Методы машинного обучения для анализа спектрально-разрешенных изображений в нейроонкологии

При проведении хирургических операций по удалению опухолей мозга критически важной для снижения частоты рецидивов является полнота удаления всех пораженных участков мозга без нарушения функциональности жизненно важных органов. Поэтому дифференциальная диагностика микроучастков опухолевой ткани с последующим их удалением или деструкцией является актуальнойзадачей, определяющей успех операции в целом. Оптическая спектроскопия за последние десятилетие показала свои преимущества при использовании в качестве инструмента интраоперационной метаболической навигации. И одним из наиболее многообещающих вариантов развития этой технологии является спектрально-разрешенная визуализация. В настоящий момент разработаны методики как спектрально-разрешенной визуализации в диффузно-отраженном свете, позволяющие, например, картировать распределение сатурации гемоглобина кислородом в зоне интереса, так и системы визуализации флуоресценции, как эндогенной, так и индуцированной введением в организм пациента флуоресцентных маркеров. Эти системы обеспечивают быстрый анализ тканей по составу исследуемых хромофоров и флуорофоров, позволяя нейрохирургу во время операции дифференцировать опухолевые и нормальные ткани, а также функционально значимые зоны. Не менее важным направлением применения спектрально-разрешенной визуализации являются методы, основанные на картировании характеристик, получаемых из спектров комбинационного рассеяния, однако, в силу меньшего сечения процесса эти методики используются ex vivo, как правило, для срочного анализа только что удаленных образцов тканей. В настоящей работе мы сделаем фокус как на физических основаниях таких методов, так и на весьма важном аспекте их применения – методах машинного обучения для обработки таких изображений и классификации тканей.

1. Aboras M., Amasha H., Ibraheem I. Early detection of melanoma using multispectral imaging and artifcial intelligence techniques // American Journal of Biomedical and Life Sciences. – 2015. – Vol. 3. – № 2–3. – P. 29–33. doi: 10.11648/j.ajbls.s.2015030203.16.

2. Kotwal A., Saragadam V., Bernstock J. D. et al. Hyperspectral imaging in neurosurgery: a review of systems, computational methods, and clinical applications // Journal of Biomedical Optics. – 2024. – Vol. 30. – № 02. doi: 10.1117/1.JBO.30.2.023512.

3. Stummer W., Stocker S., Wagner S. et al. Intraoperative Detection of Malignant Gliomas by 5-Aminolevulinic Acid-induced Porphyrin Fluorescence // Neurosurgery. – 1998. – Vol. 42. – № 3. – P. 518–526. doi: 10.1097/00006123-199803000-00017.

4. Traylor J. I., Pernik M. N., Sternisha A. C. et al. Molecular and Metabolic Mechanisms Underlying Selective 5-Aminolevulinic Acid-Induced Fluorescence in Gliomas // Cancers. – 2021. – Vol. 13. – № 3. – P. 580. doi: 10.3390/cancers13030580.

5. Ivanova-Radkevich V. I., Kuznetsova O. M., Filonenko E. V. The role of membrane transport proteins in 5-ALA-induced accumulation of protoporphyrin iX in tumor cells // Biomedical Photonics. – 2024. – Vol. 13. – № 2. – P. 43–48. doi: 10.24931/2413-9432-2024-13-2-43-48.

6. Nasir-Moin M., Wadiura L. I., Sacalean V. et al. Localization of protoporphyrin IX during glioma-resection surgery via paired stimulated Raman histology and fuorescence microscopy // Nature Biomedical Engineering. – 2024. – Vol. 8. – № 6. – P. 672–688. doi: 10.1038/s41551-024-01217-3.

7. Matsumura H., Akimoto J., Haraoka J. et al. Uptake and retention of the photosensitizer mono-l-asparthyl chlorine e6 in experimental malignant glioma // Lasers in Medical Science. – 2008. – Vol. 23. – № 3. – P. 237–245. doi: 10.1007/s10103-007-0469-3.

