SLIM Brain Database: A Multimodal Image Database of In-vivo Human Brains for Tumour Detection

Martín-Pérez, Alberto; Villa, Manuel; Olmeda, Gonzalo Rosa; Sancho, Jaime; Vazquez, Guillermo; Urbanos, Gemma; de Ternero, Alejandro Martinez; Chavarrías, Miguel; Jimenez-Roldan, Luis; Perez-Nuñez, Angel; Lagares, Alfonso; Juarez, Eduardo; Sanz, César

doi:10.21203/rs.3.rs-3629358/v1

Cited by 2 publications

(3 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This section describes the materials used to conduct the experiments to test the Libra classifier. Specifically, the two databases, Helicoid [10] and Slim Brain [30], employed in this work are detailed in Section 4.1, whereas the classification chain used as baseline is described in Section 4.2.…”

Section: Methodsmentioning

confidence: 99%

“…The preprocessing and labeling procedures applied to the captures in the Slim Brain database are detailed in [30]. The preprocessing chain consists of four main steps: cube conformation, white-dark calibration, spectral correction, and normalization.…”

Section: Databasesmentioning

confidence: 99%

“…This paper presents Libra, a low spectral resolution brain tumor classifier, which is evaluated with the Helicoid and Slim Brain database [10,30]. Furthermore, Libra is compared to the classification chain proposed in the Helicoid project, which is used as a baseline.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

LIBRA: Low spectral resolution brain tumor classifier for medical hyperspectral imaging

Villa,

Martín-Perez,

Vazquez

et al. 2024

Preprint

View full text Add to dashboard Cite

Gliomas constitute a significant challenge in neurosurgery due to their high incidence and poor prognosis. Despite advancements in tumor detection techniques using machine learning approach with hyperspectral imaging, accurately distinguishing between healthy and tumoral tissues remains challenging. Following this trend, this paper introduces Libra, a low spectral resolution classifier designed for brain tumor detection, leveraging ensemble learning techniques to enhance classification performance. Evaluated across the Helicoid and Slim Brain databases, Libra demonstrates superior tumor sensitivity and a notable increase in accuracy compared to standalone SVM and the state-of-the-art Helicoid classification chain. While facing challenges in accurately distinguishing between blood vessels and tumoral tissues, Libra performs 32 % better than Helicoid in tumor sensitivity and 25.1 % in tumor F1-score. Moreover, when using unseen data, Libra demonstrates notable improvements of 7.8 % and 6.9 % respectively. These improvements are obtained using low complexity algorithms exploiting ensemble learning techniques.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Databasesmentioning

confidence: 99%

See 1 more Smart Citation

LIBRA: Low spectral resolution brain tumor classifier for medical hyperspectral imaging

Villa,

Martín-Perez,

Vazquez

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

Spectral Analysis Comparison of Pushbroom and Snapshot Hyperspectral Cameras forIn-VivoBrain Tissues and Chromophores Identification

Martín-Pérez,

de Ternero,

Lagares

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

SignificanceHyperspectral imaging sensors have rapidly advanced, aiding in tumor diagnostics forin-vivobrain tumors. Linescan cameras effectively distinguish between pathological and healthy tissue, while snapshot cameras offer a potential alternative to reduce acquisition time.AimOur research compares linescan and snapshot hyperspectral cameras forin-vivobrain tissues and chromophores identification.ApproachWe compared a lines-can pushbroom camera and a snapshot camera using images from 10 patients with various pathologies. Objective comparisons were made using unnormalized and normalized data for healthy and pathological tissues. We utilized Interquartile Range (IQR) for the Spectral Angle Mapping (SAM), the Goodness-of-Fit Coefficient (GFC), and the Root Mean Square Error (RMSE) within the 659.95 to 951.42 nm range. Additionally, we assessed the ability of both cameras to capture tissue chromophores by analyzing absorbance from reflectance information.ResultsThe SAM metric indicates reduced dispersion and high similarity between cameras for pathological samples, with a 9.68% IQR for normalized data compared to 2.38% for unnormalized data. This pattern is consistent across GFC and RMSE metrics, regardless of tissue type. Moreover, both cameras could identify absorption peaks of certain chromophores. For instance, using the absorbance measurements of the linescan camera we obtained SAM values below 0.235 for four peaks, regardless of the tissue and type of data under inspection. These peaks are: one for cytochrome b in its oxidised form atλ= 422 nm, two for HbO2 atλ= 542 nm andλ= 576 nm, and one for water atλ= 976 nm.ConclusionThe spectral signatures of the cameras show more similarity with unnormalized data, likely due to snapshot sensor noise, resulting in noisier signatures post-normalization. Comparisons in this study suggest that snapshot cameras might be viable alternatives to linescan cameras for real-time brain tissues identification.

show abstract

SLIM Brain Database: A Multimodal Image Database of In-vivo Human Brains for Tumour Detection

Cited by 2 publications

References 34 publications

LIBRA: Low spectral resolution brain tumor classifier for medical hyperspectral imaging

LIBRA: Low spectral resolution brain tumor classifier for medical hyperspectral imaging

Spectral Analysis Comparison of Pushbroom and Snapshot Hyperspectral Cameras forIn-VivoBrain Tissues and Chromophores Identification

Contact Info

Product

Resources

About