pith. machine review for the scientific record. sign in

arxiv: 2604.26147 · v1 · submitted 2026-04-28 · 💻 cs.CV · cs.AI

Recognition: unknown

A Data-Centric Framework for Intraoperative Fluorescence Lifetime Imaging for Glioma Surgical Guidance

Silvia Noble Anbunesan , Mohamed Abul Hassan , Jinyi Qi , Lisanne Kraft , Han Sung Lee , Orin Bloch , Laura Marcu
Authors on Pith no claims yet

Pith reviewed 2026-05-07 16:39 UTC · model grok-4.3

classification 💻 cs.CV cs.AI
keywords fluorescence lifetime imaginggliomadata-centric AIconfident learningintraoperative imagingglioblastomasurgical guidancelabel refinement
0
0 comments X

The pith

A data-centric framework uses confident learning to merge seven glioma classes into three for 96 percent FLIm classification accuracy.

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

The paper establishes that integrating confident learning with class refinement on intraoperative fluorescence lifetime imaging data can systematically identify and correct label inconsistencies from an initial seven-class neuropathologist scheme. This produces a high-fidelity three-class dataset that trains a model reaching 96 percent accuracy while revealing distinct optical signatures across low, moderate, and high tumor cellularity. A sympathetic reader would care because reliable real-time biochemical contrast during glioblastoma resection could help surgeons maximize tumor removal without damaging functional brain tissue. The approach also shows that targeted evaluation of low-confidence points highlights biological factors like gray matter composition and acquisition issues like blood contamination. Blinded re-evaluation of flagged margins confirms intra-pathologist variability, supporting selective rather than exhaustive relabeling.

Core claim

The authors demonstrate that their data-centric AI framework, by applying confident learning to quantify point-level confidence in FLIm signals from 192 tissue margins across 31 GBM patients, identifies inconsistencies in the initial seven-class labeling and guides iterative merging into a three-class scheme of low, moderate, and high infiltration. The resulting dataset trains a classifier that achieves 96 percent accuracy, with SHAP analysis exposing class-specific feature importance and targeted analysis linking low-confidence predictions to biological and technical variables. Selective re-evaluation of CL-flagged margins reveals labeling variability, showing the framework improves data可靠性

What carries the argument

Confident learning (CL) to quantify FLIm point-level confidence, flag inconsistencies, and direct iterative class merging from seven to three tumor cellularity categories.

If this is right

  • The refined three-class model supports real-time intraoperative decisions that balance tumor resection with tissue preservation.
  • SHAP-derived optical signatures provide a biological basis for interpreting FLIm contrast across the infiltration spectrum.
  • Targeted analysis of low-confidence predictions identifies specific confounders like blood contamination that can be mitigated in future acquisitions.
  • Selective relabeling based on CL reduces the need for exhaustive expert review while maintaining high model performance.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same CL-driven refinement process could be tested on other label-free optical modalities used in neurosurgery to check for similar gains in label fidelity.
  • If the three-class scheme generalizes to new patients, it might enable deployment of FLIm probes with built-in uncertainty estimates for surgical navigation systems.
  • The observed pathologist variability suggests that future studies should collect multi-rater labels from the start to quantify the upper bound on achievable accuracy.

Load-bearing premise

The single expert neuropathologist's initial seven-class labeling is consistent enough for confident learning to detect and fix errors without introducing new systematic biases.

What would settle it

Independent multi-pathologist re-labeling of the CL-flagged margins showing no greater inconsistency rate than unflagged margins would falsify the claim that CL improves label reliability.

