HeiPorSPECTRAL - the Heidelberg Porcine HyperSPECTRAL Imaging Dataset of 20 Physiological Organs

Studier-Fischer, Alexander; Seidlitz, Silvia; Sellner, Jan; Bressan, Maria Antonietta; Özdemir, Berkin; Ayala, Leonardo; Odenthal, Jan; Knoedler, Samuel; Kowalewski, Karl‐Friedrich; Haney, Caelán Max; Salg, Gabriel Alexander; Dietrich, Maximilian; Kenngott, Hannes; Gockel, Ines; Hackert, Thilo; Müller‐Stich, Beat P.; Maier-Hein, Lena; Nickel, Felix

doi:10.1038/s41597-023-02315-8

Cited by 13 publications

(3 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous efforts in the state of the art employ statistical estimators to tune their parameters and reduce the influence of outliers, while using pulsioximetry data as a reference for their estimations based on RGB images [14], [15]. In recent works [38]- [40], perfusion parameters were identified for a hyperspectral system [19], based on isosbestic points, band selection, and ratio index methods. Especially when dealing with in-vivo human tissue, MSI data could identify blended spectral signatures, since the camera captures combined reflectance from the skin, influenced by the relative proportions of melanin, hemoglobin, and other chromophores, which ultimately dictate the color of the surface.…”

Section: Discussion and Final Remarksmentioning

confidence: 99%

Contactless Skin Blood Perfusion Imaging via Multispectral Information, Spectral Unmixing and Multivariable Regression

Granados-Castro,

Gutierrez-Navarro,

Mejia-Rodriguez

et al. 2024

IEEE Open J. Signal Process.

View full text Add to dashboard Cite

Noninvasive methods for assessing in-vivo skin blood perfusion parameters, such as hemoglobin oxygenation, are crucial for diagnosing and monitoring microvascular diseases. This approach is particularly beneficial for patients with compromised skin, where standard contact-based clinical devices are inappropriate. For this goal, we propose the analysis of multimodal data from an occlusion protocol applied to 18 healthy participants, which includes multispectral imaging of the whole hand and reference photoplethysmography information from the thumb. Multispectral data analysis was conducted using two different blind linear unmixing methods: principal component analysis (PCA), and extended blind endmember and abundance extraction (EBEAE). Perfusion maps for oxygenated and deoxygenated hemoglobin changes in the hand were generated using linear multivariable regression models based on the unmixing methods. Our results showed high accuracy, with R 2 -adjusted values, up to 0.90±0.08. Further analysis revealed that using more than four characteristic components during spectral unmixing did not improve the fit of the model. Bhattacharyya distance results showed that the fitted models with EBEAE were more sensitive to hemoglobin changes during occlusion stages, up to four times higher than PCA. Our study concludes that multispectral imaging with EBEAE is effective in quantifying changes in oxygenated hemoglobin levels, especially when using 3 to 4 characteristic components. Our proposed method holds promise for the noninvasive diagnosis and monitoring of superficial microvascular alterations across extensive anatomical regions.INDEX TERMS skin blood perfusion, hemoglobin, multispectral imaging, photoplethysmography, spectral unmixing.

show abstract

Section: Discussion and Final Remarksmentioning

confidence: 99%

Contactless Skin Blood Perfusion Imaging via Multispectral Information, Spectral Unmixing and Multivariable Regression

Granados-Castro,

Gutierrez-Navarro,

Mejia-Rodriguez

et al. 2024

IEEE Open J. Signal Process.

View full text Add to dashboard Cite

show abstract

“…Raw data were obtained and analyzed using the annotation software [ 35 , 36 ]. Then, data were entered into a spreadsheet and the statistical evaluation was done with GraphPad Prism version 9.2.0. for Mac (GraphPad Software, San Diego, California, USA).…”

Section: Methodsmentioning

confidence: 99%

Effects of endoluminal vacuum sponge therapy on the perfusion of gastric conduit in a porcine model for esophagectomy

