In vivo quantitative assessment of atherosclerotic plaque: A validation study on the optical coherence tomography-derived index of plaque attenuation

Abstract

Growing evidence indicates that coronary plaque instability is an independent risk factor for adverse coronary events, yet current optical coherence tomography (OCT) assessment of high-risk plaque characteristics (HRPC) relies largely on qualitative interpretation. The index of plaque attenuation (IPA) is a quantitative OCT-based metric that may provide a more objective evaluation. This retrospective observational diagnostic accuracy study assessed the performance of OCT-derived IPA for HRPC detection in patients with acute coronary syndrome or stable angina, using expert consensus qualitative OCT analysis as the reference standard. A total of 560 analyzable OCT cross‐sections from 83 patients were categorized as HRPC or non-HRPC, with the latter grouped into progression plaque characteristics (PPC) or normal vessels. Receiver operating characteristic analysis was performed. IPA₁₁ distinguished HRPC from non-HRPC with an area under the curve of 0.75 (95% confidence interval: 70.3–79.0; P < .001), sensitivity of 76.1%, and specificity of 63.8%. IPA₁₁ was higher in HRPC than in non-HRPC (66.0 [38.0–110.0] vs 20.0 [2.0–52.0]; P = .001). In the non-HRPC group, IPA₈ differentiated PPC from normal vessels with an area under the curve of 0.98 (95% CI: 95.3–98.8; P < .001), with higher values in PPC (241.0 [155.5–320.5] vs 11.0 [3.5–30.5]; P < .001). IPA offers an objective, quantitative method for automated OCT-based plaque assessment, demonstrating good accuracy for HRPC identification and excellent discrimination of PPC from normal vessels, meriting validation in larger prospective studies.

1. Introduction

Coronary artery disease remains the leading cause of death and disability worldwide.^[1] In the prospective multicenter imaging study for evaluation of chest Pain (PROMISE) trial, high-risk plaque characteristics (HRPC) were associated with a 70% increased risk of major adverse cardiovascular events, independent of other risk factors or the presence of obstructive disease.^[2] Furthermore, a nationwide survey in China analyzed data from the China Stroke High-risk Population Screening and Intervention Program (2020–2021), including 194,878 adults aged 40 and older. The study highlighted the pivotal role of HRPC in stroke occurrence, emphasizing the urgent need for their early detection.^[3] HRPC includes any identifiable plaque characteristic such as fibrous cap thickness (FCT) < 75 μm, lipid arc circumferential extension > 180°, macrophage infiltration (MΦ), rupture of the fibrous cap (RFC), or intact fibrous cap erosion (IFC). Detection of HRPC using imaging tools plays a crucial role in guiding clinical decision-making for patients with known or suspected coronary artery disease.^[4–8]

Currently, noninvasive imaging modalities, such as coronary computed tomography angiography (CCTA) and magnetic resonance imaging, are widely used to characterize atherosclerotic plaques by providing macroscopic information about their features. However, these techniques are limited in their ability to detect microscopic details. In contrast, invasive imaging techniques, including optical coherence tomography (OCT) and intravascular ultrasound (IVUS), offer significantly higher resolution, allowing for the visualization of fine structural details within the plaques.^[4] OCT, with an axial resolution of approximately 10 μm, is highly suitable for both qualitative and quantitative assessment of HRPC. It demonstrates exceptional capability in identifying features such as MΦ and IFC, as well as accurately measuring FCT.^[5,6,9–11] More recently, OCT was upgraded to Class IA recommendation in the 2024 European Society of Cardiology guidelines for the interventional therapy of chronic coronary syndrome.^[12] Although the use of OCT in HRPC has great potential, its ability to perform quantitative measurements has not yet been fully investigated, limiting its widespread clinical adoption. The OCT-derived index of plaque attenuation (IPA) offers a quantitative approach to evaluate HRPC. It is safe, reproducible, cost-effective, and clinically validated.^[13] Importantly, sub-resolution tissue morphological changes during disease onset and progression alter the optical properties. These changes can be assessed through quantitative measurements of OCT signal decay with depth. These sub-resolution changes are not directly visible in imaging, leading to poor contrast in the tissue structures provided by conventional OCT. Notably, IPA can serve as a quantitative supplement to enhance the detection and analysis of subtle variations.^[14]