read the original abstract

Accurate intraoperative assessment of glioma infiltration is essential for maximizing tumor resection while preserving functional brain tissue. Fluorescence lifetime imaging (FLIm) offers real-time, label-free biochemical contrast, but its clinical utility is challenged by biological heterogeneity, class imbalance, and variability in histopathological labeling. We present a data-centric AI (DC-AI) framework that integrates confident learning (CL), class refinement, and targeted label evaluation to develop a robust multi-class FLIm classifier for glioblastoma (GBM) resection margins. FLIm data were collected from 192 tissue margins across 31 newly diagnosed IDH-wildtype GBM patients and initially labeled into seven tumor cellularity classes by an expert neuropathologist. CL was applied to quantify FLIm point-level confidence, identify label inconsistencies, and guide iterative class merging into a three-class scheme ("low", "moderate", "high"). The resulting high-fidelity dataset enabled training a model that achieved 96% accuracy in the three-class task. SHAP analysis revealed class-specific FLIm feature importance, highlighting distinct optical signatures across the infiltration spectrum. Targeted FLIm analysis further identified biological (e.g., gray matter composition) and acquisition-related (e.g., blood contamination) contributors to low-confidence predictions. Blinded re-evaluation of margins flagged by CL demonstrated intra-pathologist variability, underscoring the value of selective relabeling rather than exhaustive review. Together, these findings demonstrate that a DC-AI framework can systematically improve data reliability, enhance model robustness, and refine biological interpretation of FLIm signals, supporting the development of clinically actionable optical tools for real-time glioma margin assessment.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a data-centric AI framework for intraoperative fluorescence lifetime imaging (FLIm) to guide glioma resection. It collects FLIm data from 192 margins in 31 GBM patients, obtains initial seven-class cellularity labels from a single expert neuropathologist, applies confident learning (CL) to detect inconsistencies and iteratively merge into a three-class scheme (low/moderate/high), trains a classifier that reaches 96% accuracy on the refined labels, uses SHAP to interpret feature importance, and performs targeted analysis plus blinded re-evaluation of low-confidence cases.

Significance. If the 96% accuracy reflects genuine discriminative power on biologically faithful labels rather than artifacts of the CL-driven merging process, the work could meaningfully advance label-free optical tools for real-time margin assessment in neurosurgery. The data-centric emphasis on systematic label refinement, combined with SHAP-based biological interpretation and selective re-evaluation, offers a practical template for handling noisy, imbalanced medical imaging data; this is especially relevant given the heterogeneity of glioma infiltration.

major comments (2)
  1. [Abstract] Abstract: the headline claim of 96% accuracy on the three-class task supplies no validation details (train/test split, cross-validation scheme, baseline models, or error analysis). Because CL uses model predictions both to flag inconsistencies and to guide class merging, the accuracy figure risks circularity unless the test partition was strictly held out from the entire CL procedure; performance against the original seven-class labels should also be reported to demonstrate that merging preserved rather than erased discriminative signal.
  2. [Abstract] Abstract (framework description): the initial seven-class labels come from a single expert with no reported inter-rater reliability or independent ground-truth validation. The subsequent CL-guided merging into three classes, followed by training and 96% accuracy, therefore rests on the untested assumption that the refined labels remain faithful to underlying biology; the blinded re-evaluation is limited to CL-flagged margins and does not quantify how much the merging step alters the infiltration spectrum.
minor comments (2)
  1. [Abstract] Abstract: the phrase 'targeted FLIm analysis further identified biological (e.g., gray matter composition) and acquisition-related (e.g., blood contamination) contributors' is stated without naming the exact statistical or visualization methods used to link these factors to low-confidence predictions.
  2. [Abstract] Abstract: SHAP analysis is invoked to reveal 'class-specific FLIm feature importance' but the manuscript does not indicate which FLIm parameters (lifetime, intensity, etc.) were input to the model or how many features were retained after any preprocessing.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for their careful reading and valuable comments, which have helped us improve the clarity and rigor of our manuscript. Below, we provide point-by-point responses to the major comments.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the headline claim of 96% accuracy on the three-class task supplies no validation details (train/test split, cross-validation scheme, baseline models, or error analysis). Because CL uses model predictions both to flag inconsistencies and to guide class merging, the accuracy figure risks circularity unless the test partition was strictly held out from the entire CL procedure; performance against the original seven-class labels should also be reported to demonstrate that merging preserved rather than erased discriminative signal.

    Authors: We agree that additional details on the validation procedure should be included in the abstract for completeness. In the revised manuscript, we will expand the abstract to describe the cross-validation scheme, train/test split, and comparison to baseline models. To address the risk of circularity, we clarify that while CL was used to refine labels on the dataset, the performance evaluation was conducted using a held-out test set that was not involved in the label refinement process. We will also report the classification performance on the original seven-class labels to show that the merging step retained important discriminative information rather than erasing it. These changes will be reflected in both the abstract and the detailed methods section. revision: yes

  2. Referee: [Abstract] Abstract (framework description): the initial seven-class labels come from a single expert with no reported inter-rater reliability or independent ground-truth validation. The subsequent CL-guided merging into three classes, followed by training and 96% accuracy, therefore rests on the untested assumption that the refined labels remain faithful to underlying biology; the blinded re-evaluation is limited to CL-flagged margins and does not quantify how much the merging step alters the infiltration spectrum.

