Deep learning of quantitative ultrasound multi-parametric images at pre-treatment to predict breast cancer response to chemotherapy

Deep learning analysis of serial digital breast tomosynthesis images in a prospective cohort of breast cancer patients who received neoadjuvant chemotherapy

PURPOSE

Different imaging tools, including digital breast tomosynthesis (DBT), are frequently used for evaluating tumor response during neoadjuvant chemotherapy (NACT). This study aimed to explore whether using artificial intelligence (AI) for serial DBT acquisitions during NACT for breast cancer can predict pathological complete response (pCR) after completion of NACT.

METHODS

A total of 149 women (mean age 53 years, pCR rate 22 %) with breast cancer treated with NACT at Skane University Hospital, Sweden, between 2014 and 2019, were prospectively included in this observational cohort study (ClinicalTrials.gov: NCT02306096). DBT images from both the cancer and contralateral healthy breasts acquired at three time points: pre-NACT, after two cycles of NACT, and after the completion of six cycles of NACT (pre-surgery) were analyzed. The deep learning AI system used to predict pCR consisted of a backbone 3D ResNet and an attention and prediction module. The GradCAM method was used to obtain insights into the model decision basis through a quantitative analysis of the importance maps on the validation set. Moreover, specific model choices were motivated by ablation studies.

RESULTS

The AI model reached an AUC of 0.83 (95% CI: 0.63-1.00) (test set). The spatial correlation of importance maps for input volumes from the same patient but at different time points was high, possibly indicating that the model focuses on the same areas during decision-making.

CONCLUSIONS

We demonstrate a high discriminative performance of our algorithm for predicting pCR/non-pCR. Availability of larger datasets would permit more comprehensive training of the models and more rigorous evaluation of their prediction performance for future patients.

Deep learning based ultrasound analysis facilitates precise distinction between parotid pleomorphic adenoma and Warthin tumor

Background

Pleomorphic adenoma (PA), often with the benign-like imaging appearances similar to Warthin tumor (WT), however, is a potentially malignant tumor with a high recurrence rate. It is worse that pathological fine-needle aspiration cytology (FNAC) is difficult to distinguish PA and WT for inexperienced pathologists. This study employed deep learning (DL) technology, which effectively utilized ultrasound images, to provide a reliable approach for discriminating PA from WT.

Methods

488 surgically confirmed patients, including 266 with PA and 222 with WT, were enrolled in this study. Two experienced ultrasound physicians independently evaluated all images to differentiate between PA and WT. The diagnostic performance of preoperative FNAC was also evaluated. During the DL study, all ultrasound images were randomly divided into training (70%), validation (20%), and test (10%) sets. Furthermore, ultrasound images that could not be diagnosed by FNAC were also randomly allocated to training (60%), validation (20%), and test (20%) sets. Five DL models were developed to classify ultrasound images as PA or WT. The robustness of these models was assessed using five-fold cross-validation. The Gradient-weighted Class Activation Mapping (Grad-CAM) technique was employed to visualize the region of interest in the DL models.

Results

In Grad-CAM analysis, the DL models accurately identified the mass as the region of interest. The area under the receiver operating characteristic curve (AUROC) of the two ultrasound physicians were 0.351 and 0.598, and FNAC achieved an AUROC of only 0.721. Meanwhile, for DL models, the AUROC value for discriminating between PA and WT in the test set was from 0.828 to 0.908. ResNet50 demonstrated the optimal performance with an AUROC of 0.908, an accuracy of 0.833, a sensitivity of 0.736, and a specificity of 0.904. In the test set of cases that FNAC failed to provide a diagnosis, DenseNet121 demonstrated the optimal performance with an AUROC of 0.897, an accuracy of 0.806, a sensitivity of 0.789, and a specificity of 0.824.

Conclusion

For the discrimination of PA and WT, DL models are superior to ultrasound and FNAC, thereby facilitating surgeons in making informed decisions regarding the most appropriate surgical approach.