8. Valdés P. A., Jacobs V., Harris B. T. et al. Quantitative fuorescence using 5-aminolevulinic acid-induced protoporphyrin IX biomarker as a surgical adjunct in low-grade glioma surgery // Journal of Neurosurgery. – 2015. – Vol. 123. – № 3. – P. 771–780. doi: 10.3171/2014.12.JNS14391.

9. Rynda A. Yu., Olyushin V. E., Rostovtsev D. M. et al. Results of microsurgical resection of glioblastomas under endoscopic and fuorescent control // Biomedical Photonics. – 2024. – Vol. 13. – № 3. – P. 20–30. doi: 10.24931/2413-9432-2024-13-3-20-30.

10. Горяйнов С. А., Потапов А. А., Лощенов В. Б. и др. Флуоресценция в нейрохирургии / С. А. Горяйнов, А. А. Потапов, В. Б. Лощенов и др., под ред. Д. Ю. Усачева, Москва: ТПС Принт, 2024.

11. Loshchenov M., Zelenkov P., Potapov A. et al. Endoscopic fuorescence visualization of 5-ALA photosensitized central nervous system tumors in the neural tissue transparency spectral range // Photonics & Lasers in Medicine. – 2014. – Vol. 3. – № 2. doi: 10.1515/plm-2013-0017.

12. Savelieva T. A., Loshchenov M. V., Borodkin A. V. et al. Combined spectroscopic and video fuorescent instrument for intraoperative navigation when removing a glial tumor ed. Z. Zalevsky, V. V. Tuchin, W. C. Blondel, Online Only, France: SPIE, 2020. C. 35. doi: 10.1117/12.2556064.

13. Udeneev A. M., Kalyagina N. A., Reps V. F. et al. Photo and spectral fuorescence analysis of the spinal cord injury area in animal models // Biomedical Photonics. – 2023. – Vol. 12. – № 3. – P. 15–20. doi: 10.24931/2413-9432-2023-12-3-16-20.

14. Wainwright J. V., Endo T., Cooper J. B. et al. The role of 5-aminolevulinic acid in spinal tumor surgery: a review // Journal of Neuro-Oncology. – 2019. – Vol. 141. – № 3. – P. 575–584. doi: 10.1007/s11060-018-03080-0.

15. Valdés P. A., Jacobs V. L., Leblond F. et al. Quantitative spectrally resolved intraoperative fuorescence imaging for neurosurgical guidance in brain tumor surgery: pre-clinical and clinical results ed. H. Hirschberg, S. J. Madsen, E. D. Jansen et al., San Francisco, California, United States, 2014. C. 892809. doi: 10.1117/12.2039090.

16. Picart T., Gautheron A., Caredda C. et al. Fluorescence-Guided Surgical Techniques in Adult Difuse Low-Grade Gliomas: State-of-the-Art and Emerging Techniques: A Systematic Review // Cancers. – 2024. – Vol. 16. – № 15. – P. 2698. doi: 10.3390/cancers16152698.

17. Maragkou T., Quint K., Pollo B. et al. Intraoperative confocal laser endomicroscopy for brain tumors - potential and challenges from a neuropathological perspective // Free Neuropathology. – 2022. – P. 24 Pages. doi: 10.17879/FREENEUROPATHOLOGY-2022-4369.

18. Sankar T., Delaney P. M., Ryan R. W. et al. Miniaturized Handheld Confocal Microscopy for Neurosurgery: Results in an Experimental Glioblastoma Model // Neurosurgery. – 2010. – Vol. 66. – № 2. – P. 410–418. doi: 10.1227/01.NEU.0000365772.66324.6F.

19. Foersch S., Heimann A., Ayyad A. et al. Confocal Laser Endomicroscopy for Diagnosis and Histomorphologic Imaging of Brain Tumors In Vivo // PLoS ONE. – 2012. – Vol. 7. – № 7. – P. e41760. doi: 10.1371/journal.pone.0041760.