    Authors: We recognize that relying on labels from a single expert neuropathologist without inter-rater reliability or independent validation is a limitation. The CL framework is specifically designed to handle label noise in such scenarios by identifying and correcting inconsistencies. The blinded re-evaluation was intentionally focused on the low-confidence cases identified by CL to efficiently assess variability, and it did demonstrate intra-pathologist variability. However, we agree that this approach does not fully quantify the impact of the class merging on the entire spectrum of infiltration. In the revision, we will include an analysis of the label distributions before and after merging and discuss the implications for biological fidelity. We will also highlight this as a limitation in the discussion section. revision: partial

standing simulated objections not resolved
  • We cannot provide inter-rater reliability or independent ground-truth validation data, as the initial labeling was performed by a single expert and no additional raters were involved in the study.

Circularity Check

0 steps flagged

No circularity: empirical label refinement and supervised training remain independent of inputs

full rationale

The paper presents a standard data-centric workflow: expert neuropathologist provides initial seven-class labels on FLIm data, confident learning is applied to flag inconsistencies and guide merging to three classes, and a classifier is then trained and evaluated on the resulting dataset, yielding 96% accuracy. No equations, derivations, fitted parameters renamed as predictions, or self-citations appear in the load-bearing steps. The accuracy metric is computed on the refined labels via conventional supervised training (with mention of blinded re-evaluation), not by construction from the same model predictions used in cleaning. The framework is self-contained against external benchmarks and does not reduce any claimed result to its own inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Abstract supplies no explicit free parameters, new entities, or formal axioms; relies on standard domain assumptions in medical imaging and machine learning.

axioms (2)
  • domain assumption Initial expert histopathological labels serve as a sufficiently accurate starting point despite known variability
    Used as input for confident learning to detect inconsistencies
  • domain assumption FLIm signals contain distinguishable biochemical features across infiltration levels
    Implicit basis for the multi-class classification task

pith-pipeline@v0.9.0 · 5611 in / 1189 out tokens · 40184 ms · 2026-05-07T16:39:55.723844+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

39 extracted references · 35 canonical work pages

  1. [1]

    State -of-the-art imaging for glioma surgery

    N. Verburg, P. C. De, and W. Hamer, “State -of-the-art imaging for glioma surgery”, doi: 10.1007/s10143-020-01337-9/Published

    work page doi:10.1007/s10143-020-01337-9/published

  2. [2]

    Mesoscopic fluorescence lifetime imaging: Fundamental principles, clinical applications and future directions,

    A. Alfonso-Garcia et al., “Mesoscopic fluorescence lifetime imaging: Fundamental principles, clinical applications and future directions,” J. Biophotonics , vol. 14, no. 6, Jun. 2021, doi: 10.1002/jbio.202000472

    work page doi:10.1002/jbio.202000472 2021

  3. [3]

    J. R. Lakowicz, Principles of fluorescence spectroscopy , Third. Baltimore: Springer New York LLC, 2006. doi: 10.1007/978-0-387- 46312-4

    work page doi:10.1007/978-0-387- 2006

  4. [4]

    Recent innovations in fluorescence lifetime imaging microscopy for biology and medicine,

    R. Datta, A. Gillette, M. Stefely, and M. C. Skala, “Recent innovations in fluorescence lifetime imaging microscopy for biology and medicine,” J. Biomed. Opt. , vol. 26, no. 07, Jul. 2021, doi: 10.1117/1.jbo.26.7.070603

    work page doi:10.1117/1.jbo.26.7.070603 2021

  5. [5]

    Augmenting Neuronavigation with Label-Free Fluorescence Lifetime Imaging for Precise Detection of Glioma Infiltration,

    S. Noble Anbunesan et al. , “Augmenting Neuronavigation with Label-Free Fluorescence Lifetime Imaging for Precise Detection of Glioma Infiltration,” 2025. 10

    2025

  6. [6]

    Fluorescence lifetime measurements and biological imaging,

    M. Y. Berezin and S. Achilefu, “Fluorescence lifetime measurements and biological imaging,” Chem. Rev., vol. 110, no. 5, pp. 2641 –84, May 2010, doi: 10.1021/cr900343z

    work page doi:10.1021/cr900343z 2010

  7. [7]

    FLImBrush: dynamic visualization of intraoperative free-hand fiber-based fluorescence lifetime imaging,

