Pith. sign in

REVIEW 2 major objections 1 minor 34 references

Deep Learning-Based Automated Quantification of TIMI Myocardial Perfusion Frame Count (DL-TMPFC) from Coronary Angiography: A Novel Framework for Rapid Assessment of Microvascular Dysfunction

T0 review · 2 major / 1 minor · reviewed 2026-06-30 · grok-4.3

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Deep learning framework automates TMPFC calculation to quantify coronary microvascular dysfunction from routine angiography.

desk verdict The paper automates TMPFC calculation via DL on a 655-patient multi-center set with strong reported agreement, but the abstract leaves the critical segmentation and frame-selection accuracy untested. read the letter →

arxiv 2605.24012 v1 pith:IM7LVYCX submitted 2026-05-20 cs.CV

Si Li , Yuanqing He , Chenkai Hu , Xiaogang Guo , Huay-Cheem Tan , Chieh Yang Koo , Xuan Zhang , Lei He
show 2 more authors
Jingyuan Zeng Shan Xiao
This is my paper · ORCID
classification cs.CV
keywords deeplearningcoronaryangiographymicrovasculardysfunctionTIMIframecountautomatedquantificationperfusionassessmentCMVD
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper develops and validates DL-TMPFC, a deep learning system that automatically computes the TIMI Myocardial Perfusion Frame Count from coronary angiograms after first excluding obstructive disease. It combines a stenosis detection network with a territory-aware segmentation network to locate perfusion territories and identify the start and end frames of contrast flow. Tested on 655 patients across three centers, the system matched expert manual counts with bias of -0.93 frames and correlation 0.98, while distinguishing microvascular dysfunction across obstructive, non-obstructive, and control cases and tracking its continuous severity. The automation removes manual calculation steps and observer variability, allowing TMPFC to move from research tool to routine clinical measure. This supplies an objective, angiography-based metric for a condition that currently lacks simple bedside quantification.

What carries the argument

Territory-aware segmentation network that locates perfusion territories and marks the first and last frames of myocardial contrast appearance for TMPFC computation.

What would settle it

A new multi-center cohort in which DL-TMPFC values deviate from simultaneous manual expert counts by more than the reported limits of agreement on average.

Watch

Extended reading notes

Core claim

DL-TMPFC delivers automatic and objective TMPFC values directly from standard angiographic sequences, achieving excellent agreement with manual expert readings and correctly classifying microvascular dysfunction across the full range of coronary artery disease presentations.

Load-bearing premise

The segmentation network will correctly define perfusion territories and mark first and last frames even when imaging angles, patient anatomy, or equipment vary from the training data.

Editorial extensions

If this is right

  • TMPFC becomes feasible as a standard reportable value on every diagnostic angiogram without added procedure time or staff effort.
  • Clinicians gain a continuous numeric score rather than a binary yes/no for microvascular dysfunction, supporting graded risk assessment.
  • Observer-to-observer differences in perfusion grading disappear because the calculation is fully deterministic once the images are acquired.
  • The same angiographic run used for stenosis assessment now also yields a microvascular metric without extra contrast or radiation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Hospitals could embed the model in existing cath-lab software so that TMPFC appears automatically in the procedural report.
  • Longitudinal tracking of the same patient’s TMPFC values before and after therapy becomes practical for monitoring treatment response.
  • The framework’s exclusion of obstructive disease first may allow combined reporting of epicardial and microvascular status from one study.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 1 minor

Summary. The manuscript presents DL-TMPFC, a deep learning framework for automated TMPFC quantification from coronary angiography to assess CMVD. It comprises a stenosis detection network to exclude obstructive CAD, a territory-aware segmentation network to identify perfusion territories, and an automatic first/last frame determination module. Validation on a 655-patient multi-center cohort (445 obstructive CAD, 100 confirmed CMVD, 110 controls) reports excellent agreement with manual expert measurements (bias -0.93 frames; 95% LoA -5.33 to +3.47; r=0.98) and accurate CMVD identification across pathologies with continuous severity capture.

