Deep Learning for the Preoperative Diagnosis of Metastatic Cervical Lymph Nodes on Contrast-Enhanced Computed ToMography in Patients with Oral Squamous Cell Carcinoma

Abstract

We investigated the value of deep learning (DL) in differentiating between benign and metastatic cervical lymph nodes (LNs) using pretreatment contrast-enhanced computed tomography (CT). This retrospective study analyzed 86 metastatic and 234 benign (non-metastatic) cervical LNs at levels I–V in 39 patients with oral squamous cell carcinoma (OSCC) who underwent preoperative CT and neck dissection. LNs were randomly divided into training (70%), validation (10%), and test (20%) sets. For the validation and test sets, cervical LNs at levels I–II were evaluated. Convolutional neural network analysis was performed using Xception architecture. Two radiologists evaluated the possibility of metastasis to cervical LNs using a 4-point scale. The area under the curve of the DL model and the radiologists’ assessments were calculated and compared at levels I–II, I, and II. In the test set, the area under the curves at levels I–II (0.898) and II (0.967) were significantly higher than those of each reader (both, p < 0.05). DL analysis of pretreatment contrast-enhanced CT can help classify cervical LNs in patients with OSCC with better diagnostic performance than radiologists’ assessments alone. DL may be a valuable diagnostic tool for differentiating between benign and metastatic cervical LNs.

Utilizing Deep Machine Learning for Prognostication of Oral Squamous Cell Carcinoma—A Systematic Review

“…The majority of these studies used a convolutional neural network (CNN) [ 2 , 15 – 22 , 24 – 26 , 28 , 31 – 36 , 38 – 41 , 43 – 45 , 48 , 49 ]. Several data types such as gene expression data [ 15 , 45 ], spectra data [ 20 , 21 , 29 , 34 , 37 , 44 , 48 ], and other image data types—anatomical [ 16 ], intraoral [ 17 ], histology [ 18 , 27 ], auto-fluorescence [ 19 , 22 ], cytology-image [ 23 ], neoplastic [ 40 ], clinical [ 28 , 36 , 38 ], oral lesions [ 42 ], computed tomography images [ 24 – 26 , 33 , 35 , 41 , 49 ], clinicopathologic [ 2 ], saliva metabolites [ 31 ], histopathological [ 30 , 32 , 43 ], and pathological [ 39 ] images have been used in the included studies.…”

“…Additionally, specificity and accuracy were also used to demonstrate the performance of the deep learning model for prognostication in OSCC [ 23 ]. Other studies used either accuracy, C-index (concordance index), F1-score, or Dice similarity coefficient (Dsc) mean value as the performance metrics for reporting the potential benefits of the deep learning model [ 2 , 18 , 18 , 26 , 28 , 29 , 31 – 33 , 39 – 41 , 44 , 45 , 49 ].…”

“…Most of these studies used either spectra data or computed tomography images [ 20 , 21 , 24 – 26 , 29 , 33 – 35 , 37 , 41 , 44 , 48 , 49 ]. The average specificity, sensitivity, area under receiving operating characteristics curve [AUC]), and accuracy for studies that used spectra data were 0.97, 0.99, 0.96, and 96.6%, respectively.…”

Deep Machine Learning for Oral Cancer: From Precise Diagnosis to Precision Medicine

“…For example, for precise diagnosis purposes, deep learning models have been used in the detection of oral cancer [24,25,[64][65][66][67][68][69][70][71][72][73][74][75]. Additionally, these models have assisted in the prediction of lymph node metastasis [27][28][29]76]. Besides, they have been reported to perform well in differentiating between precancerous and cancerous lesions [64,[77][78][79][80][81].…”

Nuclear Morphology is a Deep Learning Biomarker of Senescence Across Tissues and Species

“…After seeing accuracy converge to a steady level, the model was applied to validation data. We also compared our custom network to Xception, one of the top performing models for image classification that has been often applied to biomedical classification 24,25 . Xception achieved superior results with an f1score of 94%, accuracy of 95%, and AUC of 0.99 with validation data (Fig.…”