    M. Marsden et al. , “FLImBrush: dynamic visualization of intraoperative free-hand fiber-based fluorescence lifetime imaging,” Biomed. Opt. Express , vol. 11, no. 9, p. 5166, Sep. 2020, doi: 10.1364/boe.398357

    work page doi:10.1364/boe.398357 2020

  8. [8]

    Interobserver variation of the histopathological diagnosis in clinical trials on glioma: A clinician’s perspective,

    M. J. Van Den Bent, “Interobserver variation of the histopathological diagnosis in clinical trials on glioma: A clinician’s perspective,” Sep

  9. [9]

    doi: 10.1007/s00401-010-0725-7

    work page doi:10.1007/s00401-010-0725-7

  10. [10]

    Tumor heterogeneity: Mechanisms and bases for a reliable application of molecular marker design,

    S. J. Diaz-Cano, “Tumor heterogeneity: Mechanisms and bases for a reliable application of molecular marker design,” Feb. 2012. doi: 10.3390/ijms13021951

    work page doi:10.3390/ijms13021951 2012

  11. [11]

    A Data -Centric Approach to improve performance of deep learning models,

    N. Bhatt, N. Bhatt, P. Prajapati, V. Sorathiya, S. Alshathri, and W. El-Shafai, “A Data -Centric Approach to improve performance of deep learning models,” Sci. Rep., vol. 14, no. 1, p. 22329, Dec. 2024, doi: 10.1038/s41598-024-73643-x

    work page doi:10.1038/s41598-024-73643-x 2024

  12. [12]

    Deep learning from multiple experts improves identification of amyloid neuropathologies,

    D. R. Wong et al., “Deep learning from multiple experts improves identification of amyloid neuropathologies,” Acta Neuropathol. Commun., vol. 10, no. 1, Dec. 2022, doi: 10.1186/s40478-022-01365- 0

    work page doi:10.1186/s40478-022-01365- 2022

  13. [13]

    A survey of label-noise deep learning for medical image analysis,

    J. Shi, K. Zhang, C. Guo, Y. Yang, Y. Xu, and J. Wu, “A survey of label-noise deep learning for medical image analysis,” Jul. 01, 2024, Elsevier B.V. doi: 10.1016/j.media.2024.103166

    work page doi:10.1016/j.media.2024.103166 2024

  14. [14]

    Deep learning with noisy labels: Exploring techniques and remedies in medical image analysis,

    D. Karimi, H. Dou, S. K. Warfield, and A. Gholipour, “Deep learning with noisy labels: Exploring techniques and remedies in medical image analysis,” Med. Image Anal. , vol. 65, Oct. 2020, doi: 10.1016/j.media.2020.101759

    work page doi:10.1016/j.media.2020.101759 2020

  15. [15]

    Confident Learning: Estimating Uncertainty in Dataset Labels,

    C. G. Northcutt, L. Jiang, and I. L. Chuang, “Confident Learning: Estimating Uncertainty in Dataset Labels,” Oct. 2019, [Online]. Available: http://arxiv.org/abs/1911.00068

    work page arXiv 2019

  16. [16]

    Pulse - sampling fluorescence lifetime imaging: evaluation of photon economy,

    X. Zhou, J. Bec, K. Ehrlich, A. A. Garcia, and L. Marcu, “Pulse - sampling fluorescence lifetime imaging: evaluation of photon economy,” Opt. Lett. , vol. 48, no. 17, p. 4578, Sep. 2023, doi: 10.1364/ol.490096

    work page doi:10.1364/ol.490096 2023

  17. [17]

    Multispectral fluorescence lifetime imaging device with a silicon avalanche photodetector,

    X. Zhou, J. Bec, D. Yankelevich, and L. Marcu, “Multispectral fluorescence lifetime imaging device with a silicon avalanche photodetector,” Opt. Express, vol. 29, no. 13, p. 20105, Jun. 2021, doi: 10.1364/oe.425632

    work page doi:10.1364/oe.425632 2021

  18. [18]

    In vivo characterization of the human glioblastoma infiltrative edge with label -free intraoperative fluorescence lifetime imaging,

    A. Alfonso -Garcia et al. , “In vivo characterization of the human glioblastoma infiltrative edge with label -free intraoperative fluorescence lifetime imaging,” Biomed. Opt. Express, vol. 14, no. 5, p. 2196, May 2023, doi: 10.1364/boe.481304

    work page doi:10.1364/boe.481304 2023

  19. [19]

    Intraoperative detection of IDH-mutant glioma using fluorescence lifetime imaging,