Significance. If the segmentation and frame-selection components prove robust, the work could enable objective, observer-independent TMPFC measurement in routine angiography, supporting quantitative CMVD risk stratification and clinical translation beyond subjective TIMI grading. The multi-center cohort size and reported agreement statistics provide empirical grounding for the automation claim.

major comments (2)
  1. [Methods (DL-TMPFC framework)] Methods (DL-TMPFC framework description): The territory-aware segmentation network and first/last frame detection are load-bearing for all downstream TMPFC values and agreement metrics, yet no segmentation performance metrics (Dice/IoU per territory), frame-selection error rates, or ablation studies on imaging variations (projection angle, contrast timing) are reported. This leaves the central assumption untested and directly undermines confidence in the bias/LoA/r values and CMVD classification results.
  2. [Results (validation cohort)] Results (validation on 655-patient cohort): No details are provided on training/validation splits, external-site testing of the full pipeline, or handling of edge cases/anatomies outside the three-institution set. Any systematic territory or frame error would propagate into the reported agreement statistics, making the generalizability claim difficult to evaluate.
minor comments (1)
  1. [Abstract] Abstract: The statement that DL-TMPFC 'accurately identified CMVD across a full spectrum of coronary pathologies' would benefit from explicit reference to the specific performance metrics (e.g., sensitivity/specificity or correlation with continuous severity) supporting this claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed review. The comments highlight important aspects of transparency for the DL-TMPFC framework. We respond to each major comment below and indicate planned revisions where appropriate.

read point-by-point responses

  1. Referee: Methods (DL-TMPFC framework): The territory-aware segmentation network and first/last frame detection are load-bearing for all downstream TMPFC values and agreement metrics, yet no segmentation performance metrics (Dice/IoU per territory), frame-selection error rates, or ablation studies on imaging variations (projection angle, contrast timing) are reported. This leaves the central assumption untested and directly undermines confidence in the bias/LoA/r values and CMVD classification results.

    Authors: We agree that intermediate performance metrics would increase transparency and allow readers to better assess potential error propagation. In the revised manuscript we will report per-territory Dice and IoU scores for the segmentation network, frame-selection accuracy (mean absolute error and percentage of frames within acceptable tolerance), and a focused ablation on projection angle and contrast timing variations using the available multi-center data. The primary clinical validation remains the end-to-end agreement with expert TMPFC measurements, which directly tests the quantity of interest; however, the additional metrics will strengthen the supporting evidence. revision: yes

  2. Referee: Results (validation on 655-patient cohort): No details are provided on training/validation splits, external-site testing of the full pipeline, or handling of edge cases/anatomies outside the three-institution set. Any systematic territory or frame error would propagate into the reported agreement statistics, making the generalizability claim difficult to evaluate.

    Authors: The 655-patient cohort was collected from three independent institutions specifically to improve diversity. We will expand the Methods section to explicitly state the training/validation split ratios and any site-stratified partitioning used during model development. While the multi-center design already incorporates data from separate sites, we will add a leave-one-site-out analysis where computationally feasible. Edge cases (e.g., anomalous coronary anatomy, suboptimal contrast opacification) will be illustrated with representative examples and failure-mode discussion. These additions will be included in the revision. revision: partial

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; central claims rest on empirical comparison to independent manual annotations

full rationale

The paper describes a DL pipeline (stenosis detection + territory-aware segmentation + frame selection) whose performance is assessed solely via direct numerical agreement with separate expert manual TMPFC measurements on a held-out multi-center cohort of 655 patients. No equations, fitted parameters, or self-citations are presented that would make the reported bias, limits of agreement, or correlation reduce to quantities defined or optimized inside the same dataset. The validation therefore remains an external empirical check rather than a self-referential construction.

Assumptions & free parameters 0 free parameters · 0 assumptions · 0 invented entities

The central claim rests on empirical agreement between an automated pipeline and human annotations in a held-out cohort; no additional free parameters, mathematical axioms, or postulated entities are required beyond standard supervised deep-learning training.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Deep Learning-Based Automated Quantification of TIMI Myocardial Perfusion Frame Count (DL-TMPFC) from Coronary Angiography: A Novel Framework for Rapid Assessment of Microvascular Dysfunction." pith.science (2026). https://pith.science/paper/IM7LVYCX

@misc{pith2026260524012,
  author       = {Pith},
  title        = {Pith review of: Deep Learning-Based Automated Quantification of TIMI Myocardial Perfusion Frame Count (DL-TMPFC) from Coronary Angiography: A Novel Framework for Rapid Assessment of Microvascular Dysfunction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IM7LVYCX}},
  note         = {Machine review of arXiv:2605.24012}
}
read the original abstract

