ultimately analyzing their corresponding potential and limitations. It furnishes essential information and insights for researchers, clinicians, and policymakers, potentially propelling advancements and innovations within the domains of AI and IR. Finally, our findings indicate that although AI and robotics technologies are not yet widely applied in clinical settings, they are evolving across multiple aspects and are expected to significantly improve the processes and efficacy of interventional treatments.

Automatic reorientation by deep learning to generate short-axis SPECT myocardial perfusion images

BACKGROUND

Single photon emission computed tomography (SPECT) myocardial perfusion images (MPI) can be displayed both in traditional short-axis (SA) cardiac planes and polar maps for interpretation and quantification. It is essential to reorient the reconstructed transaxial SPECT MPI into standard SA slices. This study is aimed to develop a deep-learning-based approach for automatic reorientation of MPI.

METHODS

A total of 254 patients were enrolled, including 226 stress SPECT MPIs and 247 rest SPECT MPIs. Fivefold cross-validation with 180 stress and 201 rest MPIs was used for training and internal validation; the remaining images were used for testing. The rigid transformation parameters (translation and rotation) from manual reorientation were annotated by an experienced nuclear cardiologist and used as the reference standard. A convolutional neural network (CNN) was designed to predict the transformation parameters. Then, the derived transform was applied to the grid generator and sampler in spatial transformer network (STN) to generate the reoriented image. A loss function containing mean absolute errors for translation and mean square errors for rotation was employed. A three-stage optimization strategy was adopted for model optimization: (1) optimize the translation parameters while fixing the rotation parameters; (2) optimize rotation parameters while fixing the translation parameters; (3) optimize both translation and rotation parameters together.

RESULTS

In the test set, the Spearman determination coefficients of the translation distances and rotation angles between the model prediction and the reference standard were 0.993 in X axis, 0.992 in Y axis, 0.994 in Z axis, 0.987 along X axis, 0.990 along Y axis and 0.996 along Z axis, respectively. For the 46 stress MPIs in the test set, the Spearman determination coefficients were 0.858 in percentage of profusion defect (PPD) and 0.858 in summed stress score (SSS); for the 46 rest MPIs in the test set, the Spearman determination coefficients were 0.9 in PPD and 0.9 in summed rest score (SRS).

CONCLUSIONS

Our deep learning-based LV reorientation method is able to accurately generate the SA images. Technical validations and subsequent evaluations of measured clinical parameters show that it has great promise for clinical use.

Review of cardiovascular imaging in the Journal of Nuclear Cardiology 2022: single photon emission computed tomography

In this review, we will summarize a selection of articles on single-photon emission computed tomography published in the Journal of Nuclear Cardiology in 2022. The aim of this review is to concisely recap major advancements in the field to provide the reader a glimpse of the research published in the journal over the last year. This review will place emphasis on myocardial perfusion imaging using single-photon emission computed tomography summarizing advances in the field including in prognosis, non-perfusion variables, attenuation compensation, machine learning and camera design. It will also review nuclear imaging advances in amyloidosis, left ventricular mechanical dyssynchrony, cardiac innervation, and lung perfusion. We encourage interested readers to go back to the original articles, and editorials, for a comprehensive read as necessary but hope that this yearly review will be helpful in reminding readers of articles they have seen and attracting their attentions to ones they have missed.

Texture analysis of SPECT myocardial perfusion provides prognostic value for dilated cardiomyopathy

BACKGROUND

Texture analysis (TA) has demonstrated clinical values in extracting information, quantifying inhomogeneity, evaluating treatment outcomes, and predicting long-term prognosis for cardiac diseases. The aim of this study was to explore whether TA of SPECT myocardial perfusion could contribute to improving the prognosis of dilated cardiomyopathy (DCM) patients.

METHODS

Eighty-eight patients were recruited in our study between 2009 and 2020 who were diagnosed with DCM and underwent single-photon emission tomography myocardial perfusion imaging (SPECT MPI). Forty TA features were obtained from quantitative analysis of SPECT imaging in subjects with myocardial perfusion at rest. All patients were divided into two groups: the all-cause death group and the survival group. The prognostic value of texture parameters was assessed by Cox regression and Kaplan-Meier analysis.

RESULTS

Twenty-five all-cause deaths (28.4%) were observed during the follow-up (39.2±28.7 months). Compared with the survival group, NT-proBNP and total perfusion deficit (TPD) were higher and left ventricular ejection fraction (LVEF) was lower in the all-cause death group. In addition, 26 out of 40 texture parameters were significantly different between the two groups. Univariate Cox regression analysis revealed that NT-proBNP, LVEF, and 25 texture parameters were significantly associated with all-cause death. The multivariate Cox regression analysis showed that low gray-level emphasis (LGLE) (P = 0.010, HR = 4.698, 95% CI 1.457-15.145) and long-run low gray-level emphasis (LRLGE) (P =0.002, HR = 6.085, 95% CI 1.906-19.422) were independent predictors of the survival outcome. When added to clinical parameters, LVEF, TPD, and TA parameters, including LGLE and LRLGE, were incrementally associated with all-cause death (global chi-square statistic of 26.246 vs. 33.521; P = 0.028, global chi-square statistic of 26.246 vs. 34.711; P = 0.004).