Felinska,

Studier-Fischer,

Özdemir

et al. 2024

Surg Endosc

Self Cite

View full text Add to dashboard Cite

Background After esophagectomy, the postoperative rate of anastomotic leakage is up to 30% and is the main driver of postoperative morbidity. Contemporary management includes endoluminal vacuum sponge therapy (EndoVAC) with good success rates. Vacuum therapy improves tissue perfusion in superficial wounds, but this has not been shown for gastric conduits. This study aimed to assess gastric conduit perfusion with EndoVAC in a porcine model for esophagectomy. Material and methods A porcine model (n = 18) was used with gastric conduit formation and induction of ischemia at the cranial end of the gastric conduit with measurement of tissue perfusion over time. In three experimental groups EndoVAC therapy was then used in the gastric conduit (− 40, − 125, and − 200 mmHg). Changes in tissue perfusion and tissue edema were assessed using hyperspectral imaging. The study was approved by local authorities (Project License G-333/19, G-67/22). Results Induction of ischemia led to significant reduction of tissue oxygenation from 65.1 ± 2.5% to 44.7 ± 5.5% (p < 0.01). After EndoVAC therapy with − 125 mmHg a significant increase in tissue oxygenation to 61.9 ± 5.5% was seen after 60 min and stayed stable after 120 min (62.9 ± 9.4%, p < 0.01 vs tissue ischemia). A similar improvement was seen with EndoVAC therapy at − 200 mmHg. A nonsignificant increase in oxygenation levels was also seen after therapy with − 40 mmHg, from 46.3 ± 3.4% to 52.5 ± 4.3% and 53.9 ± 8.1% after 60 and 120 min respectively (p > 0.05). An increase in tissue edema was observed after 60 and 120 min of EndoVAC therapy with − 200 mmHg but not with − 40 and − 125 mmHg. Conclusions EndoVAC therapy with a pressure of − 125 mmHg significantly increased tissue perfusion of ischemic gastric conduit. With better understanding of underlying physiology the optimal use of EndoVAC therapy can be determined including a possible preemptive use for gastric conduits with impaired arterial perfusion or venous congestion. Graphical abstract

show abstract

“…Sparsity of real-world data: training data from emerging imaging modalities, such as HSI, lacks the diversity to represent the full range of pathologies encountered in real-word medical settings. For example, the largest publicly available HSI data set in visceral surgery, HeiPorSpectral [ 14 ], solely features images from well-perfused tissue.…”

Section: Methodsmentioning

confidence: 99%

Test-time augmentation with synthetic data addresses distribution shifts in spectral imaging

Qasim,

Motta,

Studier-Fischer

et al. 2024

Int J CARS

Self Cite

View full text Add to dashboard Cite

Purpose Surgical scene segmentation is crucial for providing context-aware surgical assistance. Recent studies highlight the significant advantages of hyperspectral imaging (HSI) over traditional RGB data in enhancing segmentation performance. Nevertheless, the current hyperspectral imaging (HSI) datasets remain limited and do not capture the full range of tissue variations encountered clinically. Methods Based on a total of 615 hyperspectral images from a total of 16 pigs, featuring porcine organs in different perfusion states, we carry out an exploration of distribution shifts in spectral imaging caused by perfusion alterations. We further introduce a novel strategy to mitigate such distribution shifts, utilizing synthetic data for test-time augmentation. Results The effect of perfusion changes on state-of-the-art (SOA) segmentation networks depended on the organ and the specific perfusion alteration induced. In the case of the kidney, we observed a performance decline of up to 93% when applying a state-of-the-art (SOA) network under ischemic conditions. Our method improved on the state-of-the-art (SOA) by up to 4.6 times. Conclusion Given its potential wide-ranging relevance to diverse pathologies, our approach may serve as a pivotal tool to enhance neural network generalization within the realm of spectral imaging.

show abstract

HeiPorSPECTRAL - the Heidelberg Porcine HyperSPECTRAL Imaging Dataset of 20 Physiological Organs

Cited by 13 publications

References 41 publications

Contactless Skin Blood Perfusion Imaging via Multispectral Information, Spectral Unmixing and Multivariable Regression

Contactless Skin Blood Perfusion Imaging via Multispectral Information, Spectral Unmixing and Multivariable Regression

Effects of endoluminal vacuum sponge therapy on the perfusion of gastric conduit in a porcine model for esophagectomy

Test-time augmentation with synthetic data addresses distribution shifts in spectral imaging

Contact Info

Product

Resources

About