Aims: Coronary microvascular dysfunction (CMVD) affects approximately 40%-60% of patients with ischemia and non-obstructive coronary arteries, yet diagnosis remains challenging due to reliance on invasive functional testing or subjective Thrombolysis In Myocardial Infarction (TIMI) flow grade. The TIMI Myocardial Perfusion Frame Count (TMPFC) offers an objective, angiography-based quantitative measure of CMVD, but its clinical translation is hindered by cumbersome manual calculation and insufficient validation. This study aims to develop and validate a deep learning-powered TMPFC calculation (DL-TMPFC), enabling integration into clinical workflows. Methods and results: DL-TMPFC framework comprised two components. A stenosis detection network first excluded obstructive coronary artery disease (CAD). A territory-aware segmentation network then identified perfusion territories and TMPFC calculation module automatically determined the first and last frames from angiographic sequences. The framework was validated in a cohort of 655 patients (445 of obstructive CAD, 100 of confirmed CMVD, 110 of control group) from three independent institutions. DL-TMPFC showed excellent agreement with expert manual measurements (bias: -0.93 frames; 95% LoA: -5.33 to +3.47; r =0.98). DL-TMPFC markedly enhanced clinical feasibility by fully automating TMPFC and removing observer dependence. Clinically, DL-TMPFC accurately identified CMVD across a full spectrum of coronary pathologies and captured the continuous severity of CMVD beyond binary classification, enabling quantitative risk stratification. Conclusion: DL-TMPFC enabled automatic, standardized, and accurate quantification of CMVD directly from routine angiography. By providing an automatic and objective measure, this tool provided immediate diagnostic information for timely recognition and management of CMVD in clinical practice.

Figures

Figures reproduced from arXiv: 2605.24012 by the authors.

Figure 3
Figure 3. Detailed description of the development of DL-driven TMPFC measurement method. n frame 𝐹𝑚𝑎𝑥 was identified as the frame with the maximum 𝐴𝑡 . Based on the smoothed 𝐴𝑡 sequence, the initial filling frame 𝐹1 and the first clearance frame 𝐹2 were detected (shown in [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 34 canonical work pages

  1. [1]

    Patient enrollment and data acquisition This retrospective analysis included 655 consecutive patients who underwent elective percutaneous coronary intervention (PCI) at the Second Affiliated Hospital of Nanchang University (Nanchang, China), the First Affiliated Hospital of Zhejiang University School of Medicine (Hangzhou, China) and the National Universi...

    work page 2023

  2. [2]

    For stenosis detection network, the training set contained opacified frames with manually annotated bounding boxes

    Proposed DL-TMPFC framework The proposed DL -TMPFC framework comprised stenosis detection phase and TMPFC calculation phase. For stenosis detection network, the training set contained opacified frames with manually annotated bounding boxes. For territory-aware segmentation network, the training set contained opacified frames with manual labels. Manual ann...

    work page

  3. [3]

    Cardiologists selected contrast -opacified frames from angiograms

    Stenosis detection network development The stenosis detection network development comprised three steps (shown in Figure 2): manual annotation, network training, and performance validation. Cardiologists selected contrast -opacified frames from angiograms. Two experts annotated stenosis bounding boxes using LabelImg7, focusing on critical segments, such a...

    work page

  4. [4]

    DL-driven TMPFC measurement The DL-driven TMPFC measurement method comprised two key components: development of a coronary multi - class territory-aware segmentation network and TMPFC measurement. 4.1 Coronary multi-class territory-aware segmentation network The segmentation network development comprised three steps (Figure 3(a)): manual annotation, model...

    work page

  5. [5]

    5.1 Technical validation: agreement versus manual TMPFC The sequences from 30 test cases of group B were used for this analysis

    Comprehensive performance validation To rigorously evaluate the clinical utility of the proposed DL -TMPFC framework, we designed a comprehensive validation study comprising three core components: 1) technical validation against expert manual measurements, 2) clinical validation against established patient phenotypes, 3) incremental value assessment. 5.1 ...

    work page

  6. [6]

    Quantitative evaluation revealed outstanding detection accuracy, with mAP50 reaching 0.991, indicating near-flawless stenosis identification

    Coronary stenosis detection network development The developed coronary stenosis detection model demonstrated exceptional performance in real -time coronary stenosis detection acr oss 196 test cas es. Quantitative evaluation revealed outstanding detection accuracy, with mAP50 reaching 0.991, indicating near-flawless stenosis identification. The mAP50-95 of...