20. Martirosyan N. L., Cavalcanti D. D., Eschbacher J. M. et al. Use of in vivo near-infrared laser confocal endomicroscopy with indocyanine green to detect the boundary of infltrative tumor: Laboratory investigation // Journal of Neurosurgery. – 2011. – Vol. 115. – № 6. – P. 1131–1138. doi: 10.3171/2011.8.JNS11559.

21. Sanai N., Eschbacher J., Hattendorf G. et al. Intraoperative Confocal Microscopy for Brain Tumors: A Feasibility Analysis in Humans // Operative Neurosurgery. – 2011. – Vol. 68. – P. ons282–ons290. doi: 10.1227/NEU.0b013e318212464e.

22. Sanai N., Snyder L. A., Honea N. J. et al. Intraoperative confocal microscopy in the visualization of 5-aminolevulinic acid fuorescence in low-grade gliomas: Clinical article // Journal of Neurosurgery. – 2011. – Vol. 115. – № 4. – P. 740–748. doi: 10.3171/2011.6.JNS11252.

23. Eschbacher J., Martirosyan N. L., Nakaji P. et al. In vivo intraoperative confocal microscopy for real-time histopathological imaging of brain tumors: Clinical article // Journal of Neurosurgery. – 2012. – Vol. 116. – № 4. – P. 854–860. doi: 10.3171/2011.12.JNS11696.

24. Pavlov V., Meyronet D., Meyer-Bisch V. et al. Intraoperative Probe-Based Confocal Laser Endomicroscopy in Surgery and Stereotactic Biopsy of Low-Grade and High-Grade Gliomas: A Feasibility Study in Humans // Neurosurgery. – 2016. – Vol. 79. – № 4. – P. 604–612. doi: 10.1227/NEU.0000000000001365.

25. Liu J. T. C., Meza D., Sanai N. Trends in Fluorescence Image-Guided Surgery for Gliomas // Neurosurgery. – 2014. – Vol. 75. – № 1. – P. 61–71. doi: 10.1227/NEU.0000000000000344.

26. Sibai M., Veilleux I., Elliott J. T. et al. Quantitative spatial frequency fuorescence imaging in the sub-difusive domain for image-guided glioma resection // Biomedical Optics Express. – 2015. – Vol. 6. – № 12. – P. 4923. doi: 10.1364/BOE.6.004923.

27. Wirth D., Kolste K., Kanick S. et al. Fluorescence depth estimation from wide-feld optical imaging data for guiding brain tumor resection: a multi-inclusion phantom study // Biomedical Optics Express. – 2017. – Vol. 8. – № 8. – P. 3656. doi: 10.1364/BOE.8.003656.

28. Belykh E., Miller E. J., Carotenuto A. et al. Progress in Confocal Laser Endomicroscopy for Neurosurgery and Technical Nuances for Brain Tumor Imaging With Fluorescein // Frontiers in Oncology. – 2019. – Vol. 9. – P. 554. doi: 10.3389/fonc.2019.00554.

29. Abramov I., Park M. T., Belykh E. et al. Intraoperative confocal laser endomicroscopy: prospective in vivo feasibility study of a clinical-grade system for brain tumors // Journal of Neurosurgery. – 2023. – Vol. 138. – № 3. – P. 587–597. doi: 10.3171/2022.5.JNS2282.

30. Belykh E., Miller E. J., Patel A. A. et al. Diagnostic Accuracy of a Confocal Laser Endomicroscope for In Vivo Diferentiation Between Normal Injured And Tumor Tissue During Fluorescein-Guided Glioma Resection: Laboratory Investigation // World Neurosurgery. – 2018. – Vol. 115. – P. e337–e348. doi: 10.1016/j.wneu.2018.04.048.

31. Acerbi F., Pollo B., De Laurentis C. et al. Ex Vivo Fluorescein-Assisted Confocal Laser Endomicroscopy (CONVIVO® System) in Patients With Glioblastoma: Results From a Prospective Study // Frontiers in Oncology. – 2020. – Vol. 10. – P. 606574. doi: 10.3389/fonc.2020.606574.

