Raman spectroscopy and supervised learning as a potential tool to identify high‐dose‐rate‐brachytherapy induced biochemical profiles of prostate cancer

Abstract

High‐dose‐rate‐brachytherapy (HDR‐BT) is an increasingly attractive alternative to external beam radiation‐therapy for patients with intermediate risk prostate cancer. Despite this, no bio‐marker based method currently exists to monitor treatment response, and the changes which take place at the biochemical level in hypo‐fractionated HDR‐BT remain poorly understood. The aim of this pilot study is to assess the capability of Raman spectroscopy (RS) combined with principal component analysis (PCA) and random‐forest classification (RF) to identify radiation response profiles after a single dose of 13.5 Gy in a cohort of nine patients. We here demonstrate, as a proof‐of‐concept, how RS‐PCA‐RF could be utilised as an effective tool in radiation response monitoring, specifically assessing the importance of low variance PCs in complex sample sets. As RS provides information on the biochemical composition of tissue samples, this technique could provide insight into the changes which take place on the biochemical level, as result of HDR‐BT treatment.

Understanding radiation response and cell cycle variation in brain tumour cells using Raman spectroscopy

“…An advantage of using RF for classification is the ability to evaluate the importance of the identified predictor variables for correct classification. 73,74 The ten most important PCs to correctly classify the data as irradiated and non-irradiated for 1, 4 and 24 hours post irradiation are shown in Fig. S3.…”

Radiation treatment response and hypoxia biomarkers revealed by machine learning assisted Raman spectroscopy in tumour cells and xenograft tissues

“…In previous research, RS has demonstrated the potential to be incorporated into radiation therapy as a label-free technique to reveal the biochemical dynamics and assess the treatment response for various types of cancer. [9][10][11][12][13][14][15][16][17][59][60][61][62][63] Nevertheless, conventional unsupervised dimension reduction techniques can provide limited resolution in distinguishing responses from different chemicals. [9][10][11][12][13][14][15][16][17][59][60][61][62][63] In this manuscript, the GBR-NMF-RF-SHAP data analytical framework has demonstrated an outstanding capability to resolve signals such as radiation response and hypoxia indicators corresponding to constrained chemicals.…”

“…[9][10][11][12][13][14][15][16][17][59][60][61][62][63] Nevertheless, conventional unsupervised dimension reduction techniques can provide limited resolution in distinguishing responses from different chemicals. [9][10][11][12][13][14][15][16][17][59][60][61][62][63] In this manuscript, the GBR-NMF-RF-SHAP data analytical framework has demonstrated an outstanding capability to resolve signals such as radiation response and hypoxia indicators corresponding to constrained chemicals. Moreover, GBR-NMF-RF-SHAP also exhibited efficiency in analyses across various types of cancer data (cell and tissue).…”

“…To analyze the spectral data acquired from RS, dimensionality reduction techniques are typically applied to parse the covariant features in the spectra 8 . Unsupervised dimensionality reduction techniques such as principal component analysis (PCA) and non-negative matrix factorization (NMF) have helped discover glycogen and lipids as radiation response Raman biomarkers in cancer cells and tissue samples [9][10][11][12][13][14][15][16][17] . However, unsupervised dimensionality reduction methods can also lead to interpretability problems when used to interpret the biochemical information contained within RS data, as biological components do not necessarily segregate along individual reduced dimensions (eg.…”

Recent Advances in Raman Spectral Imaging in Cell Diagnosis and Gene Expression Prediction

“…In this manner, the complex Raman scattering spectra of a cell can be effectively used as a “cell fingerprint” using machine learning in combination. Because of the high-content data obtained by Raman spectroscopy, it is compatible to other machine learning methods such as random-forest classification, support vector machine or k-means classification and being used in cancer diagnosis or evaluation of cancer treatment [ 33 , 34 , 35 ]. Use of Raman spectra in cancer discrimination is recently becoming popular in combination with machine-learning, including artificial intelligence [ 35 , 36 , 37 ], which is expected to further develop in the near future.…”