Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography

Valentin Koch; Olle Holmberg; Edna Blum; Ece Sancar; Alp Aytekin; Masaru Seguchi; Erion Xhepa; Jens Wiebe; Salvatore Cassese; Sebastian Kufner; Thorsten Kessler; Hendrik Sager; Felix Voll; Tobias Rheude; Tobias Lenz; Adnan Kastrati; Heribert Schunkert; Julia A Schnabel; Michael Joner; Carsten Marr; Philipp Nicol

doi:10.1038/s43856-025-00835-5

Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography

Valentin Koch, Olle Holmberg, Edna Blum, Ece Sancar, Alp Aytekin, Masaru Seguchi, Erion Xhepa, Jens Wiebe, Salvatore Cassese, Sebastian Kufner, Thorsten Kessler, Hendrik Sager, Felix Voll, Tobias Rheude, Tobias Lenz, Adnan Kastrati, Heribert Schunkert, Julia A Schnabel, Michael Joner, Carsten MarrPhilipp Nicol

Biomedical Computing

Research output: Contribution to journal › Article › peer-review

Abstract

BACKGROUND: Accurate interpretation of optical coherence tomography (OCT) pullbacks is critical for assessing vascular healing after percutaneous coronary intervention (PCI). Manual analysis is time-consuming and subjective, highlighting the need for a fully automated solution.

METHODS: In this study, 1148 frames from 92 OCT pullbacks were manually annotated to classify neointima as homogeneous, heterogeneous, neoatherosclerosis, or not analyzable on a quadrant level. Stent and lumen contours were annotated in 305 frames for segmentation of the lumen, stent struts, and neointima. We used these annotations to train a deep learning algorithm called DeepNeo. Performance was further evaluated in an animal model (male New Zealand White Rabbits) of neoatherosclerosis using co-registered histopathology images as the gold standard.

RESULTS: DeepNeo demonstrates a strong classification performance for neointimal tissue, achieving an overall accuracy of 75%, which is comparable to manual classification accuracy by two clinical experts (75% and 71%). In the animal model of neoatherosclerosis, DeepNeo achieves an accuracy of 87% when compared with histopathological findings. For segmentation tasks in human pullbacks, the algorithm shows strong performance with mean Dice overlap scores of 0.99 for the lumen, 0.66 for stent struts, and 0.86 for neointima.

CONCLUSIONS: To the best of our knowledge, DeepNeo is the first deep learning algorithm enabling fully automated segmentation and classification of neointimal tissue with performance comparable to human experts. It could standardize vascular healing assessments after PCI, support therapeutic decisions, and improve risk detection for cardiac events.

Original language	English
Article number	124
Pages (from-to)	124
Journal	Communications Medicine
Volume	5
Issue number	1
DOIs	https://doi.org/10.1038/s43856-025-00835-5
Publication status	Published - 17 Apr 2025

Access to Document

10.1038/s43856-025-00835-5

Cite this

Koch, V., Holmberg, O., Blum, E., Sancar, E., Aytekin, A., Seguchi, M., Xhepa, E., Wiebe, J., Cassese, S., Kufner, S., Kessler, T., Sager, H., Voll, F., Rheude, T., Lenz, T., Kastrati, A., Schunkert, H., Schnabel, J. A., Joner, M., ... Nicol, P. (2025). Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography. Communications Medicine, 5(1), 124. Article 124. https://doi.org/10.1038/s43856-025-00835-5

@article{370cc189ce04439fa7f86f7a3e70bf68,

title = "Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography",

abstract = "BACKGROUND: Accurate interpretation of optical coherence tomography (OCT) pullbacks is critical for assessing vascular healing after percutaneous coronary intervention (PCI). Manual analysis is time-consuming and subjective, highlighting the need for a fully automated solution.METHODS: In this study, 1148 frames from 92 OCT pullbacks were manually annotated to classify neointima as homogeneous, heterogeneous, neoatherosclerosis, or not analyzable on a quadrant level. Stent and lumen contours were annotated in 305 frames for segmentation of the lumen, stent struts, and neointima. We used these annotations to train a deep learning algorithm called DeepNeo. Performance was further evaluated in an animal model (male New Zealand White Rabbits) of neoatherosclerosis using co-registered histopathology images as the gold standard.RESULTS: DeepNeo demonstrates a strong classification performance for neointimal tissue, achieving an overall accuracy of 75%, which is comparable to manual classification accuracy by two clinical experts (75% and 71%). In the animal model of neoatherosclerosis, DeepNeo achieves an accuracy of 87% when compared with histopathological findings. For segmentation tasks in human pullbacks, the algorithm shows strong performance with mean Dice overlap scores of 0.99 for the lumen, 0.66 for stent struts, and 0.86 for neointima.CONCLUSIONS: To the best of our knowledge, DeepNeo is the first deep learning algorithm enabling fully automated segmentation and classification of neointimal tissue with performance comparable to human experts. It could standardize vascular healing assessments after PCI, support therapeutic decisions, and improve risk detection for cardiac events.",

author = "Valentin Koch and Olle Holmberg and Edna Blum and Ece Sancar and Alp Aytekin and Masaru Seguchi and Erion Xhepa and Jens Wiebe and Salvatore Cassese and Sebastian Kufner and Thorsten Kessler and Hendrik Sager and Felix Voll and Tobias Rheude and Tobias Lenz and Adnan Kastrati and Heribert Schunkert and Schnabel, {Julia A} and Michael Joner and Carsten Marr and Philipp Nicol",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2025.",

