Deep Learning of Histopathological Features for the Prediction of Tumour Molecular Genetics

Development and deployment of a histopathology-based deep learning algorithm for patient prescreening in a clinical trial

Accurate identification of genetic alterations in tumors, such as Fibroblast Growth Factor Receptor, is crucial for treating with targeted therapies; however, molecular testing can delay patient care due to the time and tissue required. Successful development, validation, and deployment of an AI-based, biomarker-detection algorithm could reduce screening cost and accelerate patient recruitment. Here, we develop a deep-learning algorithm using >3000 H&E-stained whole slide images from patients with advanced urothelial cancers, optimized for high sensitivity to avoid ruling out trial-eligible patients. The algorithm is validated on a dataset of 350 patients, achieving an area under the curve of 0.75, specificity of 31.8% at 88.7% sensitivity, and projected 28.7% reduction in molecular testing. We successfully deploy the system in a non-interventional study comprising 89 global study clinical sites and demonstrate its potential to prioritize/deprioritize molecular testing resources and provide substantial cost savings in the drug development and clinical settings.

Surrogate Biomarker Prediction from Whole-Slide Images for Evaluating Overall Survival in Lung Adenocarcinoma

BACKGROUND

Recent advances in computational pathology have shown potential in predicting biomarkers from haematoxylin and eosin (H&E) whole-slide images (WSI). However, predicting the outcome directly from WSIs remains a substantial challenge. In this study, we aimed to investigate how gene expression, predicted from WSIs, could be used to evaluate overall survival (OS) in patients with lung adenocarcinoma (LUAD).

METHODS

Differentially expressed genes (DEGs) were identified from The Cancer Genome Atlas (TCGA)-LUAD cohort. Cox regression analysis was performed on DEGs to identify the gene prognostics of OS. Attention-based multiple instance learning (AMIL) models were trained to predict the expression of identified prognostic genes from WSIs using the TCGA-LUAD dataset. Models were externally validated in the Clinical Proteomic Tumour Analysis Consortium (CPTAC)-LUAD dataset. The prognostic value of predicted gene expression values was then compared to the true gene expression measurements.

RESULTS

The expression of 239 prognostic genes could be predicted in TCGA-LUAD with cross-validated Pearson's R > 0.4. Predicted gene expression demonstrated prognostic performance, attaining a cross-validated concordance index of up to 0.615 in TCGA-LUAD through Cox regression. In total, 36 genes had predicted expression in the external validation cohort that was prognostic of OS.

CONCLUSIONS

Gene expression predicted from WSIs is an effective method of evaluating OS in patients with LUAD. These results may open up new avenues of cost- and time-efficient prognosis assessment in LUAD treatment.

HistoMIL: A Python package for training multiple instance learning models on histopathology slides

Hematoxylin and eosin (H&E) stained slides are widely used in disease diagnosis. Remarkable advances in deep learning have made it possible to detect complex molecular patterns in these histopathology slides, suggesting automated approaches could help inform pathologists' decisions. Multiple instance learning (MIL) algorithms have shown promise in this context, outperforming transfer learning (TL) methods for various tasks, but their implementation and usage remains complex. We introduce HistoMIL, a Python package designed to streamline the implementation, training and inference process of MIL-based algorithms for computational pathologists and biomedical researchers. It integrates a self-supervised learning module for feature encoding, and a full pipeline encompassing TL and three MIL algorithms: ABMIL, DSMIL, and TransMIL. The PyTorch Lightning framework enables effortless customization and algorithm implementation. We illustrate HistoMIL's capabilities by building predictive models for 2,487 cancer hallmark genes on breast cancer histology slides, achieving AUROC performances of up to 85%.

Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images

Image-based AI has thrived as a potentially revolutionary tool for predicting molecular biomarker statuses, which aids in categorizing patients for appropriate medical treatments. However, many methods using hematoxylin and eosin-stained (H&E) whole-slide images (WSIs) have been found to be inefficient because of the presence of numerous uninformative or irrelevant image patches. In this study, we introduced the region of biomarker relevance (ROB) concept to identify the morphological areas most closely associated with biomarkers for accurate status prediction. We actualized this concept within a framework called saliency ROB search (SRS) to enable efficient and effective predictions. By evaluating various lung adenocarcinoma (LUAD) biomarkers, we showcased the superior performance of SRS compared to current state-of-the-art AI approaches. These findings suggest that AI tools, built on the ROB concept, can achieve enhanced molecular biomarker prediction accuracy from pathological images.

