Wei-Chieh Huang

The traditional interpretation of Intravascular Ultrasound (IVUS) images during Percutaneous Coronary Intervention (PCI) is time-intensive and inconsistent, relying heavily on physician expertise. Regulatory restrictions and privacy concerns further hinder data integration across hospital systems, complicating collaborative analysis. To address these challenges, a parallel 2D U-Net model with a multi-stage segmentation architecture has been developed, utilizing federated learning to enable secure data analysis across institutions while preserving privacy. The model segments plaques by identifying and subtracting the External Elastic Membrane (EEM) and lumen areas, with preprocessing converting Cartesian to polar coordinates for improved computational efficiency. Achieving a Dice Similarity Coefficient (DSC) of 0.706, the model effectively identifies plaques and detects circular boundaries in real-time. Collaborative efforts with domain experts enhance plaque burden interpretation through precise quantitative measurements. Future advancements may involve integrating advanced federated learning techniques and expanding datasets to further improve performance and applicability. This adaptable technology holds promise for environments handling sensitive, distributed data, offering potential to optimize outcomes in medical imaging and intervention.

Background: The role of routine intravascular imaging in percutaneous coronary intervention (PCI) for acute myocardial infarction (AMI) remains unclear. This study evaluated the clinical outcomes of PCI guided by different imaging modalities in AMI patients.Materials and methods: Data from AMI patients who had undergone PCI between 2012 and 2022 were analyzed. The mean follow-up was 12.9 ± 1.73 months. The imaging modality-either intravascular ultrasound (IVUS), optical coherence tomography (OCT), or angiography alone-was selected at the operator's discretion. The primary endpoint was major adverse cardiac events (MACEs), including cardiovascular (CV) death, myocardial infarction (MI), target vessel revascularization.Results: Of the 1,304 PCIs performed, 47.5% (n = 620) were guided by angiography alone, 37.0% (n = 483) by IVUS, and 15.4% (n = 201) by OCT. PCI guided by intravascular imaging modalities was associated with lower 1-year rates of MI (1.3%, P = 0.001) and MACE (5.2%, P = 0.036). OCT-guided PCI was linked to lower rates of 1-year CV death (IVUS vs. OCT: 6.2% vs. 1.5%, P = 0.016) and MACE (IVUS vs. OCT: 6.4% vs. 2.5%, P = 0.032). Intravascular imaging modalities and diabetes were identified as predictors of better and worse 1-year MACE outcomes, respectively.Conclusion: PCI guided by intravascular imaging modalities resulted in improved 1-year clinical outcomes compared to angiography-guided PCI alone in AMI patients. OCT-guided PCI was associated with lower 1-year MACE rates compared to IVUS-guided PCI. Therefore, intravascular imaging should be recommended for PCI in AMI, with OCT being particularly considered when appropriate.

Aims. We aimed to evaluate the acute performance and short‐ and long‐term outcomes of optical coherence tomography (OCT)‐guided Magmaris deployment. Methods. This was a retrospective study of 28 consecutive patients (23 men, mean age: 59.8 years) with 28 Magmaris implantations in de novo coronary lesions. OCT was performed at the baseline and after the final postdilatation. The choice of stent and postdilatation balloon size was based on OCT measurements. The following indices were determined using OCT: prestenting minimum lumen diameter and area, poststenting minimum lumen diameter and area, acute lumen area gain, residual area stenosis, eccentricity and symmetry indices, incomplete strut apposition, strut fracture, tissue prolapse, and edge dissection. Results. Before the stenting, OCT analysis revealed a minimal lumen area of 1.55 ± 0.59 mm², a minimal lumen diameter of 1.19 ± 0.38 mm, a minimal scaffold area of 6.78 ± 1.58 mm², and a minimal scaffold diameter of 2.88 ± 0.50 mm². The prolapse area was 1.2 ± 1.5 mm². The mean percentage of RAS was 13.3 ± 7.1% and 6 (21.4%) patients had scaffold RAS more than 20%. Only one proximal edge intimal dissection was noted. The mean eccentricity index was 0.86 ± 0.04 and symmetry index 0.33 ± 0.08. ISA analysis showed that the percentage of malapposed struts was 1.5%. There were no short‐term cardiovascular events, and only 2 incidents of target lesion failure (TLF) occurred 13 months later. Conclusion. The Magmaris has excellent acute mechanical performance and no short‐term cardiovascular events occurred. There were only 2 TLFs that occurred 13 months later. It is suitable and feasible to treat vessels using the Magmaris.

Aims: Patients with high-flow arteriovenous (AV) access are at risk of developing high-output cardiac failure (HOCF) and subsequent hospitalization. However, diagnosing HOCF is challenging and often requires invasive procedures. The role of systemic vascular resistance (SVR) in diagnosing HOCF is underestimated, and its predictive value is limited. Our study aims to identify non-invasive risk factors for HOCF to facilitate early diagnosis and timely surgical interventions. Methods and results: We included 109 patients with high-flow AV access who underwent serial echocardiography. The retrospective cohort was divided into two groups based on their hospitalization due to HOCF. The two groups were matched for age and gender. After a mean follow-up of 25.1 months, 19 patients (17.4%) were hospitalized due to HOCF. The two groups had similar baseline characteristics. However, the HOCF group had a higher value of vascular access blood flow (Qa) (2168 ± 856 vs. 1828 ± 617 mL/min; P = 0.045). Echocardiographic analysis revealed that the HOCF group had more pronounced left ventricular diastolic dysfunction (E/e′: 21.1 ± 7.3 vs. 16.2 ± 5.9; P = 0.002), more severe pulmonary hypertension (right ventricular systolic pressure: 41.4 ± 16.7 vs. 32.2 ± 12.8; P = 0.009), a higher Doppler-derived cardiac index (CI) (4.3 ± 0.8 vs. 3.7 ± 1.1; P = 0.031), and a lower Doppler-derived estimated SVR (eSVR) value (5.5 ± 0.3 vs. 6.9 ± 0.2; P = 0.002) than the non-HOCF group. Using multivariable Cox regression analysis, a low eSVR value (<6) emerged as an independent predictor of HOCF hospitalization with a hazard ratio of 9.084 (95% confidence interval, 2.33–35.39; P = 0.001). Receiver operating characteristic curve analysis indicated that CI/eSVR values more accurately predicted HOCF hospitalization [sensitivity: 94.7%, specificity: 51.0%, area under the curve (AUC): 0.75, P < 0.001] than the Qa/cardiac output ratio (AUC: 0.50, P = 0.955), Qa values ≥ 2000 mL/min (AUC: 0.60, P = 0.181), and Qa values indexed for height in metres (AUC: 0.65, P = 0.040). Conclusions: In patients with high-flow AV access, low eSVR values obtained through non-invasive Doppler echocardiography were associated with a high rate of HOCF hospitalizations. Therefore, routine eSVR screening in these patients might expedite the diagnosis of HOCF.