year = "2025",

month = apr,

day = "17",

doi = "10.1038/s43856-025-00835-5",

language = "English",

volume = "5",

pages = "124",

journal = "Communications Medicine",

issn = "2730-664X",

publisher = "Nature Research",

number = "1",

}

Koch, V, Holmberg, O, Blum, E, Sancar, E, Aytekin, A, Seguchi, M, Xhepa, E, Wiebe, J, Cassese, S, Kufner, S, Kessler, T, Sager, H, Voll, F, Rheude, T, Lenz, T, Kastrati, A, Schunkert, H, Schnabel, JA, Joner, M, Marr, C & Nicol, P 2025, 'Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography', Communications Medicine, vol. 5, no. 1, 124, pp. 124. https://doi.org/10.1038/s43856-025-00835-5

TY - JOUR

T1 - Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography

AU - Koch, Valentin

AU - Holmberg, Olle

AU - Blum, Edna

AU - Sancar, Ece

AU - Aytekin, Alp

AU - Seguchi, Masaru

AU - Xhepa, Erion

AU - Wiebe, Jens

AU - Cassese, Salvatore

AU - Kufner, Sebastian

AU - Kessler, Thorsten

AU - Sager, Hendrik

AU - Voll, Felix

AU - Rheude, Tobias

AU - Lenz, Tobias

AU - Kastrati, Adnan

AU - Schunkert, Heribert

AU - Schnabel, Julia A

AU - Joner, Michael

AU - Marr, Carsten

AU - Nicol, Philipp

PY - 2025/4/17

Y1 - 2025/4/17

N2 - BACKGROUND: Accurate interpretation of optical coherence tomography (OCT) pullbacks is critical for assessing vascular healing after percutaneous coronary intervention (PCI). Manual analysis is time-consuming and subjective, highlighting the need for a fully automated solution.METHODS: In this study, 1148 frames from 92 OCT pullbacks were manually annotated to classify neointima as homogeneous, heterogeneous, neoatherosclerosis, or not analyzable on a quadrant level. Stent and lumen contours were annotated in 305 frames for segmentation of the lumen, stent struts, and neointima. We used these annotations to train a deep learning algorithm called DeepNeo. Performance was further evaluated in an animal model (male New Zealand White Rabbits) of neoatherosclerosis using co-registered histopathology images as the gold standard.RESULTS: DeepNeo demonstrates a strong classification performance for neointimal tissue, achieving an overall accuracy of 75%, which is comparable to manual classification accuracy by two clinical experts (75% and 71%). In the animal model of neoatherosclerosis, DeepNeo achieves an accuracy of 87% when compared with histopathological findings. For segmentation tasks in human pullbacks, the algorithm shows strong performance with mean Dice overlap scores of 0.99 for the lumen, 0.66 for stent struts, and 0.86 for neointima.CONCLUSIONS: To the best of our knowledge, DeepNeo is the first deep learning algorithm enabling fully automated segmentation and classification of neointimal tissue with performance comparable to human experts. It could standardize vascular healing assessments after PCI, support therapeutic decisions, and improve risk detection for cardiac events.

AB - BACKGROUND: Accurate interpretation of optical coherence tomography (OCT) pullbacks is critical for assessing vascular healing after percutaneous coronary intervention (PCI). Manual analysis is time-consuming and subjective, highlighting the need for a fully automated solution.METHODS: In this study, 1148 frames from 92 OCT pullbacks were manually annotated to classify neointima as homogeneous, heterogeneous, neoatherosclerosis, or not analyzable on a quadrant level. Stent and lumen contours were annotated in 305 frames for segmentation of the lumen, stent struts, and neointima. We used these annotations to train a deep learning algorithm called DeepNeo. Performance was further evaluated in an animal model (male New Zealand White Rabbits) of neoatherosclerosis using co-registered histopathology images as the gold standard.RESULTS: DeepNeo demonstrates a strong classification performance for neointimal tissue, achieving an overall accuracy of 75%, which is comparable to manual classification accuracy by two clinical experts (75% and 71%). In the animal model of neoatherosclerosis, DeepNeo achieves an accuracy of 87% when compared with histopathological findings. For segmentation tasks in human pullbacks, the algorithm shows strong performance with mean Dice overlap scores of 0.99 for the lumen, 0.66 for stent struts, and 0.86 for neointima.CONCLUSIONS: To the best of our knowledge, DeepNeo is the first deep learning algorithm enabling fully automated segmentation and classification of neointimal tissue with performance comparable to human experts. It could standardize vascular healing assessments after PCI, support therapeutic decisions, and improve risk detection for cardiac events.

UR - http://www.scopus.com/inward/record.url?scp=105003113203&partnerID=8YFLogxK

U2 - 10.1038/s43856-025-00835-5

DO - 10.1038/s43856-025-00835-5

M3 - Article

C2 - 40247001

SN - 2730-664X

VL - 5

SP - 124

JO - Communications Medicine

JF - Communications Medicine

IS - 1

M1 - 124

ER -

Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography

Abstract

Access to Document

Other files and links

Fingerprint

Cite this