Preliminary evaluation of deep learning for first-line diagnostic prediction of tumor mutational status

The detection of tumour gene mutations by DNA or RNA sequencing is crucial for the prescription of effective targeted therapies. Recent developments showed promising results for tumoral mutational status prediction using new deep learning based methods on histopathological images. However, it is still unknown whether these methods can be useful aside from sequencing methods for efficient population diagnosis. In this retrospective study, we use a standard prediction pipeline based on a convolutional neural network for the detection of cancer driver genomic alterations in The Cancer Genome Atlas (TCGA) breast (BRCA, n = 719), lung (LUAD, n = 541) and colon (COAD, n = 459) cancer datasets. We propose 3 diagnostic strategies using deep learning methods as first-line diagnostic tools. Focusing on cancer driver genes such as KRAS, EGFR or TP53, we show that these methods help reduce DNA sequencing by up to 49.9% with a high sensitivity (95%). In a context of limited resources, these methods increase sensitivity up to 69.8% at a 30% capacity of DNA sequencing tests, up to 85.1% at a 50% capacity, and up to 91.8% at a 70% capacity. These methods can also be used to prioritize patients with a positive predictive value up to 90.6% in the 10% patient most at risk of being mutated. Limitations of this study include the lack of external validation on non-TCGA data, dependence on prevalence of mutations in datasets, and use of a standard DL method on a limited dataset. Future studies using state-of-the-art methods and larger datasets are needed for better evaluation and clinical implementation.

Multimodal deep learning to predict prognosis in adult and pediatric brain tumors

BACKGROUND

The introduction of deep learning in both imaging and genomics has significantly advanced the analysis of biomedical data. For complex diseases such as cancer, different data modalities may reveal different disease characteristics, and the integration of imaging with genomic data has the potential to unravel additional information than when using these data sources in isolation. Here, we propose a DL framework that combines these two modalities with the aim to predict brain tumor prognosis.

METHODS

Using two separate glioma cohorts of 783 adults and 305 pediatric patients we developed a DL framework that can fuse histopathology images with gene expression profiles. Three strategies for data fusion were implemented and compared: early, late, and joint fusion. Additional validation of the adult glioma models was done on an independent cohort of 97 adult patients.

RESULTS

Here we show that the developed multimodal data models achieve better prediction results compared to the single data models, but also lead to the identification of more relevant biological pathways. When testing our adult models on a third brain tumor dataset, we show our multimodal framework is able to generalize and performs better on new data from different cohorts. Leveraging the concept of transfer learning, we demonstrate how our pediatric multimodal models can be used to predict prognosis for two more rare (less available samples) pediatric brain tumors.

CONCLUSIONS

Our study illustrates that a multimodal data fusion approach can be successfully implemented and customized to model clinical outcome of adult and pediatric brain tumors.

Artificial intelligence for multimodal data integration in oncology

In oncology, the patient state is characterized by a whole spectrum of modalities, ranging from radiology, histology, and genomics to electronic health records. Current artificial intelligence (AI) models operate mainly in the realm of a single modality, neglecting the broader clinical context, which inevitably diminishes their potential. Integration of different data modalities provides opportunities to increase robustness and accuracy of diagnostic and prognostic models, bringing AI closer to clinical practice. AI models are also capable of discovering novel patterns within and across modalities suitable for explaining differences in patient outcomes or treatment resistance. The insights gleaned from such models can guide exploration studies and contribute to the discovery of novel biomarkers and therapeutic targets. To support these advances, here we present a synopsis of AI methods and strategies for multimodal data fusion and association discovery. We outline approaches for AI interpretability and directions for AI-driven exploration through multimodal data interconnections. We examine challenges in clinical adoption and discuss emerging solutions.