32. Giannoni L., Bonaudo C., Marradi M. et al. Optical characterisation and study of ex vivo glioma tissue for hyperspectral imaging during neurosurgery ed. D. Contini, Y. Hoshi, T. D. O’Sullivan, Munich, Germany: SPIE, 2023. C. 81. doi: 10.1117/12.2670854.

33. Fabelo H., Ortega S., Kabwama S. et al. HELICoiD project: a new use of hyperspectral imaging for brain cancer detection in real-time during neurosurgical operations ed. D. P. Bannon, Baltimore, Maryland, United States:, 2016. C. 986002. doi: 10.1117/12.2223075.

34. Fabelo H., Ortega S., Lazcano R. et al. An Intraoperative Visualization System Using Hyperspectral Imaging to Aid in Brain Tumor Delineation // Sensors. – 2018. – Vol. 18. – № 2. – P. 430. doi: 10.3390/s18020430.

35. Ezhov I., Giannoni L., Shit S. et al. Identifying chromophore fngerprints of brain tumor tissue on hyperspectral imaging using principal component analysis ed. D. Contini, Y. Hoshi, T. D. O’Sullivan, Munich, Germany: SPIE, 2023. C. 78. doi: 10.1117/12.2670775.

36. Martinez B., Leon R., Fabelo H. et al. Most Relevant Spectral Bands Identifcation for Brain Cancer Detection Using Hyperspectral Imaging // Sensors. – 2019. – Vol. 19. – № 24. – P. 5481. doi: 10.3390/s19245481.

37. DePaoli D., Lemoine É., Ember K. et al. Rise of Raman spectroscopy in neurosurgery: a review // Journal of Biomedical Optics. – 2020. – Vol. 25. – № 05. – P. 1. doi: 10.1117/1.JBO.25.5.050901.

38. Romanishkin I. D., Savelieva T. A., Ospanov A. et al. Classifcation of intracranial tumors based on optical-spectral analysis // Biomedical Photonics. – 2023. – Vol. 12. – № 3. – P. 4–10. doi: 10.24931/2413-9432-2023-12-3-4-10.

39. Romanishkin I., Savelieva T., Kosyrkova A. et al. Diferentiation of glioblastoma tissues using spontaneous Raman scattering with dimensionality reduction and data classifcation // Frontiers in Oncology. – 2022. – Vol. 12. – P. 944210. doi: 10.3389/fonc.2022.944210.

40. Hollon T., Orringer D. A. Label-free brain tumor imaging using Raman-based methods // Journal of Neuro-Oncology. – 2021. – Vol. 151. – № 3. – P. 393–402. doi: 10.1007/s11060-019-03380-z.

41. Kast R., Auner G., Yurgelevic S. et al. Identifcation of regions of normal grey matter and white matter from pathologic glioblastoma and necrosis in frozen sections using Raman imaging // Journal of Neuro-Oncology. – 2015. – Vol. 125. – № 2. – P. 287–295. doi: 10.1007/s11060-015-1929-4.

42. Hollon T., Jiang C., Chowdury A. et al. Artifcial-intelligence-based molecular classifcation of difuse gliomas using rapid, label-free optical imaging // Nature Medicine. – 2023. – Vol. 29. – № 4. – P. 828–832. doi: 10.1038/s41591-023-02252-4.

43. Hollon T. C., Pandian B., Adapa A. R. et al. Near real-time intraoperative brain tumor diagnosis using stimulated Raman histology and deep neural networks // Nature Medicine. – 2020. – Vol. 26. – № 1. – P. 52–58. doi: 10.1038/s41591-019-0715-9.

44. Evans C. L., Xu X., Kesari S. et al. Chemically-selective imaging of brain structures with CARS microscopy // Optics Express. – 2007. – Vol. 15. – № 19. – P. 12076. doi: 10.1364/OE.15.012076.