Fusion of transfer learning models with LSTM for detection of breast cancer using ultrasound images

Breast Cancer (BC) is one of the top reasons for fatality in women worldwide. As a result, timely identification is critical for successful therapy and excellent survival rates. Transfer Learning (TL) approaches have recently shown promise in aiding in the early recognition of BC. In this work, three TL models, MobileNetV2, ResNet50, and VGG16, were combined with LSTM to extract the features from Ultrasound Images (USIs). Furthermore, the Synthetic Minority Over-sampling Technique (SMOTE) with Tomek (SMOTETomek) was employed to balance the extracted features. The proposed method with VGG16 achieved an F1 score of 99.0 %, Matthews Correlation Coefficient (MCC) and Kappa Coefficient of 98.9 % with an Area Under Curve (AUC) of 1.0. The K-fold method was applied for cross-validation and achieved an average F1 score of 96 %. Moreover, the Gradient-weighted Class Activation Mapping (Grad-CAM) method was applied for visualization, and the Local Interpretable Model-agnostic Explanations (LIME) method was applied for interpretability. The Normal Approximation Interval (NAI) and bootstrapping methods were used to calculate Confidence Intervals (CIs). The proposed method achieved a Lower CI (LCI), Upper CI (UCI), and Mean CI (MCI) of 96.50 %, 99.75 %, and 98.13 %, respectively, with the NAI, while 95 % LCI of 93.81 %, an UCI of 96.00 %, and a bootstrap mean of 94.90 % with the bootstrap method. Furthermore, the performance of the six state-of-the-art (SOTA) TL models, such as Xception, NASNetMobile, InceptionResNetV2, MobileNetV2, ResNet50, and VGG16, were compared with the proposed method.

Transfer learning of pre-treatment quantitative ultrasound multi-parametric images for the prediction of breast cancer response to neoadjuvant chemotherapy

Locally advanced breast cancer (LABC) is a severe type of cancer with a poor prognosis, despite advancements in therapy. As the disease is often inoperable, current guidelines suggest upfront aggressive neoadjuvant chemotherapy (NAC). Complete pathological response to chemotherapy is linked to improved survival, but conventional clinical assessments like physical exams, mammography, and imaging are limited in detecting early response. Early detection of tissue response can improve complete pathological response and patient survival while reducing exposure to ineffective and potentially harmful treatments. A rapid, cost-effective modality without the need for exogenous contrast agents would be valuable for evaluating neoadjuvant therapy response. Conventional ultrasound provides information about tissue echogenicity, but image comparisons are difficult due to instrument-dependent settings and imaging parameters. Quantitative ultrasound (QUS) overcomes this by using normalized power spectra to calculate quantitative metrics. This study used a novel transfer learning-based approach to predict LABC response to neoadjuvant chemotherapy using QUS imaging at pre-treatment. Using data from 174 patients, QUS parametric images of breast tumors with margins were generated. The ground truth response to therapy for each patient was based on standard clinical and pathological criteria. The Residual Network (ResNet) deep learning architecture was used to extract features from the parametric QUS maps. This was followed by SelectKBest and Synthetic Minority Oversampling (SMOTE) techniques for feature selection and data balancing, respectively. The Support Vector Machine (SVM) algorithm was employed to classify patients into two distinct categories: nonresponders (NR) and responders (RR). Evaluation results on an unseen test set demonstrate that the transfer learning-based approach using spectral slope parametric maps had the best performance in the identification of nonresponders with precision, recall, F1-score, and balanced accuracy of 100, 71, 83, and 86%, respectively. The transfer learning-based approach has many advantages over conventional deep learning methods since it reduces the need for large image datasets for training and shortens the training time. The results of this study demonstrate the potential of transfer learning in predicting LABC response to neoadjuvant chemotherapy before the start of treatment using quantitative ultrasound imaging. Prediction of NAC response before treatment can aid clinicians in customizing ineffectual treatment regimens for individual patients.

Near-Infrared Spectral Tomography for Predicting Residual Cancer Burden during Early-Stage Neoadjuvant Chemotherapy for Breast Cancer

PURPOSE

The aim of this study is to investigate whether near-infrared spectral tomography (NIRST) might serve as a reliable prognostic tool to predict residual cancer burden (RCB) in patients with breast cancer undergoing neoadjuvant chemotherapy (NAC) based upon early treatment response measurements.