After OCT scanning, the light attenuation coefficient of each pixel is assessed. This creates a spectrum ranging from blue to yellow, which intuitively reflects the plaque stability. The proportion of pixels exhibiting “positive” light attenuation characteristics is then calculated and converted into an IPA value, allowing for quantitative analysis. The IPA values range from 0 to 1000, with higher values indicating greater plaque instability. This enables more informed decisions regarding interventional treatment.^[13–15] Nevertheless, IPA has yet to be validated for HRPC in vivo. Given this limitation, the present study was conducted to assess the ability of IPA to quantify HRPC.

2. Methods

2.1. Study population

The present study is a retrospective clinical trial designed to evaluate the quantitative capability of IPA in assessing HRPC within OCT images. Patients diagnosed with acute coronary syndromes or stable angina pectoris who underwent OCT imaging prior to percutaneous coronary intervention at the Chinese PLA General Hospital (Beijing, China) between January 2016 and December 2020 were included in the study. OCT imaging is used to analyze various types of atherosclerotic plaques, including those exhibiting HRPC. OCT images of poor quality, including those with massive thrombus or chronic total occlusion, were excluded during plaque screening. These images were not incorporated into the final analysis. The present study was approved by the ethics committee [Approval S2023-114-01] and was in accordance with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

2.2. OCT imaging acquisition and analysis

The OCT images were obtained using a frequency-domain OCT system and a Pathfinder 164 catheter (Corinas P60, Vivolight Co., Ltd., Shenzhen, China). OCT images were obtained using an automatic pull-back device with an acquisition speed of 20 mm/s, with a consistent pull-back time of 3 seconds. All matched images were submitted to the Image Core Laboratory (Senior Department of Cardiology) for off-line analysis. Two experienced cardiologists, using the LabelMe tool, analyzed and labeled areas of interest in the OCT images blindly, focusing on plaque types within each 1-mm segment of coronary lesions. In cases of discordance between the observers, a third investigator acted as a referee to achieve a consensus on the classification, and the majority opinion was used as the final plaque classification. All regions of interest were further classified into the following categories based on the “Consensus standards for acquisition, measurement, and reporting of intravascular OCT studies” document from the International Working Group for Intravascular OCT Standardization and Validation^[6,16]: MΦ as defined by a confluent or punctate highly backscattering focal region within the artery wall (as shown in Fig. 1A); TCFA as defined by a high backscattering and a relatively homogeneous OCT signal (thickness ≤ 65 µm) overlying a signal-poor region (as shown in Fig. 1B); IFC as defined by the presence of an attached thrombus (less backscattering, homogeneous, and low attenuation) overlying an intact and visualized plaque or luminal surface irregularity at the culprit lesion in the absence of thrombus, or attenuation of the underlying plaque by thrombus without superficial lipid or calcification immediately proximal or distal to the site of thrombus (as shown in Fig. 1C); RFC, defined by the presence of a fibrous cap (high backscattering) discontinuity with a cavity formation (as shown in Fig. 1D). FA, defined by a high backscattering and a relatively homogeneous OCT signal (thickness > 65 µm) overlying a signal-poor region (as shown in Fig. 1E); Calc exhibits signal-poor regions with sharply delineated borders and limited shadowing (as shown in Fig. 1F); Normal vessel as defined by layered architecture, comprising a highly backscattering intima, media with low backscattering, and highly backscattering adventitia (as shown in Fig. 1G).

Figure 1.

