Corrigendum: Deep learning for the detection of anatomical tissue structures and neoplasms of the skin on scanned histopathological tissue sections

[This corrects the article DOI: 10.3389/fonc.2022.1022967.].

Deep learning for the detection of anatomical tissue structures and neoplasms of the skin on scanned histopathological tissue sections

Basal cell carcinoma (BCC), squamous cell carcinoma (SqCC) and melanoma are among the most common cancer types. Correct diagnosis based on histological evaluation after biopsy or excision is paramount for adequate therapy stratification. Deep learning on histological slides has been suggested to complement and improve routine diagnostics, but publicly available curated and annotated data and usable models trained to distinguish common skin tumors are rare and often lack heterogeneous non-tumor categories. A total of 16 classes from 386 cases were manually annotated on scanned histological slides, 129,364 100 x 100 µm (~395 x 395 px) image tiles were extracted and split into a training, validation and test set. An EfficientV2 neuronal network was trained and optimized to classify image categories. Cross entropy loss, balanced accuracy and Matthews correlation coefficient were used for model evaluation. Image and patient data were assessed with confusion matrices. Application of the model to an external set of whole slides facilitated localization of melanoma and non-tumor tissue. Automated differentiation of BCC, SqCC, melanoma, naevi and non-tumor tissue structures was possible, and a high diagnostic accuracy was achieved in the validation (98%) and test (97%) set. In summary, we provide a curated dataset including the most common neoplasms of the skin and various anatomical compartments to enable researchers to train, validate and improve deep learning models. Automated classification of skin tumors by deep learning techniques is possible with high accuracy, facilitates tumor localization and has the potential to support and improve routine diagnostics.

Drug-protein interaction prediction via variational autoencoders and attention mechanisms

During the process of drug discovery, exploring drug-protein interactions (DPIs) is a key step. With the rapid development of biological data, computer-aided methods are much faster than biological experiments. Deep learning methods have become popular and are mainly used to extract the characteristics of drugs and proteins for further DPIs prediction. Since the prediction of DPIs through machine learning cannot fully extract effective features, in our work, we propose a deep learning framework that uses variational autoencoders and attention mechanisms; it utilizes convolutional neural networks (CNNs) to obtain local features and attention mechanisms to obtain important information about drugs and proteins, which is very important for predicting DPIs. Compared with some machine learning methods on the C.elegans and human datasets, our approach provides a better effect. On the BindingDB dataset, its accuracy (ACC) and area under the curve (AUC) reach 0.862 and 0.913, respectively. To verify the robustness of the model, multiclass classification tasks are performed on Davis and KIBA datasets, and the ACC values reach 0.850 and 0.841, respectively, thus further demonstrating the effectiveness of the model.

A deep learning based multimodal fusion model for skin lesion diagnosis using smartphone collected clinical images and metadata

Introduction

Skin cancer is one of the most common types of cancer. An accessible tool to the public can help screening for malign lesion. We aimed to develop a deep learning model to classify skin lesion using clinical images and meta information collected from smartphones.

Methods

A deep neural network was developed with two encoders for extracting information from image data and metadata. A multimodal fusion module with intra-modality self-attention and inter-modality cross-attention was proposed to effectively combine image features and meta features. The model was trained on tested on a public dataset and compared with other state-of-the-art methods using five-fold cross-validation.

Results

Including metadata is shown to significantly improve a model's performance. Our model outperformed other metadata fusion methods in terms of accuracy, balanced accuracy and area under the receiver-operating characteristic curve, with an averaged value of 0.768±0.022, 0.775±0.022 and 0.947±0.007.

Conclusion

A deep learning model using smartphone collected images and metadata for skin lesion diagnosis was successfully developed. The proposed model showed promising performance and could be a potential tool for skin cancer screening.

Identification of Snails and Schistosoma of Medical Importance via Convolutional Neural Networks: A Proof-of-Concept Application for Human Schistosomiasis

In recent decades, computer vision has proven remarkably effective in addressing diverse issues in public health, from determining the diagnosis, prognosis, and treatment of diseases in humans to predicting infectious disease outbreaks. Here, we investigate whether convolutional neural networks (CNNs) can also demonstrate effectiveness in classifying the environmental stages of parasites of public health importance and their invertebrate hosts. We used schistosomiasis as a reference model. Schistosomiasis is a debilitating parasitic disease transmitted to humans via snail intermediate hosts. The parasite affects more than 200 million people in tropical and subtropical regions. We trained our CNN, a feed-forward neural network, on a limited dataset of 5,500 images of snails and 5,100 images of cercariae obtained from schistosomiasis transmission sites in the Senegal River Basin, a region in western Africa that is hyper-endemic for the disease. The image set included both images of two snail genera that are relevant to schistosomiasis transmission – that is, Bulinus spp. and Biomphalaria pfeifferi – as well as snail images that are non-component hosts for human schistosomiasis. Cercariae shed from Bi. pfeifferi and Bulinus spp. snails were classified into 11 categories, of which only two, S. haematobium and S. mansoni, are major etiological agents of human schistosomiasis. The algorithms, trained on 80% of the snail and parasite dataset, achieved 99% and 91% accuracy for snail and parasite classification, respectively, when used on the hold-out validation dataset – a performance comparable to that of experienced parasitologists. The promising results of this proof-of-concept study suggests that this CNN model, and potentially similar replicable models, have the potential to support the classification of snails and parasite of medical importance. In remote field settings where machine learning algorithms can be deployed on cost-effective and widely used mobile devices, such as smartphones, these models can be a valuable complement to laboratory identification by trained technicians. Future efforts must be dedicated to increasing dataset sizes for model training and validation, as well as testing these algorithms in diverse transmission settings and geographies.

BENDR: Using Transformers and a Contrastive Self-Supervised Learning Task to Learn From Massive Amounts of EEG Data

Deep neural networks (DNNs) used for brain-computer interface (BCI) classification are commonly expected to learn general features when trained across a variety of contexts, such that these features could be fine-tuned to specific contexts. While some success is found in such an approach, we suggest that this interpretation is limited and an alternative would better leverage the newly (publicly) available massive electroencephalography (EEG) datasets. We consider how to adapt techniques and architectures used for language modeling (LM) that appear capable of ingesting awesome amounts of data toward the development of encephalography modeling with DNNs in the same vein. We specifically adapt an approach effectively used for automatic speech recognition, which similarly (to LMs) uses a self-supervised training objective to learn compressed representations of raw data signals. After adaptation to EEG, we find that a single pre-trained model is capable of modeling completely novel raw EEG sequences recorded with differing hardware, and different subjects performing different tasks. Furthermore, both the internal representations of this model and the entire architecture can be fine-tuned to a variety of downstream BCI and EEG classification tasks, outperforming prior work in more task-specific (sleep stage classification) self-supervision.