EXPERIMENTAL DESIGN

A total of thirty-five patients with breast cancer receiving NAC were included in this study. NIRST imaging was performed at multiple time points, including: before treatment, at end of the first cycle, at the mid-point, and post-NAC treatments. From reconstructed NIRST images, average values of total hemoglobin (HbT) were obtained for both the tumor region and contralateral breast at each time point. RCB scores/classes were assessed by a pathologist using histologic slides of the surgical specimen obtained after completing NAC. Logistic regression of the normalized early percentage change of HbT in the tumor region (ΔHbT%) was used to predict RCB and determine its significance as an indicator for differentiating cases within each RCB class.

RESULTS

The ΔHbT% at the end of the first cycle, as compared with pretreatment levels, showed excellent prognostic capability in differentiating RCB-0 from RCB-I/II/III or RCB-II from RCB-0/I/III (P < 0.001). Corresponding area under the curve (AUC) values for these comparisons were 0.97 and 0.94, and accuracy values were 0.90 and 0.83, respectively.

CONCLUSIONS

NIRST holds promise as a potential clinical tool that can be seamlessly integrated into existing clinical workflow within the infusion suite. By providing early assessment of RCB, NIRST has potential to improve breast cancer patient management strategies.

Exploring Neoadjuvant Chemotherapy, Predictive Models, Radiomic, and Pathological Markers in Breast Cancer: A Comprehensive Review

Breast cancer retains its position as the most prevalent form of malignancy among females on a global scale. The careful selection of appropriate treatment for each patient holds paramount importance in effectively managing breast cancer. Neoadjuvant chemotherapy (NACT) plays a pivotal role in the comprehensive treatment of this disease. Administering chemotherapy before surgery, NACT becomes a powerful tool in reducing tumor size, potentially enabling fewer invasive surgical procedures and even rendering initially inoperable tumors amenable to surgery. However, a significant challenge lies in the varying responses exhibited by different patients towards NACT. To address this challenge, researchers have focused on developing prediction models that can identify those who would benefit from NACT and those who would not. Such models have the potential to reduce treatment costs and contribute to a more efficient and accurate management of breast cancer. Therefore, this review has two objectives: first, to identify the most effective radiomic markers correlated with NACT response, and second, to explore whether integrating radiomic markers extracted from radiological images with pathological markers can enhance the predictive accuracy of NACT response. This review will delve into addressing these research questions and also shed light on the emerging research direction of leveraging artificial intelligence techniques for predicting NACT response, thereby shaping the future landscape of breast cancer treatment.

Predicting head and neck cancer treatment outcomes with pre-treatment quantitative ultrasound texture features and optimising machine learning classifiers with texture-of-texture features

Aim

Cancer treatments with radiation present a challenging physical toll for patients, which can be justified by the potential reduction in cancerous tissue with treatment. However, there remain patients for whom treatments do not yield desired outcomes. Radiomics involves using biomedical images to determine imaging features which, when used in tandem with retrospective treatment outcomes, can train machine learning (ML) classifiers to create predictive models. In this study we investigated whether pre-treatment imaging features from index lymph node (LN) quantitative ultrasound (QUS) scans parametric maps of head & neck (H&N) cancer patients can provide predictive information about treatment outcomes.

Methods

72 H&N cancer patients with bulky metastatic LN involvement were recruited for study. Involved bulky neck nodes were scanned with ultrasound prior to the start of treatment for each patient. QUS parametric maps and related radiomics texture-based features were determined and used to train two ML classifiers (support vector machines (SVM) and k-nearest neighbour (k-NN)) for predictive modeling using retrospectively labelled binary treatment outcomes, as determined clinically 3-months after completion of treatment. Additionally, novel higher-order texture-of-texture (TOT) features were incorporated and evaluated in regards to improved predictive model performance.

Results

It was found that a 7-feature multivariable model of QUS texture features using a support vector machine (SVM) classifier demonstrated 81% sensitivity, 76% specificity, 79% accuracy, 86% precision and an area under the curve (AUC) of 0.82 in separating responding from non-responding patients. All performance metrics improved after implementation of TOT features to 85% sensitivity, 80% specificity, 83% accuracy, 89% precision and AUC of 0.85. Similar trends were found with k-NN classifier.