    work page

  7. [7]

    Qualitative assessment across all the test datasets, mean Likert score was 4.5 with 82% rated more than 4

    Coronary multi-class territory-aware segmentation network development The segmentation network achieved exceptional accuracy on the 1 96 hold-out test cases. Qualitative assessment across all the test datasets, mean Likert score was 4.5 with 82% rated more than 4. Qualitative performance across LAD, LCX and RCA revealed mean Likert scores of 4.4 for LAD t...

    work page

  8. [8]

    The median filter window size for 𝐹1, 𝐹2 was 4% of the sequence length, with a maximum of 3 and a maximum of 11 (odd numbers only)

    TMPFC measurement algorithm The automated identification of the initial and final frames for TMPFC calculation relied on adaptive parameters (𝑁1, 𝛿1, 𝑁2, 𝛿2), whose optimal values were empirically derived from the training dataset. The median filter window size for 𝐹1, 𝐹2 was 4% of the sequence length, with a maximum of 3 and a maximum of 11 (odd numbers ...

    work page

Show all 34 references
  1. [9]

    As detailed in Figure 6(a), the Bland -Altman analysis revealed a mean bias of -0.93 frames, with 95% limits of agreement (LoA) ranging from -5.33 to +3.47 frames

    Comprehensive performance validation 4.1 Technical validation: agreement versus manual TMPFC The proposed DL -TMPFC framework demonstrated excellent agreement with expert manual measurements. As detailed in Figure 6(a), the Bland -Altman analysis revealed a mean bias of -0.93 ...

  2. [10]

    Consent 14 Written informed consent was obtained from all individual participants included in the study

  3. [11]

    Data availability statement The datasets and software code are available in the github repository, [https://github.com/DrThink - ai/DL-TMPFC]

  4. [12]

    Conflict of interest None declared

  5. [13]

    YQH: Data curation, Software, Formal analysis, Resources

    Author contributions SL: Data curation, Investigation, Methodology, Software, Validation, Visua lization, Writing -original draft. YQH: Data curation, Software, Formal analysis, Resources. CKH, XGG, HCT: Conceptualization, Data curation, Formal analysis, Project administration...

  6. [14]

    G., Berry C., Escaned J., Maas A

    Kunadian V ., Chieffo A., Camici P . G., Berry C., Escaned J., Maas A. H. E. M., et al., An EAPCI Expert Consensus Document on Ischaemia with Non-Obstructive Coronary Arteries in Collaboration with European Society of Cardiology Working Group on Coronary Pathophysiology & Micr...

    2020

  7. [15]

    A., FFR, iFR, CFR, and IMR: Results from clinical trials., Cardiovasc Revasc Med, 2025, 71: 16-21

    Yang C., Wong C., Teradaa K., Tremmel J. A., FFR, iFR, CFR, and IMR: Results from clinical trials., Cardiovasc Revasc Med, 2025, 71: 16-21

    2025

  8. [16]

    P., et al., TIMI Myocardial Perfusion Frame Count: A New Method to Assess Myocardial Perfusion and Its Predictive Value for Short-Term Prognosis

    Ding S., Pu J., Qiao Z.Q., Shan P., Song W., Du Y . P., et al., TIMI Myocardial Perfusion Frame Count: A New Method to Assess Myocardial Perfusion and Its Predictive Value for Short-Term Prognosis. Catheter Cardiovasc Interv, 2010, 75(5): 722-732

    2010

  9. [17]

    Ge H., Pu J., Ding S., Li Zheng, Xu J. R., He B., V alue of TIMI Myocardial Perfusion Frame Count for Prediction of Microvascular Dysfunction in Reperfused STEMI Patients, View of Cardiac Magnetic Resonance Early After Reperfusion. Am J Cardiol, 2013, 111(7): 40B

    2013

  10. [18]

    D., Ostovaneh M

    Cong C., Kato Y ., Vasconcellos H. D., Ostovaneh M. R., Lima J. A. C., Ambale-Venkatesh B., Deep learning-based end-to-end automated stenosis classification and localization on catheter coronary angiography. Front Cardiovasc Med, 2023, 10: 944135

    2023

  11. [19]

