Deep Learning for Head and Neck Cancer Detection
For patients undergoing surgical cancer resection of squamous cell carcinoma (SCCa), cancer-free surgical margins are essential for good prognosis. We developed a method to use hyperspectral imaging, a non-contact optical imaging modality, and convolutional neural networks (CNNs) to perform an optical biopsy of ex-vivo, surgical gross-tissue specimens undergoing surgical cancer resection. Supervised machine learning and artificial intelligence algorithms have demonstrated the ability to classify images after being allowed to learn features from training or example images. One such method, convolutional neural networks (CNNs), have demonstrated astounding performance at image classification tasks due to their capacity for robust handling of training sample variance and ability to extract features from large training data sizes. a simple binary classification is performed, i.e. cancer versus normal, and second, multi-class sub-classification of normal upper aerodigestive tract samples is investigated. If proven to be reliable and generalizable, this method could help provide intra-operative diagnostic information beyond palpation and visual inspection to the surgeon’s resources, perhaps enabling surgeons to achieve more accurate cuts and biopsies, or as a computer-aided diagnostic tool for physicians diagnosing and treating these types of cancer.
Figure 1: Modified inception module for use in the 2D-CNN architecture for classifying HSI of tissues from the upper aerodigestive tract.
Figure 2: Convolutional neural network (CNN) architectures implemented for classification of HSI of thyroid tissue (left) and tissue from the upper aerodigestive tract (right).
Figure 3: Representative results of binary cancer classification. Left: HSI-RGB composite images with cancer ROI outlined. Center: Respective histological gold standard with corresponding ROI outlined. Right: Artificially colored CNN classification results. True positive results representing correct cancer identification are visualized in red, and false negatives representing incorrect normal identification are shown in yellow. Tissue shown in grayscale represents tissue that is not classified due to the tissue surface containing glare pixels causing insufficient area to produce the necessary patch-size for classification.