Representative cases of IPA image overlaid on the OCT image for coronary plaque characteristics and normal vessel (color: blue-red-yellow representing the maximum attenuation coefficient per A-line within the range of 0–12 mm^–1). (A) macrophage infiltration (white box), IPA: 125. (B) thin-cap fibroatheroma, IPA: 248. (C) erosion of an intact fibrous cap (white arrows), IPA: 233. (D) Rupture of the fibrous cap, IPA: 120. (E) Fibroatheroma, IPA: 73. (F) Fibrocalcific plaque, IPA: 78. (G) Normal vessel, IPA: 5. Asterisks indicate the guide-wire shadowing artifact. HRPC = high-risk plaque characteristics, IPA = index of plaque attenuation, OCT = optical coherence tomography, PPC = progression plaque characteristics.

Based on the criteria for the presence of HRPC derived from autopsy studies, TCFA, MΦ, IFC, and RFC were assigned to the HRPC group; normal vessel, FA, and Calc were assigned to the non-HRPC group. The non-HRPC group was further classified according to the guidelines of the updated classification scheme for atherosclerotic lesions: FA and Calc were assigned to the progression plaque characteristics (PPC) group; normal vessel was assigned to the normal vessel group.^[10,17,18]

2.3. Index of plaque attenuation

IPA is a measure of optical properties that is defined based on the index of optical attenuation to quantify tissue characteristics by fitting the OCT signal to a single scattering model.

$${\displaystyle \begin{array}{l}{\displaystyle \langle I_d\rangle =T\left(r\right)S\left(r\right)I_0{e}^{-{\mu }_{t^r}}}\end{array}}$$ (1)

where T(r) is the point spread function of the catheter and S(r) describes the OCT signal roll-off with depth. The attenuation coefficient ${\mu }_t$ is a parameter of interest. The model was fitted in the polar image, in every A-line starting from the edge of the lumen, in small windows of varying lengths to extract the attenuation coefficients. The accuracy of the extracted ${\mu }_t$ is approximately 1 mm⁻¹. The data analysis resulted in an attenuation image per frame of pullback (Fig. 2A).^[14,19]

Figure 2.

OCT image and attenuation image of a coronary atherosclerotic lesion in vivo. (A) Cross-sectional image of OCT and the corresponding attenuation image (color: blue-red-yellow representing the maximum attenuation coefficient per A-line within the range of 0–12 mm^–1). (B) Longitudinal image of OCT and the corresponding attenuation image in the entire vessel. OCT = optical coherence tomography.

Longitudinal attenuation maps were constructed to visualize the optical properties of the vessel’s intimal tissue throughout the entire pullback, and these were mapped to a blue-red-yellow color scale. The color coding indicated the maximum attenuation coefficient per A-line within the range of 0–12 mm⁻¹. In the en-face map, the X-axis indicated the pull-back position in millimeters (pixels every 0.2 mm), and the Y-axis indicated the circumferential position in degrees (pixels every 1°) as if the vessel had been split open along its longitudinal axis. Such a map may highlight regions with high attenuation characteristics throughout the pullback (Fig. 2B).

IPA is the ratio of the number of pixels with an attenuation coefficient greater than a certain threshold $x$ to the total number of pixels in the attenuation map, multiplied by a factor of 1000. The mathematical expression of this relationship is as follows:

$${\displaystyle \begin{array}{l}{\displaystyle IP{A}_x=\frac{N_{({\mu }_t>x)}}{N_{total}}\times 1,000}\end{array}}$$ (2)

where ${\mathrm{N}}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}$ is the total number of pixels, ${\mathrm{N}}_{({\mathrm{\mu }}_{\mathrm{t}}>x)}$ is the number of pixels greater than the threshold $x$, and $x$ is the threshold of the attenuation coefficient, with a maximum value of 12 mm⁻¹, which enables the IPA to represent a particular tissue type.

2.4. Statistical analysis

Data distribution was assessed using the Kolmogorov–Smirnov test. Continuous variables were expressed as the mean ± standard deviation or median with interquartile range were compared using Student t test, the Mann–Whitney U test or the Kruskal–Wallis test, as appropriate. Categorical outcome data were represented as n (%) and compared using the chi-squared test or Fisher exact test, as appropriate. Receiver operating characteristic (ROC) curve analyses were also performed. Sensitivity, and specificity with 95% confidence intervals (CIs) were obtained, and the maximum Youden index was used as a criterion for selecting the optimal cutoff point. The P-value (P < .05) with a 2-tailed test was used to determine the significance. All analyses were performed using SPSS version 26.0 (IBM Corp., Armonk).