    BMC Cardiovasc Disord, 2023, 23, 407

    Farhad A., Reza R., Azamossadat H., Ali G., Arash R., Mehrad A., et al., Artificial intelligence in estimating fractional flow re- serve: a systematic literature review of techniques. BMC Cardiovasc Disord, 2023, 23, 407

    2023

  12. [20]

    Github, 2015

    Tzutalin, LabelImg (Version 1.8.6). Github, 2015. URL: https://github.com/HumanSignal/labelImg

    2015

  13. [21]

    L., Interrater reliability: the kappa statistic., Biochem Med (Zagreb), 2012, 22(3): 276-82

    McHugh M. L., Interrater reliability: the kappa statistic., Biochem Med (Zagreb), 2012, 22(3): 276-82

    2012

  14. [22]

    J., Ye Q

    Tian Y . J., Ye Q. X., Doermann D., YOLOv12: Attention-Centric Real-Time Object Detectors [preprint]. arXiv.org., 2025, URL: https://doi.org/10.48550/arXiv.2502.12524

    2025 doi arXiv

  15. [23]

    Int J Comput Vis, 2010, 88: 303-338

    Everingham M., Van Gool L., Williams C.K.I., Winn J., Zisserman A., The Pascal Visual Object Classes (VOC) Challenge. Int J Comput Vis, 2010, 88: 303-338

    2010

  16. [24]

    H., Yang T

    Sze V ., Chen Y . H., Yang T. J., Emer J. S., Efficient Processing of Deep Neural Networks: A Tutorial and Survey. Proc. IEEE, 2017, 105(12): 2295-2329

    2017

  17. [25]

    A., Gao Y., Gerig G., ITK-SNAP: An interactive tool for semi -automatic segmentation of multi -modality biomedical images

    Y ushkevich P. A., Gao Y., Gerig G., ITK-SNAP: An interactive tool for semi -automatic segmentation of multi -modality biomedical images. EMBC, 2016, 3342-3345

    2016

  18. [26]

    H., Zhou M., Liu D., Yan Z

    Gao Y . H., Zhou M., Liu D., Yan Z. N. , Zhang S. T., Metaxas D. N. , A Data -scalable Transformer for Medical Image Segmentation: Architecture, Model Efficiency, and Benchmark [preprint]. arXiv.org., 2023, URL: https://arxiv.org/pdf/2203. 00131. 15

    2023

  19. [27]

    Archives of Psychology, 1932, 140: 1-55

    Likert R., A Technique for the Measurement of Attitudes. Archives of Psychology, 1932, 140: 1-55

    1932

  20. [28]

    Müller D., Soto -Rey I., Kramer F., Towards a guideline for evaluation metrics in medical image segmentation., BMC Res Notes, 2022, 15: 210

    2022

  21. [29]

    M., Altman D

    Bland J. M., Altman D. G., Statistical methods for assessing agreement between two methods of clinical measurement. Lancet, 1986, 1(8476): 307-10

    1986

  22. [30]

    Madrigal L., Statistics for Anthropology, 2nd ed., Cambridge University Press, 2012

    2012

  23. [31]

    H., Knatterud G., Roberts R., Border J., Cohen L

    Chesebro J. H., Knatterud G., Roberts R., Border J., Cohen L. S., Dalen J., et al., Thrombolysis in Myocardial Infarction (TIMI) Trial, Phase I: A comparison between intravenous tissue plasminogen activator and intravenous streptokinase. Clinical findings through hospital disc...

    1987

  24. [32]

    M., Cannon C

    Gibson C. M., Cannon C. P ., Daley W. L., Dodge J. T., Alexander B., Marble S. J., et al., TIMI frame count: a quantitative method of assessing coronary artery flow. Circulation, 1996, 93(5): 879-88

    1996

  25. [33]

    R., Use and misuse of the receiver operating characteristic curve in risk prediction

    Cook N. R., Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation, 2007, 115(7): 928 - 35

    2007

  26. [34]

    A., Nonparametric Statistical Methods., 3rd ed., Wiley, 2013

    Hollander M., Wolfe D. A., Nonparametric Statistical Methods., 3rd ed., Wiley, 2013. Figure legends Figure 1. Schematic of the proposed DL-TMPFC framework. Figure 2. Detailed description of the development of coronary stenosis detection network model. Figure 3. Detailed descri...

    2013

Pith tools

Reviewed June 30, 2026 · model on record in the stance chip above.