    S. Noble Anbunesan et al., “Intraoperative detection of IDH-mutant glioma using fluorescence lifetime imaging,” J. Biophotonics, vol. 16, no. 4, Dec. 2022, doi: 10.1002/jbio.202200291

    work page doi:10.1002/jbio.202200291 2022

  20. [20]

    Halos and related structures.Physica Scripta2013,2013, 014001

    J. Liu, Y. Sun, J. Qi, and L. Marcu, “A novel method for fast and robust estimation of fluorescence decay dynamics using constrained least-squares deconvolution with Laguerre expansion,” Phys. Med. Biol., vol. 57, no. 4, pp. 843 –865, Feb. 2012, doi: 10.1088/0031 - 9155/57/4/843

    work page doi:10.1088/0031 2012

  21. [21]

    D., Methods Appl Fluoresc 2016, 4 (4), 042001

    F. Fereidouni, D. Gorpas, D. Ma, H. Fatakdawala, and L. Marcu, “Rapid fluorescence lifetime estimation with modified phasor approach and Laguerre deconvolution: a comparative study,” Methods Appl. Fluoresc. , vol. 5, no. 3, 2017, doi: 10.1088/2050 - 6120/aa7b62

    work page doi:10.1088/2050 2017

  22. [22]

    Ser: Graduate Texts in Mathematics (GTM), vol

    Z. Xu et al. , “Denoising for Relaxing: Unsupervised Domain Adaptive Fundus Image Segmentation Without Source Data,” in Medical Image Computing and Computer Assisted Intervention – MICCAI 2022, 2022, pp. 214 –224. doi: 10.1007/978 -3-031-16443- 9_21

    work page doi:10.1007/978 2022

  23. [23]

    Unsupervised domain adaptation for histopathology image segmentation with incomplete labels,

    H. Zhou et al., “Unsupervised domain adaptation for histopathology image segmentation with incomplete labels,” Comput. Biol. Med. , vol. 171, Mar. 2024, doi: 10.1016/j.compbiomed.2024.108226

    work page doi:10.1016/j.compbiomed.2024.108226 2024

  24. [24]

    Converting tabular data into images for deep learning with convolutional neural networks,

    Y. Zhu et al., “Converting tabular data into images for deep learning with convolutional neural networks,” Sci. Rep., vol. 11, no. 1, Dec. 2021, doi: 10.1038/s41598-021-90923-y

    work page doi:10.1038/s41598-021-90923-y 2021

  25. [25]

    A closer look at deep learning methods on tabular datasets, 2025

    H.-J. Ye, S.-Y. Liu, H.-R. Cai, Q.-L. Zhou, and D.-C. Zhan, “A Closer Look at Deep Learning Methods on Tabular Datasets,” Nov. 2025, [Online]. Available: http://arxiv.org/abs/2407.00956

    work page arXiv 2025

  26. [26]

    Tabular Data: Deep Learning is Not All You Need,

    R. Shwartz-Ziv and A. Armon, “Tabular Data: Deep Learning is Not All You Need,” Nov. 2021, [Online]. Available: http://arxiv.org/abs/2106.03253

    work page arXiv 2021

  27. [27]

    Intraoperative Margin Assessment in Oral and Oropharyngeal Cancer Using Label -Free Fluorescence Lifetime Imaging and Machine Learning,

    M. Marsden et al., “Intraoperative Margin Assessment in Oral and Oropharyngeal Cancer Using Label -Free Fluorescence Lifetime Imaging and Machine Learning,” IEEE Trans. Biomed. Eng., vol. 68, no. 3, pp. 857–868, Mar. 2021, doi: 10.1109/TBME.2020.3010480

    work page doi:10.1109/tbme.2020.3010480 2021

  28. [28]

    The Intraoperative Utility of Raman Spectroscopy for Neurosurgical Oncology,

    J.-S. Chen, J. Y. Oh, T. C. Hollon, S. L. Hervey-Jumper, J. S. Young, and M. S. Berger, “The Intraoperative Utility of Raman Spectroscopy for Neurosurgical Oncology,” Cancers (Basel)., vol. 17, no. 24, p. 3920, Dec. 2025, doi: 10.3390/cancers17243920

    work page doi:10.3390/cancers17243920 2025

  29. [29]