3. Results

3.1. Study population and plaque characterization

A total of 95 patients (560 OCT images) who underwent OCT examination between January 2016 and December 2020 were enrolled in the study. Twelve patients (79 OCT images) were excluded owing to poor imaging quality, resulting in a study population of 83 patients (481 OCT images) for the analysis (Fig. 3). The average patient age was 57.2 ± 14 years, and the cohort included 58 males (69.9%) and 25 females (30.1%). The OCT images were categorized into 2 groups: HRPC and non-HRPC. The non-HRPC group comprised 318 images (318/481, 66.1%), including normal vessel tissues in 108 images (22%), FA in 117 images (24%), Calc in 93 images (19%). The HRPC group comprised 163 images (163/481, 33.9%), featuring TCFA in 52 images (11%), MΦ in 41 images (9%), RFC in 43 images (9%) and IFC in 27 images (6%), data are presented in Table 1. The outputs of the IPA for the HRPC and non-HRPC groups are illustrated in Figure 1.

Figure 3.

The study flow diagram. ACS = acute coronary syndrome, IPA = index of plaque attenuation, OCT = optical coherence tomography, PCI = percutaneous coronary intervention.

Table 1.

IPA values of non-HRPC group versus HRPC group (N = 481).

Non-HRPC (N = 318)				HRPC (N = 163)				Non-HRPC versus HRPC
IPAx	Normal vessel (N = 108)	PPC (N = 210)						P-Value
		FA (N = 117)	Calc (N = 93)	TCFA (N = 52)	MΦ (N = 41)	RFC (N = 43)	IFC (N = 27)
IPA8	10.0	194.0	288.0	324.0	272.0	236.0	280.0	.002
	(2.5–27.5)	(120.0–297.0)	(215.0–358.0)	(265.0–418.5)	(131.0–396.0)	(172.0–358.0)	(244.0–306.0)
IPA8.5	8.0	152.0	246.0	274.0	230.0	200.0	228.0	.011
	(0–22.0)	(86.0–240.0)	(159.0–299.0)	(226.0–370.0)	(88.0–335.0)	(231.0–286.0)	(194.0–242.0)
IPA9	6.0	112.0	190.0	224.0	186.0	136.0	182.0	.092
	(0–14.0)	(61.0–188.0)	(127.0–255.0)	(184.0–321.5)	(61.0–279.0)	(82.0–186.0)	(154.0–198.0)
IPA9.5	3.0	88.0	146.0	181.0	140.0	102.0	140.0	.221
	(0–11.5)	(42.0–140.0)	(95.0–213.0)	(146.5–280.0)	(44.0–225.5)	(64.0–142.0)	(114.0–158.0)
IPA10	2.0	62.0	104.0	144.0	106.0	62.0	96.0	.158
	(0–8.0)	(27.0–110.0)	(69.0–176.0)	(114.0–231.5)	(34.0–178.0)	(27.0–110.0)	(76.0–118.0)
IPA10.5	0	40.0	78.0	114.0	76.0	78.0	68.0	.049
	(0–6.0)	(18.0–80.0)	(48.0–145.0)	(86.0–195.5)	(23.0–134.0)	(48.0–124.0)	(48.0–90.0)
IPA11	0	26.0	52.0	86.0	56.0	56.0	44.0	.001
	(0–5.5)	(11.0–55.0)	(32.0–117.0)	(66.0–151.0)	(13.0–103.0)	(38.0–112.0)	(16.0–90.0)
IPA11.5	0	16.0	42.0	68.0	32.0	48.0	32.0	<.001
	(0–4.0)	(6.0–40.0)	(22.0–43.0)	(37.0–115.5)	(9.0–80.0)	(34.0–94.0)	(16.0–48.0)
IPA12	0	10.0	26.0	49.0	24.0	40.0	20.0	<.001
	(0–2.0)	(2.0–29.0)	(14.0–69.0)	(26.0–96.0)	(4.0–68.0)	(24.0–82.0)	(8.0–32.0)

Data are presented as median (interquartile range). Bold values indicate statistical significance (P < .05).