45. Uckermann O., Galli R., Tamosaityte S. et al. Label-Free Delineation of Brain Tumors by Coherent Anti-Stokes Raman Scattering Microscopy in an Orthotopic Mouse Model and Human Glioblastoma // PLoS ONE. – 2014. – Vol. 9. – № 9. – P. e107115. doi: 10.1371/journal.pone.0107115.

46. Mascagni P., Alapatt D., Sestini L. et al. Computer vision in surgery: from potential to clinical value // npj Digital Medicine. – 2022. – Vol. 5. – № 1. – P. 163. doi: 10.1038/s41746-022-00707-5.

47. Rau A., Edwards P. J. E., Ahmad O. F. et al. Implicit domain adaptation with conditional generative adversarial networks for depth prediction in endoscopy // International Journal of Computer Assisted Radiology and Surgery. – 2019. – Vol. 14. – № 7. – P. 1167–1176. doi: 10.1007/s11548-019-01962-w.

48. Kassem H., Alapatt D., Mascagni P. et al. Federated Cycling (FedCy): Semi-Supervised Federated Learning of Surgical Phases // IEEE Transactions on Medical Imaging. – 2023. – Vol. 42. – № 7. – P. 1920–1931. doi: 10.1109/TMI.2022.3222126.

49. Orringer D. A., Pandian B., Niknafs Y. S. et al. Rapid intraoperative histology of unprocessed surgical specimens via fbre-laser-based stimulated Raman scattering microscopy // Nature Biomedical Engineering. – 2017. – Vol. 1. – № 2. – P. 0027. doi: 10.1038/s41551-016-0027.

50. Leon R., Fabelo H., Ortega S. et al. Hyperspectral imaging benchmark based on machine learning for intraoperative brain tumour detection // npj Precision Oncology. – 2023. – Vol. 7. – № 1. – P. 119. doi: 10.1038/s41698-023-00475-9.

51. Fabelo H., Halicek M., Ortega S. et al. Deep Learning-Based Framework for In Vivo Identifcation of Glioblastoma Tumor using Hyperspectral Images of Human Brain // Sensors. – 2019. – Vol. 19. – № 4. – P. 920. doi: 10.3390/s19040920.

52. Ravi D., Fabelo H., Callic G. M. et al. Manifold Embedding and Semantic Segmentation for Intraoperative Guidance With Hyperspectral Brain Imaging // IEEE Transactions on Medical Imaging. – 2017. – Vol. 36. – № 9. – P. 1845–1857. doi: 10.1109/TMI.2017.2695523.

53. Fabelo H., Ortega S., Ravi D. et al. Spatio-spectral classifcation of hyperspectral images for brain cancer detection during surgical operations // PLOS ONE. – 2018. – Vol. 13. – № 3. – P. e0193721. doi: 10.1371/journal.pone.0193721.

54. Florimbi G., Fabelo H., Torti E. et al. Accelerating the K-Nearest Neighbors Filtering Algorithm to Optimize the Real-Time Classifcation of Human Brain Tumor in Hyperspectral Images // Sensors. – 2018. – Vol. 18. – № 7. – P. 2314. doi: 10.3390/s18072314.

55. Zhang Y., Cao G., Li X. et al. Cascaded Random Forest for Hyperspectral Image Classifcation // IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. – 2018. – Vol. 11. – № 4. – P. 1082–1094. doi: 10.1109/JSTARS.2018.2809781.

56. Xia J., Ghamisi P., Yokoya N. et al. Random Forest Ensembles and Extended Multiextinction Profles for Hyperspectral Image Classifcation // IEEE Transactions on Geoscience and Remote Sensing. – 2018. – Vol. 56. – № 1. – P. 202–216. doi: 10.1109/TGRS.2017.2744662.

57. Ayaz H., Tormey D., McLoughlin I. et al. Hyperspectral Brain Tissue Classifcation using a Fast and Compact 3D CNN Approach Genova, Italy: IEEE, 2022. C. 1–4. doi: 10.1109/IPAS55744.2022.10053044.

