Dermatologist-Level Classification of Skin Cancer Using Cascaded Ensembling of Convolutional Neural Network and Handcrafted Features Based Deep Neural Network

By A. Sharma, Shamik Tiwari, Gaurav Aggarwal, Nitika Goenka, Anil Kumar, P. Chakrabarti, T. Chakrabarti, R. Goňo, Zbigniew Leonowicz, M. Jasinski. 2022

332 citations recorded by the source corpus.

Skin cancer is caused due to unusual development of skin cells and deadly type cancer. Early diagnosis is very significant and can avoid some categories of skin cancers, such as melanoma and focal cell carcinoma. The recognition and the classification of skin malignant growth in the beginning time is expensive and challenging. The deep learning architectures such as recurrent networks and convolutional neural networks (ConvNets) are developed in the past, which are proven appropriate for non-handcrafted extraction of complex features. To additional expand the efficiency of the ConvNet models, a cascaded ensembled network that uses an integration of ConvNet and handcrafted features based multi-layer perceptron is proposed in this work. This offered model utilizes the convolutional neural network model to mine non-handcrafted image features and colour moments and texture features as handcrafted features. It is demonstrated that accuracy of ensembled deep learning model is improved to 98.3% from 85.3% of convolutional neural network model.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Epidemiology, Oncology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Epidemiology

Citations in the corpus

Contrastive Learning of Medical Visual Representations from Paired Images and Text (2020, methodology): “Medical image understanding has the potential to transform healthcare and has seen rapid progress with deep learning (Gulshan et al., 2016; Esteva et al., 2017; De Fauw et al., 2018; Rajpurkar et al., 2018b).”
Med-BERT: pretrained contextualized embeddings on large-scale structured electronic health records for disease prediction (2020, background): “Given enough training samples, deep-learning models can achieve comparable or even better performance than domain experts in the diagnosis of certain diseases [15-19].”
Adversary Instantiation: Lower Bounds for Differentially Private Machine Learning (2021, methodology): “With machine learning now being used to train models on sensitive user data, ranging from medical images [19], to personal email [8] and text messages [53], it is becoming ever more important that these models are privacy preserving.”
Accelerating Federated Learning via Momentum Gradient Descent (2019, background): “Recently, data-intensive machine learning has been applied in various fields, such as autonomous driving [1], speech recognition [2], image classification [3] and disease detection [4] since this technique provides beneficial solutions to extract the useful information hidden in data.”
Making the Most of Text Semantics to Improve Biomedical Vision-Language Processing (2022, background): “Advances in deep learning have enabled automated diagnosis systems that operate near or above expert-level performance, paving the way for the use of machine learning systems to improve healthcare workflows, for example by supporting fast triaging and assisting medical professionals to reduce errors”
A Survey of Deep Learning for Scientific Discovery (2020, background): “pathology [135, 55], analyzing brain scans (PET, fMRI) [202, 45].”
Deterministic Neural Networks with Appropriate Inductive Biases Capture Epistemic and Aleatoric Uncertainty (2021, unknown): “This matters in particular for noisy and ambiguous datasets as might be found in safety-critical applications like autonomous driving [Huang and Chen, 2020] and medical diagnosis [Esteva et al., 2017, Filos et al., 2019].”
Deep Learning for Medical Anomaly Detection – A Survey (2020, methodology): “Example architectures include the CNN structures in References [33, 112], where they employ supervised learning to identify anomalies in retina images and for automated classification of skin lesions, respectively.”