Calc = fibrocalcific plaque, FA = fibroatheroma, HRPC = high-risk plaque characteristics, IFC = erosion of an intact fibrous cap, IPA = index of plaque attenuation, MΦ = macrophage infiltration, PPC = progression plaque characteristics, RFC = rupture of the fibrous cap, TCFA = thin-cap fibroatheroma.

3.2. Quantitative capability of IPA for HRPC and PPC

The ROC curve was constructed for IPA values with different attenuation thresholds × as well as for HRPC and non-HRPC differentiation. Figure 4A presents the area under the curve (AUC) values corresponding to different threshold levels of × in IPA. The threshold “x” that corresponds to the highest AUC is recognized as the optimal threshold, signifying its discriminative superiority. IPA, with a threshold of x = 11 mm⁻¹ (denoted as IPA₁₁) was the strongest differentiator between HRPC and non-HRPC (Fig. 4A). The accuracy value was 0.75 (95% CI: 70.3–79.0). The corresponding sensitivity value was 76.1% (95% CI: 68.8–82.4), specificity value was 63.4% (95% CI: 58.3–69.1), and optimal cutoff value of IPA₁₁ was 36 (Fig. 4B).

Figure 4.

Receiver operating characteristic curve analysis for HRPC and non-HRPC differentiation. (A) The maximum thresholds “x” in IPA_x. (B) The AUC indicates that IPA₁₁ can significantly differentiate HRPC and non-HRPC. AUC = the area under the curve, CI = confidence interval, HRPC = high-risk plaque characteristics, IPA = index of plaque attenuation.

Additionally, the ROC curve was constructed for IPA values with different attenuation thresholds × as well as for PPC and normal vessel tissue differentiation. IPA with a threshold x = 8 mm⁻¹ (denoted as IPA₈) was the strongest differentiator between PPC and normal vessel tissues (Fig. 5A). The accuracy value was 0.98 (95% CI: 95.3–98.8). The corresponding sensitivity value was 94.3% (95% CI: 90.2–97.0), specificity value was 94.4% (95% CI: 88.3–97.9), and optimal cutoff value of IPA₈ was 76 (Fig. 5B).

Figure 5.

Receiver operating characteristic curve analysis for PPC and normal vessel differentiation. (A) The maximum thresholds “x” in IPA_x. (B) The AUC indicates that IPA₈ can significantly differentiate PPC and normal vessel. AUC = the area under the curve, CI = confidence interval, IPA = index of plaque attenuation, PPC = progression plaque characteristics.

3.3. Comparison with IPA values of HRPC group versus non-HRPC and PPC group versus normal vessel group

We compared the IPA values that were regarded as the most significant differentiators, measured by OCT in the HRPC and non-HRPC groups. HRPC group had significant higher IPA₁₁ values than the non-HRPC group [HRPC group: 66.0 (38.0–110.0) vs non-HRPC group: 20.0 (2.0–52.0); P = .001], as shown in Figure 6A. Additionally, the PPC group had significantly higher IPA₈ values than the normal vessel group: [PPC group: 241.0 (155.5–320.5) vs normal vessel group: 11.0 (3.5–30.5); P < .001], as shown in Figure 6B.

Figure 6.