    OCT -Guided Surgery for Gliomas: Current Concept and Future Perspectives,

    K. Yashin et al. , “OCT -Guided Surgery for Gliomas: Current Concept and Future Perspectives,” Feb. 01, 2022, Multidisciplinary Digital Publishing Institute (MDPI) . doi: 10.3390/diagnostics12020335

    work page doi:10.3390/diagnostics12020335 2022

  30. [30]

    Machine Learning: Algorithms, Real -World Applications and Research Directions,

    I. H. Sarker, “Machine Learning: Algorithms, Real -World Applications and Research Directions,” SN Comput. Sci., vol. 2, no. 3, May 2021, doi: 10.1007/s42979-021-00592-x

    work page doi:10.1007/s42979-021-00592-x 2021

  31. [31]

    Label -Free Macroscopic Fluorescence Lifetime Imaging of Brain Tumors,

    M. Lukina et al. , “Label -Free Macroscopic Fluorescence Lifetime Imaging of Brain Tumors,” Front. Oncol., vol. 11, May 2021, doi: 10.3389/fonc.2021.666059

    work page doi:10.3389/fonc.2021.666059 2021

  32. [32]

    Label -free metabolic and structural profiling of dynamic biological samples using multimodal optical microscopy with sensorless adaptive optics,

    R. R. Iyer et al. , “Label -free metabolic and structural profiling of dynamic biological samples using multimodal optical microscopy with sensorless adaptive optics,” Sci. Rep., vol. 12, no. 1, p. 3438, Mar. 2022, doi: 10.1038/s41598-022-06926-w

    work page doi:10.1038/s41598-022-06926-w 2022

  33. [33]

    Fluorescence Lifetime Spectroscopy of Glioblastoma Multiforme,

    L. Marcu et al. , “Fluorescence Lifetime Spectroscopy of Glioblastoma Multiforme,” Photochem. Photobiol. , vol. 80, p. 3, 2004

    2004

  34. [34]

    Distinction of brain tissue, low grade and high grade glioma with time -resolved fluorescence spectroscopy,

    W. H. Yong et al., “Distinction of brain tissue, low grade and high grade glioma with time -resolved fluorescence spectroscopy,” Frontiers in Bioscience, vol. 11, no. 2 P.1199-1590, pp. 1255–1263, 2006, doi: 10.2741/1878

    work page doi:10.2741/1878 2006

  35. [35]

    Real‐time augmented reality for delineation of surgical margins during neurosurgery using autofluorescence lifetime contrast,

    A. Alfonso‐Garcia et al. , “Real‐time augmented reality for delineation of surgical margins during neurosurgery using autofluorescence lifetime contrast,” J. Biophotonics, vol. 13, no. 1, Jan. 2020, doi: 10.1002/jbio.201900108

    work page doi:10.1002/jbio.201900108 2020

  36. [36]

    Fast model -free deconvolution of fluorescence decay for analysis of biological systems,

    J. A. Jo, Q. Fang, T. Papaioannou, and L. Marcu, “Fast model -free deconvolution of fluorescence decay for analysis of biological systems,” J. Biomed. Opt. , vol. 9, no. 4, p. 743, 2004, doi: 10.1117/1.1752919

    work page doi:10.1117/1.1752919 2004

  37. [37]

    Understanding the Warburg Effect: The Metabolic Requirements of Cell Proliferation,

    M. G. Vander Heiden, L. C. Cantley, and C. B. Thompson, “Understanding the Warburg Effect: The Metabolic Requirements of Cell Proliferation,” Science (1979)., vol. 324, no. 5930, pp. 1029 – 1033, May 2009, [Online]. Available: https://www.science.org

    1979

  38. [38]

    Metabolomic differentiation of tumor core versus edge in glioma,

    M. E. Baxter, H. A. Miller, J. Chen, B. J. Williams, and H. B. Frieboes, “Metabolomic differentiation of tumor core versus edge in glioma,” Neurosurg. Focus , vol. 54, no. 6, 2023, doi: 10.3171/2023.3.FOCUS2379

    work page doi:10.3171/2023.3.focus2379 2023

  39. [39]

    Fluorescence lifetime imaging microscopy: fundamentals and advances in instrumentation, analysis, and applications,

    R. Datta, T. M. Heaster, J. T. Sharick, A. A. Gillette, and M. C. Skala, “Fluorescence lifetime imaging microscopy: fundamentals and advances in instrumentation, analysis, and applications,” J. Biomed. Opt., vol. 25, no. 07, p. 1, May 2020, doi: 10.1117/1.jbo.25.7.071203

    work page doi:10.1117/1.jbo.25.7.071203 2020