Conclusion

Binary H&N cancer treatment outcomes can be predicted with QUS texture features acquired from index LNs. Prediction efficacy improved by implementing TOT features following methodology outlined in this work.

Deep Learning Model Based on Dual-Modal Ultrasound and Molecular Data for Predicting Response to Neoadjuvant Chemotherapy in Breast Cancer

RATIONALE AND OBJECTIVES

To carry out radiomics analysis/deep convolutional neural network (CNN) based on B-mode ultrasound (BUS) and shear wave elastography (SWE) to predict response to neoadjuvant chemotherapy (NAC) in breast cancer patients.

MATERIALS AND METHODS

In this prospective study, 255 breast cancer patients who received NAC between September 2016 and December 2021 were included. Radiomics models were designed using a support vector machine classifier based on US images obtained before treatment, including BUS and SWE. And CNN models also were developed using ResNet architecture. The final predictive model was developed by combining the dual-modal US and independently associated clinicopathologic characteristics. The predictive performances of the models were assessed with five-fold cross-validation.

RESULTS

Pretreatment SWE performed better than BUS in predicting the response to NAC for breast cancer for both the CNN and radiomics models (P < 0.001). The predictive results of the CNN models were significantly better than the radiomics models, with AUCs of 0.72 versus 0.69 for BUS and 0.80 versus 0.77 for SWE, respectively (P = 0.003). The CNN model based on the dual-modal US and molecular data exhibited outstanding performance in predicting NAC response, with an accuracy of 83.60% ± 2.63%, a sensitivity of 87.76% ± 6.44%, and a specificity of 77.45% ± 4.38%.

CONCLUSION

The pretreatment CNN model based on the dual-modal US and molecular data achieved excellent performance for predicting the response to chemotherapy in breast cancer. Therefore, this model has the potential to serve as a non-invasive objective biomarker to predict NAC response and aid clinicians with individual treatments.

A review on deep learning applications in highly multiplexed tissue imaging data analysis

Since its introduction into the field of oncology, deep learning (DL) has impacted clinical discoveries and biomarker predictions. DL-driven discoveries and predictions in oncology are based on a variety of biological data such as genomics, proteomics, and imaging data. DL-based computational frameworks can predict genetic variant effects on gene expression, as well as protein structures based on amino acid sequences. Furthermore, DL algorithms can capture valuable mechanistic biological information from several spatial "omics" technologies, such as spatial transcriptomics and spatial proteomics. Here, we review the impact that the combination of artificial intelligence (AI) with spatial omics technologies has had on oncology, focusing on DL and its applications in biomedical image analysis, encompassing cell segmentation, cell phenotype identification, cancer prognostication, and therapy prediction. We highlight the advantages of using highly multiplexed images (spatial proteomics data) compared to single-stained, conventional histopathological ("simple") images, as the former can provide deep mechanistic insights that cannot be obtained by the latter, even with the aid of explainable AI. Furthermore, we provide the reader with the advantages/disadvantages of DL-based pipelines used in preprocessing highly multiplexed images (cell segmentation, cell type annotation). Therefore, this review also guides the reader to choose the DL-based pipeline that best fits their data. In conclusion, DL continues to be established as an essential tool in discovering novel biological mechanisms when combined with technologies such as highly multiplexed tissue imaging data. In balance with conventional medical data, its role in clinical routine will become more important, supporting diagnosis and prognosis in oncology, enhancing clinical decision-making, and improving the quality of care for patients. Since its introduction into the field of oncology, deep learning (DL) has impacted clinical discoveries and biomarker predictions. DL-driven discoveries and predictions in oncology are based on a variety of biological data such as genomics, proteomics, and imaging data. DL-based computational frameworks can predict genetic variant effects on gene expression, as well as protein structures based on amino acid sequences. Furthermore, DL algorithms can capture valuable mechanistic biological information from several spatial "omics" technologies, such as spatial transcriptomics and spatial proteomics. Here, we review the impact that the combination of artificial intelligence (AI) with spatial omics technologies has had on oncology, focusing on DL and its applications in biomedical image analysis, encompassing cell segmentation, cell phenotype identification, cancer prognostication, and therapy prediction. We highlight the advantages of using highly multiplexed images (spatial proteomics data) compared to single-stained, conventional histopathological ("simple") images, as the former can provide deep mechanistic insights that cannot be obtained by the latter, even with the aid of explainable AI. Furthermore, we provide the reader with the advantages/disadvantages of the DL-based pipelines used in preprocessing the highly multiplexed images (cell segmentation, cell type annotation). Therefore, this review also guides the reader to choose the DL-based pipeline that best fits their data. In conclusion, DL continues to be established as an essential tool in discovering novel biological mechanisms when combined with technologies such as highly multiplexed tissue imaging data. In balance with conventional medical data, its role in clinical routine will become more important, supporting diagnosis and prognosis in oncology, enhancing clinical decision-making, and improving the quality of care for patients.