58. Baig N., Fabelo H., Ortega S. et al. Empirical Mode Decomposition Based Hyperspectral Data Analysis for Brain Tumor Classifcation Mexico: IEEE, 2021. C. 2274–2277. doi: 10.1109/EMBC46164.2021.9629676.

59. Cruz-Guerrero I. A., Leon R., Campos-Delgado D. U. et al. Classifcation of Hyperspectral In Vivo Brain Tissue Based on Linear Unmixing // Applied Sciences. – 2020. – Vol. 10. – № 16. – P. 5686. doi: 10.3390/app10165686.

60. Ruiz L., Martin A., Urbanos G. et al. Multiclass Brain Tumor Classifcation Using Hyperspectral Imaging and Supervised Machine Learning Segovia, Spain: IEEE, 2020. C. 1–6. doi: 10.1109/DCIS51330.2020.9268650.

61. Salvador R., Fabelo H., Lazcano R. et al. Demo: HELICoiD tool demonstrator for real-time brain cancer detection Rennes, France: IEEE, 2016. C. 237–238. doi: 10.1109/DASIP.2016.7853831.

62. Sutradhar P., Sancho J., Villa M. et al. Exploration of Realtime Brain tumor classifcation from Hyperspectral Images in Heterogeneous Embedded MPSoC Pamplona, Spain: IEEE, 2022. C. 01–06. doi: 10.1109/DCIS55711.2022.9970064.

63. Torti E., Florimbi G., Castelli F. et al. Parallel K-Means Clustering for Brain Cancer Detection Using Hyperspectral Images // Electronics. – 2018. – Vol. 7. – № 11. – P. 283. doi: 10.3390/electronics7110283.

64. Urbanos G., Martín A., Vázquez G. et al. Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classifcation // Sensors. – 2021. – Vol. 21. – № 11. – P. 3827. doi: 10.3390/s21113827.

65. Manni F., Van Der Sommen F., Fabelo H. et al. Hyperspectral Imaging for Glioblastoma Surgery: Improving Tumor Identifcation Using a Deep Spectral-Spatial Approach // Sensors. – 2020. – Vol. 20. – № 23. – P. 6955. doi: 10.3390/s20236955.

66. Fürtjes G., Reinecke D., Von Spreckelsen N. et al. Intraoperative microscopic autofuorescence detection and characterization in brain tumors using stimulated Raman histology and two-photon fuorescence // Frontiers in Oncology. – 2023. – Vol. 13. – P. 1146031. doi: 10.3389/fonc.2023.1146031.

67. Reinecke D., Von Spreckelsen N., Mawrin C. et al. Novel rapid intraoperative qualitative tumor detection by a residual convolutional neural network using label-free stimulated Raman scattering microscopy // Acta Neuropathologica Communications. – 2022. – Vol. 10. – № 1. – P. 109. doi: 10.1186/s40478-022-01411-x.

68. Shirshin E. A., Yakimov B. P., Darvin M. E. et al. Label-Free Multiphoton Microscopy: The Origin of Fluorophores and Capabilities for Analyzing Biochemical Processes // Biochemistry (Moscow). – 2019. – Vol. 84. – № S1. – P. 69–88. doi: 10.1134/S0006297919140050.

69. Uckermann O., Galli R., Mark G. et al. Label-free multiphoton imaging allows brain tumor recognition based on texture analysis—a study of 382 tumor patients // Neuro-Oncology Advances. – 2020. – Vol. 2. – № 1. – P. vdaa035. doi: 10.1093/noajnl/vdaa035.

70. Lita A., Sjöberg J., Păcioianu D. et al. Raman-based machine-learning platform reveals unique metabolic diferences between IDHmut and IDHwt glioma // Neuro-Oncology. – 2024. – Vol. 26. – № 11. – P. 1994–2009. doi: 10.1093/neuonc/noae101.

71. Morais C. L. M., Lilo T., Ashton K. M. et al. Determination of meningioma brain tumour grades using Raman microspectroscopy imaging // The Analyst. – 2019. – Vol. 144. – № 23. – P. 7024–7031. doi: 10.1039/C9AN01551E.