Objective: Leveraging patient data through machine learning techniques in disease care offers a multitude of substantial benefits. Nonetheless, the inherent nature of patient data poses several challenges. Prevalent cases amass substantial longitudinal data owing to their patient volume and consistent follow-ups, however, longitudinal laboratory data are renowned for their irregularity, temporality, absenteeism, and sparsity; In contrast, recruitment for rare or specific cases is often constrained due to their limited patient size and episodic observations. This study employed self-supervised learning (SSL) to pretrain a generalized laboratory progress (GLP) model that captures the overall progression of six common laboratory markers in prevalent cardiovascular cases, with the intention of transferring this knowledge to aid in the detection of specific cardiovascular event. Methods and procedures: GLP implemented a two-stage training approach, leveraging the information embedded within interpolated data and amplify the performance of SSL. After GLP pretraining, it is transferred for target vessel revascularization (TVR) detection. Results: The proposed two-stage training improved the performance of pure SSL, and the transferability of GLP exhibited distinctiveness. After GLP processing, the classification exhibited a notable enhancement, with averaged accuracy rising from 0.63 to 0.90. All evaluated metrics demonstrated substantial superiority ( ${p} < 0.01$ ) compared to prior GLP processing. Conclusion: Our study effectively engages in translational engineering by transferring patient progression of cardiovascular laboratory parameters from one patient group to another, transcending the limitations of data availability. The transferability of disease progression optimized the strategies of examinations and treatments, and improves patient prognosis while using commonly available laboratory parameters. The potential for expanding this approach to encompass other diseases holds great promise. Clinical impact: Our study effectively transposes patient progression from one cohort to another, surpassing the constraints of episodic observation. The transferability of disease progression contributed to cardiovascular event assessment.

Background: Primary aldosteronism (PA) has been associated with atherosclerosis beyond the extent of essential hypertension, but the impact of albuminuria remains unknown. Objective: To investigate the effect of concomitant albuminuria on arterial stiffness in PA. Design: Prospective cohort study. Methods: A prospective cohort study was conducted to evaluate the association of albuminuria (>30 mg/g in morning spot urine) with arterial stiffness, as measured non-invasively by pulse wave velocity (PWV) in patients with PA. Propensity score matching (PSM) with age, sex, diabetes, systolic and diastolic blood pressure, creatinine, potassium, number of antihypertensive medications, and hypertension history was used to balance baseline characteristics. The effects of albuminuria on PWV before and 1 year after treatment were analyzed. Results: A total of 840 patients with PA were enrolled, of whom 243 had concomitant albuminuria. After PSM, there were no significant differences in baseline demographic parameters except alpha-blocker and spironolactone use. PWV was greater in the presence of albuminuria ( p = 0.012) and positively correlated with urine albumin–creatinine ratio. Multivariable regression analysis identified albuminuria, age, body weight, systolic blood pressure, and calcium channel blocker use as independent predictors of PWV. As for treatment response, only PA patients with albuminuria showed significant improvements in PWV after PSM ( p = 0.001). The magnitude of improvement in PWV increased with urine albumin–creatinine ratio and reached plateau when it exceeded 100 mg/g according to restricted cubic spline analysis. Conclusion: Concomitant albuminuria in PA was associated with greater arterial stiffness and more substantial improvement after targeted treatment. Both the baseline and the improved extent of PWV increased in correlation with rising urine albumin–creatinine ratio levels, reaching a plateau when the urine albumin–creatinine ratio surpassed 100 mg/g.

Primary aldosteronism (PA) is associated with higher arterial stiffness compared to essential hypertension (EH). However, few studies have compared different pulse wave velocity (PWV) parameters to detect aldosterone-induced arterial stiffness. In this study, we aimed to compare the sensitivity in detecting aldosterone-induced arterial stiffness between brachial-ankle PWV (baPWV) and heart-ankle PWV (haPWV). We prospectively enrolled 1006 PA patients and 983 EH patients. Detailed medical history, basic biochemistry data and two PWV measurements (baPWV and haPWV) were collected in both groups. We performed analysis on the original cohort and two propensity score matching (PSM) models (model 1 adjusted for age and sex; model 2 adjusted for age, sex, systolic and diastolic blood pressure). The DeLong test was used to compare areas under receiver operating characteristic curves (AUCs) between baPWV and haPWV to predict PA. In all models, the PA patients had significantly higher baPWV compared to the EH patients. The AUC of haPWV was greater than that of baPWV. In conclusion, haPWV seems to be a better PWV parameter than baPWV in detecting aldosterone-induced arterial stiffness.