Deep Learning in Different Ultrasound Methods for Breast Cancer, from Diagnosis to Prognosis: Current Trends, Challenges, and an Analysis

Breast cancer is the second-leading cause of mortality among women around the world. Ultrasound (US) is one of the noninvasive imaging modalities used to diagnose breast lesions and monitor the prognosis of cancer patients. It has the highest sensitivity for diagnosing breast masses, but it shows increased false negativity due to its high operator dependency. Underserved areas do not have sufficient US expertise to diagnose breast lesions, resulting in delayed management of breast lesions. Deep learning neural networks may have the potential to facilitate early decision-making by physicians by rapidly yet accurately diagnosing and monitoring their prognosis. This article reviews the recent research trends on neural networks for breast mass ultrasound, including and beyond diagnosis. We discussed original research recently conducted to analyze which modes of ultrasound and which models have been used for which purposes, and where they show the best performance. Our analysis reveals that lesion classification showed the highest performance compared to those used for other purposes. We also found that fewer studies were performed for prognosis than diagnosis. We also discussed the limitations and future directions of ongoing research on neural networks for breast ultrasound.

Artificial Intelligence for Breast US

US is a widely available, commonly used, and indispensable imaging modality for breast evaluation. It is often the primary imaging modality for the detection and diagnosis of breast cancer in low-resource settings. In addition, it is frequently employed as a supplemental screening tool via either whole breast handheld US or automated breast US among women with dense breasts. In recent years, a variety of artificial intelligence systems have been developed to assist radiologists with the detection and diagnosis of breast lesions on US. This article reviews the background and evidence supporting the use of artificial intelligence tools for breast US, describes implementation strategies and impact on clinical workflow, and discusses potential emerging roles and future directions.

Artificial Intelligence in Breast Ultrasound: From Diagnosis to Prognosis-A Rapid Review

BACKGROUND

Ultrasound (US) is a fundamental diagnostic tool in breast imaging. However, US remains an operator-dependent examination. Research into and the application of artificial intelligence (AI) in breast US are increasing. The aim of this rapid review was to assess the current development of US-based artificial intelligence in the field of breast cancer.

METHODS

Two investigators with experience in medical research performed literature searching and data extraction on PubMed. The studies included in this rapid review evaluated the role of artificial intelligence concerning BC diagnosis, prognosis, molecular subtypes of breast cancer, axillary lymph node status, and the response to neoadjuvant chemotherapy. The mean values of sensitivity, specificity, and AUC were calculated for the main study categories with a meta-analytical approach.

RESULTS

A total of 58 main studies, all published after 2017, were included. Only 9/58 studies were prospective (15.5%); 13/58 studies (22.4%) used an ML approach. The vast majority (77.6%) used DL systems. Most studies were conducted for the diagnosis or classification of BC (55.1%). At present, all the included studies showed that AI has excellent performance in breast cancer diagnosis, prognosis, and treatment strategy.

CONCLUSIONS

US-based AI has great potential and research value in the field of breast cancer diagnosis, treatment, and prognosis. More prospective and multicenter studies are needed to assess the potential impact of AI in breast ultrasound.

