3.1.6. Brain Cancer

In this section, we will review the application of optical-based biosensors and machine learning for brain cancer detection.

A supervised machine learning approach is commonly used for the identification and classification of cancer cell gestures, enabling early diagnosis. Hasan et al. [94] developed a system that captures time-lapse images of cancer cells and analyzes their morphological changes over time using image-processing techniques (Figure 6). The system also incorporates machine learning algorithms for the automated classification of cancer cells based on their dynamic morphology. As a proof of concept, the morphologies of human glioblastoma (hGBM), which causes brain tumors [95,96], and astrocyte cells were used. The cells were captured and imaged with an optical microscope. Three different classifier models, the SVM, RF, and naïve Bayes classifier (NBC), were trained with the known dataset using machine learning algorithms. All the classifier models detected the cancer cells with an average accuracy of at least 82%.

In another study, Hossain et al. [97] employed a sensor-based portable microwave brain imaging (SMBI) system to obtain the reconstructed microwave (RMW) brain images. The proposed method consists of a segmentation model called MicrowaveSegNet (MsegNet) and a classifier called BrainImageNet (BINet). A dataset of 300 RMW brain image samples was used to create an original dataset, which was then augmented to make 6000 training images for a five-fold cross-validation. The performance of MsegNet and BINet was

compared to state-of-the-art segmentation and classification models, and the proposed models achieved impressive results. The proposed cascaded model has the potential to be used in sensor-based SMBI systems to investigate the progression of brain cancer disease.

**Figure 6.** The schematic shows an overview of the dynamic morphological analysis of cell gestures. Reprinted with permission from [94]. Copyright 2018 Elsevier.
