Classification of imbalanced oral cancer image data from high-risk population

Song, Bofan; Li, Shaobai; Sunny, Sumsum; Gurushanth, Keerthi; Mendonca, Pramila; Mukhia, Nirza; Patrick, Sanjana; Gurudath, Shubha; Raghavan, Subhashini; Imchen, Tsusennaro; Leivon, Shirley T.; Kolur, Trupti; Shetty, Vivek; Rangappa, Vidya Bhushan; Ramesh, Rohan Michael; Peterson, T.W.; Pillai, Vijay; Wilder‐Smith, Petra; Sigamani, Alben; Suresh, Amritha; Kuriakose, Moni Abraham; Birur, Praveen; Liang, Rongguang

doi:10.1117/1.jbo.26.10.105001

Cited by 17 publications

(9 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this work, the performance of CNN-based image classification models works well to identify the OSCC and OPMDs. The results, particularly in DenseNet-169 and ResNet-101, achieved near-perfect AUC and showed performance similar to the classification of multiclass image of OSCC and OPMDs on oral images as a CNN model of the study of Fu et al [34], Tanriver et al [35] and Song et al [36] but more accurate than the studies of Welikala et al [37]. The difference in the performance of models may be from variations in the class distribution of each study.…”

Section: Plos Onementioning

confidence: 59%

See 1 more Smart Citation

AI-based analysis of oral lesions using novel deep convolutional neural networks for early detection of oral cancer

et al. 2022

View full text Add to dashboard Cite

Artificial intelligence (AI) applications in oncology have been developed rapidly with reported successes in recent years. This work aims to evaluate the performance of deep convolutional neural network (CNN) algorithms for the classification and detection of oral potentially malignant disorders (OPMDs) and oral squamous cell carcinoma (OSCC) in oral photographic images. A dataset comprising 980 oral photographic images was divided into 365 images of OSCC, 315 images of OPMDs and 300 images of non-pathological images. Multiclass image classification models were created by using DenseNet-169, ResNet-101, SqueezeNet and Swin-S. Multiclass object detection models were fabricated by using faster R-CNN, YOLOv5, RetinaNet and CenterNet2. The AUC of multiclass image classification of the best CNN models, DenseNet-196, was 1.00 and 0.98 on OSCC and OPMDs, respectively. The AUC of the best multiclass CNN-base object detection models, Faster R-CNN, was 0.88 and 0.64 on OSCC and OPMDs, respectively. In comparison, DenseNet-196 yielded the best multiclass image classification performance with AUC of 1.00 and 0.98 on OSCC and OPMD, respectively. These values were inline with the performance of experts and superior to those of general practictioners (GPs). In conclusion, CNN-based models have potential for the identification of OSCC and OPMDs in oral photographic images and are expected to be a diagnostic tool to assist GPs for the early detection of oral cancer.

show abstract

Section: Plos Onementioning

confidence: 59%

“…To the best of our knowledge, this is the first study to use Swin-S for classification of oral lesions. Previous studies [35][36][37]39,…”

Section: Plos Onementioning

confidence: 98%

AI-based analysis of oral lesions using novel deep convolutional neural networks for early detection of oral cancer

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Artificial intelligence has also been used to identify potentially malignant oral lesions [ 44 ]. The deep learning-based classification model MobileNet for processing the images and real-time classification has been reported in resource-restricted areas irrespective of internet connectivity [ 44 ]. This study used a dual-mode oral cancer screening system based on a smartphone to run the deep learning-based technique.…”

Section: Resultsmentioning

confidence: 99%

“…However, higher sensitivity and specificity of 94% and 95.5% were seen in determining the need for further referral [36]. Artificial intelligence has also been used to identify potentially malignant oral lesions [44]. e deep learningbased classification model MobileNet for processing the images and real-time classification has been reported in resource-restricted areas irrespective of internet connectivity [44].…”

Section: Characteristics Of Clinical Studies Using Mhealth For Oralmentioning

confidence: 99%

Mobile Health (mHealth) Technology in Early Detection and Diagnosis of Oral Cancer-A Scoping Review of the Current Scenario and Feasibility

Dailah

2022

Journal of Healthcare Engineering

View full text Add to dashboard Cite

Objective. Oral cancer is one of the most common types of cancer with dreadful consequences. But it can be detected early without much expensive equipment. Screening and early detection of oral cancer using Mobile health (mHealth) technology are reported due to the availability of the extensive network of mobile phones across populations. Therefore, we aimed to explore the existing literature regarding mHealth feasibility in the early detection of oral cancer. Materials and Method. An extensive search was conducted to explore the literature on the feasibility of mobile health for early oral cancer. Clinical studies reporting kappa agreement between on-site dentists and offsite health care workers/dentists in the early detection of oral cancer were included in this review. Studies describing the development of a diagnostic device, app development, and qualitative interviews among practitioners trained in using mobile health were also included in this review for a broader perspective on mHealth. Results. While most of the studies described various diagnostic accuracies using mHealth for oral cancer early detection, few studies reported the development of mobile applications, novel device designs for mHealth applications, and the feasibility of a few mHealth programs for early oral cancer detection. Community health workers equipped with a mobile phone-based app could identify “abnormal” oral lesions. Overall, many studies reported high sensitivity, specificity, and Kappa value of agreement. Effectiveness, advantages, and barriers in oral cancer screening using mHealth are also described. Conclusion. The overall results show that remote diagnosis for early detection of oral cancer using mHealth was found useful in remote settings.

show abstract

“…Based on numerous studies, deep learning has been applied to mineral mapping 4 , hyperspectral image classification problems 5 6 , oral cancer image classification 7 , breast cancer image classification 8 , skin cancer prediction 9 , potato and rice disease prediction 10 , and many others. Among them, in the field of diagnosing breast cancer, histopathological image analysis is an important technique in its early stages, but its efficiency is not guaranteed in practical applications 11 .…”

Section: Introductionmentioning

confidence: 99%

A novel breast cancer image classification model based on multiscale texture feature analysis and dynamic learning

Guo,

Yuan,

Shi

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Assistive medical image classifiers can greatly reduce the workload of medical personnel. However, traditional machine learning methods require large amounts of well-labeled data and long learning times to solve medical image classification problems, which can lead to high training costs and poor applicability. To address this problem, a novel unsupervised breast cancer image classification model based on multiscale texture analysis and a dynamic learning strategy for mammograms is proposed in this paper. First, a gray-level cooccurrence matrix and Tamura coarseness are used to transfer images to multiscale texture feature vectors. Then, an unsupervised dynamic learning mechanism is used to classify these vectors. In the simulation experiments with a resolution of 40 pixels, the accuracy, precision, F1-score and AUC of the proposed method reach 91.500%, 92.780%, 91.370%, and 91.500%, respectively. The experimental results show that the proposed method can provide an effective reference for breast cancer diagnosis.

show abstract

Classification of imbalanced oral cancer image data from high-risk population

Cited by 17 publications

References 22 publications

AI-based analysis of oral lesions using novel deep convolutional neural networks for early detection of oral cancer

AI-based analysis of oral lesions using novel deep convolutional neural networks for early detection of oral cancer

Mobile Health (mHealth) Technology in Early Detection and Diagnosis of Oral Cancer-A Scoping Review of the Current Scenario and Feasibility

A novel breast cancer image classification model based on multiscale texture feature analysis and dynamic learning

Contact Info

Product

Resources

About