CONCLUSION

TA based on gated SPECT MPI could discover independent prognostic predictors of all-cause death in medically treated patients with DCM. Moreover, TA parameters, including LGLE and LRLGE, independent of the total perfusion deficit of the cardiac myocardium, appeared to provide incremental prognostic value for DCM patients.

Three-dimensional Fusion of Myocardial Perfusion SPECT and Invasive Coronary Angiography Guides Coronary Revascularization

BACKGROUND

SPECT myocardial perfusion imaging (SPECT MPI) and invasive coronary angiography (ICA) provide complementary clinical information in the diagnosis of coronary artery disease (CAD). We have developed an approach for 3D fusion of perfusion data from SPECT MPI and coronary anatomy from ICA. In this study, we aimed to evaluate its clinical value when compared to the traditional side-by-side readings.

METHODS

Thirty-six CAD patients who had at least one stenosis ≥ 50% were retrospectively enrolled. Based on the presence of a perfusion defect in a territory subtended by a coronary vessel, all vessels were classified as matched, unmatched, or normal groups via both the fusion and side-by-side analysis. The treatments recommended by the fusion and side-by-side analysis were compared with those that the patients received. Major adverse cardiac events (MACE), defined as all-cause death, myocardial infarction, unstable angina requiring hospitalization, and unplanned revascularization, were assessed.

RESULTS

The overall vessel-based concordance was 78.7% between the fusion and side-by-side analysis. Compared with the side-by-side analysis, 23 coronary arteries (29 equivocal segments) of 19 patients were reclassified via fusion of data. In the matched, unmatched, and normal groups, the numbers of vessels with hemodynamically significant stenosis which caused reversible defect were 37 vs 53, 28 vs 14, and 43 vs 41 (P < .01) when comparing the side-by-side analysis with the fusion, and the revascularization ratios per vessel were 69% vs 88%, 29% vs 10%, and 2% vs 2% between them. During the five-year follow-up, 8 patients (22.2%) experienced MACE. Patients who received the same treatment as the guidance of 3D fusion results (n = 22) had superior outcomes when compared with those who did not (n = 14) (P < .01).

CONCLUSIONS

Compared with the side-by-side analysis, the 3D fusion of SPECT MPI and ICA provided incremental diagnostic and prognostic value.

Development of an approach to extracting coronary arteries and detecting stenosis in invasive coronary angiograms

Purpose: In stable coronary artery disease (CAD), reduction in mortality and/or myocardial infarction with revascularization over medical therapy has not been reliably achieved. Coronary arteries are usually extracted to perform stenosis detection. As such, developing accurate segmentation of vascular structures and quantification of coronary arterial stenosis in invasive coronary angiograms (ICA) is necessary. Approach: A multi-input and multiscale (MIMS) U-Net with a two-stage recurrent training strategy was proposed for the automatic vessel segmentation. The proposed model generated a refined prediction map with the following two training stages: (i) stage I coarsely segmented the major coronary arteries from preprocessed single-channel ICAs and generated the probability map of arteries; and (ii) during the stage II, a three-channel image consisting of the original preprocessed image, a generated probability map, and an edge-enhanced image generated from the preprocessed image was fed to the proposed MIMS U-Net to produce the final segmentation result. After segmentation, an arterial stenosis detection algorithm was developed to extract vascular centerlines and calculate arterial diameters to evaluate stenotic level. Results: Experimental results demonstrated that the proposed method achieved an average Dice similarity coefficient of 0.8329, an average sensitivity of 0.8281, and an average specificity of 0.9979 in our dataset with 294 ICAs obtained from 73 patients. Moreover, our stenosis detection algorithm achieved a true positive rate of 0.6668 and a positive predictive value of 0.7043. Conclusions: Our proposed approach has great promise for clinical use and could help physicians improve diagnosis and therapeutic decisions for CAD.

Automatic identification of end-diastolic and end-systolic cardiac frames from invasive coronary angiography videos

BACKGROUND

Automatic identification of proper image frames at the end-diastolic (ED) and end-systolic (ES) frames during the review of invasive coronary angiograms (ICA) is important to assess blood flow during a cardiac cycle, reconstruct the 3D arterial anatomy from bi-planar views, and generate the complementary fusion map with myocardial images. The current identification method primarily relies on visual interpretation, making it not only time-consuming but also less reproducible.

OBJECITVE

In this paper, we propose a new method to automatically identify angiographic image frames associated with the ED and ES cardiac phases.

METHOD

A detection algorithm is first used to detect the key points (i.e. landmarks) of coronary arteries, and then an optical flow method is employed to track the trajectories of the selected key points. The ED and ES frames are identified based on all these trajectories. Our method was tested with 62 ICA videos from two separate medical centers.

RESULTS

Comparing consensus interpretations by two human expert readers, excellent agreement was achieved by the proposed algorithm: the agreement rates within a one-frame range were 92.99% and 92.73% for the automatic identification of the ED and ES image frames, respectively.

CONCLUSION

The proposed automated method showed great potential for being an integral part of automated ICA image analysis.