P03.18 Detection of human brain cancer in pathological slides using hyperspectral images

Abstract

INTRODUCTION: Extraventricular neurocytomas are rare brain tumors with a reported worldwide incidence of only 0.13%. They originate from neuroepithelial tissue and may present throughout the central nervous system. They affect mostly young adults, and have a favorable prognosis. Due to their rarity, they are not well characterized and most features are derived from case reports. Extraventricular neurocytomas tend to be well circumscribed, contrast-enhancing and heterogeneously solid, often partly or mainly cystic. They share histological features with the more common central neurocytomas but are often more complex, less cellular, and more likely to contain ganglion cells. The primary differential diagnoses are oligodendroglioma, ganglioglioma, gangliocytoma, and dysembryoplastic neuropithelial tumor (DNET). Complete surgical resection is currently the optimal treatment choice. It is not clear whether tumors with anaplasia have a higher relapse rate or if they require adjuvant therapy. MATERIALS AND METHODS: We report on a 48 year-old woman with a history of depression who was admitted for focal seizures characterized by left gaze tonic deviation over the last few months. The neurological examination was otherwise normal. Routine EEG showed no epileptic activity. MRI identified an intra-axial heterogeneous round mass in the right frontal lobe with faint heterogeneous enhancement after contrast administration and perilesional oedema. A low-grade glial tumor was suspected. RESULTS: A left frontal craniotomy was performed. Intraoperative findings revealed a grayish and friable mass without superficial expression and with well-defined limits. A gross total resection was achieved. Histological analysis revealed a lesion with rounded cells displaying a "saltand-pepper" chromatin pattern, Homer Wright rosettes and perivascular pseudorosettes. Immunostaining was positive for GFAP, sinaptophysin and NSE. The Ki-67 labeling index of the tumor was 25%. Diagnosis was extraventricular neurocytoma with anaplastic features. Magnetic resonance imaging was performed at 3 months post-surgery and demonstrated no evidence of tumor recurrence. CONCLUSION: Although neurocytomas are often located in the lateral ventricles, they occur throughout the neuroaxis. This is the main reason why they are not always considered in the differential diagnosis of intraparenchymal cerebral tumors. The diagnosis is aided by radiologic and histologic features. If a gross total resection is achieved, one can expect a good prognosis.

Adaptive deep learning for head and neck cancer detection using hyperspectral imaging

“…Fourier coefficients [13], normalized difference nuclear index [14], sparse representation [15], box-plot and the watershed method [16], superpixel method [9], markov random fields [17,18], and morphological method [19], were used for hyperspectral image processing and quantification analysis; (3) Machine learning techniques. Many of the advancements have been done in cancer identification using traditional machine learning classification models, such as linear discriminant analysis [20][21][22][23][24][25][26], quadratic discriminant analysis [21], support vector machine [12,17,[20][21][22][27][28][29][30][31][32][33][34][35][36][37], decision trees [22], k-nearest neighbors algorithm [22,38], k-means [12,19,39], naïve bayes [22], random forests [21,22,34,37], maximum likelihood [40], minimum spanning forest [31], gaussian m...…”

Detecting brain tumor in pathological slides using hyperspectral imaging

“…Three different classifiers have been employed to automatically distinguish between tumor and normal tissue, using as features only the spectral information of the tissues. A qualitative description of this methodology has been recently published [16].…”