Hyperspectral and multispectral image processing for gross-level tumor detection in skin lesions: a systematic review

Abstract

Significance: Skin cancer is one of the most prevalent cancers worldwide. In the advent of medical digitization and telepathology, hyper/multispectral imaging (HMSI) allows for noninvasive, nonionizing tissue evaluation at a macroscopic level.Aim: We aim to summarize proposed frameworks and recent trends in HMSI-based classification and segmentation of gross-level skin tissue.Approach: A systematic review was performed, targeting HMSI-based systems for the classification and segmentation of skin lesions during gross pathology, including melanoma, pigmented lesions, and bruises. The review adhered to the 2020 Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines. For eligible reports published from 2010 to 2020, trends in HMSI acquisition, preprocessing, and analysis were identified.Results: HMSI-based frameworks for skin tissue classification and segmentation vary greatly. Most reports implemented simple image processing or machine learning, due to small training datasets. Methodologies were evaluated on heavily curated datasets, with the majority targeting melanoma detection. The choice of preprocessing scheme influenced the performance of the system. Some form of dimension reduction is commonly applied to avoid redundancies that are inherent in HMSI systems.Conclusions: To use HMSI for tumor margin detection in practice, the focus of system evaluation should shift toward the explainability and robustness of the decision-making process.

Effects of dimension reduction of hyperspectral images in skin gross pathology

“…in the past decade in various settings, in vivo, at gross level or during histology. 4,5 The concentrations of chromophore molecules, mainly melanin and hemoglobin, affect the attenuation of radiation through skin layers. 6,7 Therefore, spectral signatures of the tissue could give insights about its contents.…”

“…The suitability of dimension reduction has not been investigated comparatively in segmentation tasks during skin gross pathology. 5 Components produced by a suitable dimension reduction method could assist the discrimination between tumor and nonneoplastic tissue. At the same time, the visualization of important components could assist the diagnosis of tumor margins and be more familiar to medical staff.…”

“…Histology is costly and time‐consuming, adding delays in the diagnostic process and increased burden to short‐staffed pathology laboratories 3 . Applications of HSI for the diagnosis of skin pathologies have been investigated in the past decade in various settings, in vivo, at gross level or during histology 4,5 . The concentrations of chromophore molecules, mainly melanin and hemoglobin, affect the attenuation of radiation through skin layers 6,7 .…”

“…Until recently, gross pathology research focused only on the classification of entire tissue images or manually selected tissue areas. HSI applications in gross pathology show potential for tasks, such as the detection of tumor margins, identification of disease type, and monitoring of disease progression 5 . Previous studies investigated the effects of dimension reduction in different multidimensional datasets, including biological data, 33 general HSI, 34 or remote sensing HSI 35 .…”

“…However, there is a research gap regarding the use of dimension reduction in tasks that utilize medical and/or gross pathology HSI data. The suitability of dimension reduction has not been investigated comparatively in segmentation tasks during skin gross pathology 5 . Components produced by a suitable dimension reduction method could assist the discrimination between tumor and nonneoplastic tissue.…”