Implementation of Non-Invasive Quantitative Ultrasound in Clinical Cancer Imaging

Quantitative ultrasound (QUS) is a non-invasive novel technique that allows treatment response monitoring. Studies have shown that QUS backscatter variables strongly correlate with changes observed microscopically. Increases in cell death result in significant alterations in ultrasound backscatter parameters. In particular, the parameters related to scatterer size and scatterer concentration tend to increase in relation to cell death. The use of QUS in monitoring tumor response has been discussed in several preclinical and clinical studies. Most of the preclinical studies have utilized QUS for evaluating cell death response by differentiating between viable cells and dead cells. In addition, clinical studies have incorporated QUS mostly for tissue characterization, including classifying benign versus malignant breast lesions, as well as responder versus non-responder patients. In this review, we highlight some of the important findings of previous preclinical and clinical studies and expand the applicability and therapeutic benefits of QUS in clinical settings. We summarized some recent clinical research advances in ultrasound-based radiomics analysis for monitoring and predicting treatment response and characterizing benign and malignant breast lesions. We also discuss current challenges, limitations, and future prospects of QUS-radiomics.

Improving breast cancer diagnosis by incorporating raw ultrasound parameters into machine learning

The improved diagnostic accuracy of ultrasound breast examinations remains an important goal. In this study, we propose a biophysical feature-based machine learning method for breast cancer detection to improve the performance beyond a benchmark deep learning algorithm and to furthermore provide a color overlay visual map of the probability of malignancy within a lesion. This overall framework is termed disease-specific imaging. Previously, 150 breast lesions were segmented and classified utilizing a modified fully convolutional network and a modified GoogLeNet, respectively. In this study multiparametric analysis was performed within the contoured lesions. Features were extracted from ultrasound radiofrequency, envelope, and log-compressed data based on biophysical and morphological models. The support vector machine with a Gaussian kernel constructed a nonlinear hyperplane, and we calculated the distance between the hyperplane and each feature's data point in multiparametric space. The distance can quantitatively assess a lesion and suggest the probability of malignancy that is color-coded and overlaid onto B-mode images. Training and evaluation were performed on in vivo patient data. The overall accuracy for the most common types and sizes of breast lesions in our study exceeded 98.0% for classification and 0.98 for an area under the receiver operating characteristic curve, which is more precise than the performance of radiologists and a deep learning system. Further, the correlation between the probability and Breast Imaging Reporting and Data System enables a quantitative guideline to predict breast cancer. Therefore, we anticipate that the proposed framework can help radiologists achieve more accurate and convenient breast cancer classification and detection.

Prediction of response to neoadjuvant chemotherapy in breast cancer with recurrent neural networks and raw ultrasound signals

Objective. Prediction of the response to neoadjuvant chemotherapy (NAC) in breast cancer is important for patient outcomes. In this work, we propose a deep learning based approach to NAC response prediction in ultrasound (US) imaging. Approach. We develop recurrent neural networks that can process serial US imaging data to predict chemotherapy outcomes. We present models that can process either raw radio-frequency (RF) US data or regular US images. The proposed approach is evaluated based on 204 sequences of US data from 51 breast cancers. Each sequence included US data collected before the chemotherapy and after each subsequent dose, up to the 4th course. We investigate three pre-trained convolutional neural networks (CNNs) as back-bone feature extractors for the recurrent network. The CNNs were pre-trained using raw US RF data, US b-mode images and RGB images from the ImageNet dataset. The first two networks were developed using US data collected from malignant and benign breast masses. Main results. For the pre-treatment data, the better performing network, with back-bone CNN pre-trained on US images, achieved area under the receiver operating curve (AUC) of 0.81 (±0.04). Performance of the recurrent networks improved with each course of the chemotherapy. For the 4th course, the better performing model, based on the CNN pre-trained with RGB images, achieved AUC value of 0.93 (±0.03). Statistical analysis based on the DeLong test presented that there were no significant differences in AUC values between the pre-trained networks at each stage of the chemotherapy (p-values > 0.05). Significance. Our study demonstrates the feasibility of using recurrent neural networks for the NAC response prediction in breast cancer US.