Box plot of IPA₁₁ values for HRPC group versus non-HRPC group and IPA₈ values for PPC group versus normal vessel group. (A) non-HRPC group (blue): normal vessel, fibroatheroma, fibrocalcific plaque. HRPC group (red): macrophage infiltration, thin-cap fibroatheroma, erosion of an intact fibrous cap, rupture of the fibrous cap. IPA₁₁ values of the HRPC group [66.0 (38.0–110.0)] were significantly higher than the non-HRPC group [20.0 (2.0–52.0)]. (B) Normal vessel group (blue): normal vessel. PPC group (red): fibroatheroma, fibrocalcific plaque. IPA₈ values of the PPC group [241.0 (155.5–320.5)] were significantly higher than the normal vessel group [11.0 (3.5–30.5)]. HRPC = high-risk plaque characteristics, IPA = index of plaque attenuation, PPC = progression plaque characteristics.

4. Discussion

This study is the first to demonstrate the in vivo feasibility of a novel approach and assess the potential of IPA in quantifying atherosclerotic plaques in a clinical setting. The main findings were as follows: First, IPA₁₁ exhibited the highest ability to differentiate between HRPC and non-HRPC (AUC = 0.75), with IPA values in the HRPC group were significantly higher than those in the non-HRPC group. Second, IPA₈ demonstrated the highest ability to differentiate between PPC and normal vessel (AUC = 0.98), with IPA values in the PPC group were significantly higher than those in the normal vessel group. In vivo validation demonstrated that IPA was a reliable quantitative tool for assessing atherosclerotic plaques and provided crucial insights into plaque stability.

Evidence suggests that HRPC has prognostic value in identifying individuals at high risk for acute events.^{[10,17,20,21]} The progression of atherosclerotic plaque determined by CCTA imaging (PARADIGM) study showed that in lesions ≥ 2 HRPC, statin therapy resulted in a greater change in annual percent atheroma volume compared to lesions without HRPC.^[22] Consequently, imaging-based quantification could be improved by choosing the most appropriate imaging modality for diagnosis and treatment. Currently, according to the clinical consensus recommendations on the appropriateness of each imaging technique, CCTA enables quantitative assessment of coronary plaque with respect to dimensions, composition, location and related risk of future cardiovascular events. Magnetic resonance imaging enables coronary plaque visualization and serves as a radiation-free, second-line imaging option for noninvasive coronary angiography. Positron emission tomography offers the highest potential for the assessment of coronary plaque inflammation.^[4] In contrast, IVUS and OCT are essential invasive imaging modalities that provide high-resolution images, enabling assessment of plaques at high risk of rupture.^[23] Briefly, IVUS provides cross-sectional vessel images (100–250 µm) for assessing luminal and media-adventitia borders, plaque burden, and composition.^[4,24] OCT provides high-resolution imaging (10–20 µm), allowing for detailed visualization of vessel wall layers and microstructural features that are beyond the detection capabilities of other imaging techniques. These include MΦ, neo-vessels, micro-calcifications, characterization of thrombus types and FCT over lipid-rich plaques.^[16] However, quantitative assessment of plaque characterization is currently a challenging, time-consuming, and difficult-to-systematize process that relies heavily on operator subjective interpretation and skill. Artificial intelligence (AI) may reduce cost and improve value.^[25] Recent studies have demonstrated that AI-guided CCTA enables the rapid quantification of HRPC, which provides independent predictive value for future atherosclerotic cardiovascular disease events during short- to intermediate-term follow-up. These findings highlight the potential of AI-guided CCTA in improving risk stratification and early intervention strategies for atherosclerotic cardiovascular disease.^[8,26] Additionally, AI algorithms can classify plaque components in IVUS and OCT images, boosting the diagnostic accuracy and workflow efficiency. They can also predict coronary events by integrating imaging data with clinical and demographic information.^[27] Currently, to the best of our knowledge, there is no AI research capable of simultaneously quantifying multiple HRPC.