Deep learning-based prediction of molecular cancer biomarkers from tissue slides: A new tool for precision oncology

Molecular tests are necessary to stratify cancer patients for targeted therapy. However, high cost and technical barriers limit the application of these tests, hindering optimal treatment. Recently, deep learning (DL) has been applied to predict molecular test results from digitized images of tissue slides. Furthermore, treatment response and prognosis can be predicted from tissue slides using DL. In this review, we summarized DL-based studies regarding the prediction of genetic mutation, microsatellite instability, tumor mutational burden, molecular subtypes, gene expression, treatment response, and prognosis directly from hematoxylin- and eosin-stained tissue slides. Although performance needs to be improved, these studies clearly demonstrated the feasibility of DL-based prediction of key molecular features in cancer tissues. With the accumulation of data and technical advances, the performance of the DL system could be improved in the near future. Therefore, we expect that DL could provide cost- and time-effective alternative tools for patient stratification in the era of precision oncology.

Validating Automatic Concept-Based Explanations for AI-Based Digital Histopathology

Digital histopathology poses several challenges such as label noise, class imbalance, limited availability of labelled data, and several latent biases to deep learning, negatively influencing transparency, reproducibility, and classification performance. In particular, biases are well known to cause poor generalization. Proposed tools from explainable artificial intelligence (XAI), bias detection, and bias discovery suffer from technical challenges, complexity, unintuitive usage, inherent biases, or a semantic gap. A promising XAI method, not studied in the context of digital histopathology is automated concept-based explanation (ACE). It automatically extracts visual concepts from image data. Our objective is to evaluate ACE's technical validity following design science principals and to compare it to Guided Gradient-weighted Class Activation Mapping (Grad-CAM), a conventional pixel-wise explanation method. To that extent, we created and studied five convolutional neural networks (CNNs) in four different skin cancer settings. Our results demonstrate that ACE is a valid tool for gaining insights into the decision process of histopathological CNNs that can go beyond explanations from the control method. ACE validly visualized a class sampling ratio bias, measurement bias, sampling bias, and class-correlated bias. Furthermore, the complementary use with Guided Grad-CAM offers several benefits. Finally, we propose practical solutions for several technical challenges. In contradiction to results from the literature, we noticed lower intuitiveness in some dermatopathology scenarios as compared to concept-based explanations on real-world images.

Artificial Intelligence for Predicting Microsatellite Instability Based on Tumor Histomorphology: A Systematic Review

Microsatellite instability (MSI)/defective DNA mismatch repair (dMMR) is receiving more attention as a biomarker for eligibility for immune checkpoint inhibitors in advanced diseases. However, due to high costs and resource limitations, MSI/dMMR testing is not widely performed. Some attempts are in progress to predict MSI/dMMR status through histomorphological features on H&E slides using artificial intelligence (AI) technology. In this study, the potential predictive role of this new methodology was reviewed through a systematic review. Studies up to September 2021 were searched through PubMed and Embase database searches. The design and results of each study were summarized, and the risk of bias for each study was evaluated. For colorectal cancer, AI-based systems showed excellent performance with the highest standard of 0.972; for gastric and endometrial cancers they showed a relatively low but satisfactory performance, with the highest standard of 0.81 and 0.82, respectively. However, analyzing the risk of bias, most studies were evaluated at high-risk. AI-based systems showed a high potential in predicting the MSI/dMMR status of different cancer types, and particularly of colorectal cancers. Therefore, a confirmation test should be required only for the results that are positive in the AI test.

Estimation with Uncertainty via Conditional Generative Adversarial Networks

Conventional predictive Artificial Neural Networks (ANNs) commonly employ deterministic weight matrices; therefore, their prediction is a point estimate. Such a deterministic nature in ANNs causes the limitations of using ANNs for medical diagnosis, law problems, and portfolio management in which not only discovering the prediction but also the uncertainty of the prediction is essentially required. In order to address such a problem, we propose a predictive probabilistic neural network model, which corresponds to a different manner of using the generator in the conditional Generative Adversarial Network (cGAN) that has been routinely used for conditional sample generation. By reversing the input and output of ordinary cGAN, the model can be successfully used as a predictive model; moreover, the model is robust against noises since adversarial training is employed. In addition, to measure the uncertainty of predictions, we introduce the entropy and relative entropy for regression problems and classification problems, respectively. The proposed framework is applied to stock market data and an image classification task. As a result, the proposed framework shows superior estimation performance, especially on noisy data; moreover, it is demonstrated that the proposed framework can properly estimate the uncertainty of predictions.