The automatic algorithm of OCT-derived IPA enabled in vivo quantification of atherosclerotic plaque throughout the pull-back procedure. This eliminates the need for time-intensive frame-by-frame inspection and manual measurement. Pathological studies have demonstrated that the optical attenuation coefficient in OCT images facilitates the differentiation of high-risk plaque components (e.g., lipid core and MΦ) from low-risk plaque components (e.g., fibrous tissue, calcified tissue, and healthy tissue). Furthermore, it enables quantitative evaluation of various types of atherosclerotic tissues in autopsy samples.^[14,28] Van Soest et al highlighted the promise of OCT-derived IPA in quantifying various plaque characteristics in vivo, reporting that TCFA exhibits significantly higher IPA values than FA [TCFA: 141 (98–159) vs FA: 60 (37–103); P < .001].^[13] However, the study did not quantify MΦ, RFC, and IFC, which are critical for comprehensive HRPC assessments.^[13] Our observation showed that higher IPA values were associated with HRPC compared to non-HRPC.

Research has shown that color-coded IPA₁₁ displayed on the longitudinal OCT section, similar to the block chemogram used in near-infrared spectroscopy-intravascular ultrasound, highlights the presence of TCFA, in sections with IPA₁₁ > 110 (AUC = 0.825).^[13] We also found that the ability of IPA₁₁ for distinguishing HRPC and non-HRPC was 0.75 (95% CI: 70.3–79.0), and the HRPC group (TCFA, MΦ, RFC, IFC) had significantly higher IPA values than the non-HRPC group [HRPC group: 66.0 (38.0–110.0) vs non-HRPC group: 20.0 (2.0–52.0); P = .001]. Interestingly, we found that fibrocalcific plaques of the non-HRPC group in the data set exhibited elevated IPA₁₁, contradicting previous pathological studies. Earlier reports on the attenuation of purely calcified plaques demonstrated low attenuation.^[14,15] A possible explanation for this discrepancy is the variation in the lipid-necrotic core, macrophage content, and cap thickness among different fibrocalcific plaques, which may contribute to higher IPA values.^[11] Additionally, the present investigation indicates that the ability of IPA₈ for distinguishing PPC and normal vessel was 0.98 (95% CI: 95.3–98.8); PPC group had significantly higher IPA values than normal vessel group [PPC group: 241.0 (155.5–320.5) vs normal vessel group: 11.0 (3.5–30.5); P < .001]. This method significantly reduces the time required for OCT technicians to analyze abnormal coronary plaques. Overall, IPA enables the automated quantification of atherosclerotic coronary plaques, aiding in the prediction of disease progression and event risk. Furthermore, it has the potential to improve risk stratification and identify patients who may benefit from more intensive therapies.^[2,7,22,29]

Although IPA technology is highly promising, further clinical studies are essential to substantiate its precise correlation with clinical outcomes and to establish an IPA threshold predictive of cardiovascular events. Moreover, the role of the IPA in guiding percutaneous coronary intervention and postoperative medication strategies warrants additional exploration through a large number of clinical studies. To date, the ongoing OCT imaging for coronary plaque assessment and risk evaluation (OCT-CARE) study (ClinicalTrial.org Identifier: ChiCTR2300069716) serves as a valuable tool to confirm the effectiveness of IPA. Our study has several limitations. First, the IPA values of different components within fibrocalcific plaques require further exploratory studies using clinical and ex vivo data. Second, no clear-cut distinction between patients with acute coronary syndromes or stable angina pectoris was made for the present analyses. Finally, this was a retrospective study with a small number of patients. Hence, there was a possibility of selection bias. Data pooling with other large-scale studies using IPA might improve the statistical power in future analyses.

5. Conclusions

Our findings demonstrate the potential of IPA for the quantification of atherosclerotic plaques, particularly HRPC, offering valuable insights into plaque stability. However, its role remains promising but still understudied, warranting further investigation to fully validate its clinical utility.

Author contributions

Conceptualization: Shanshan Zhou, Yundai Chen.

Data curation: Xingxuan Cai, Yingyun Hu, Mingyi Wang.

Formal analysis: Jing Jing, Kai Wei, Shanshan Zhou.

Methodology: Xingxuan Cai.

Project administration: Yundai Chen.

Software: Yihui Cao, Quanmao Lu.

Writing – original draft: Xingxuan Cai.

Writing – review & editing: Yihui Cao, Shanshan Zhou.