Joint multi-omics discriminant analysis with consistent representation learning using PANDA.

Integrative multi-omics analysis provides deeper insight and enables better and more realistic modeling of the underlying biology and causes of diseases than does single omics analysis. Although several integrative multi-omics analysis methods have been proposed and demonstrated promising results in integrating distinct omics datasets, inconsistent distribution of the different omics data, which is caused by technology variations, poses a challenge for paired integrative multi-omics methods. In addition, the existing discriminant analysis-based integrative methods do not effectively exploit correlation and consistent discriminant structures, necessitating a compromise between correlation and discrimination in using these methods. Herein we present PAN-omics Discriminant Analysis (PANDA), a joint discriminant analysis method that seeks omics-specific discriminant common spaces by jointly learning consistent discriminant latent representations for each omics. PANDA jointly maximizes between-class and minimizes within-class omics variations in a common space and simultaneously models the relationships among omics at the consistency representation and cross-omics correlation levels, overcoming the need for compromise between discrimination and correlation as with the existing integrative multi-omics methods. Because of the consistency representation learning incorporated into the objective function of PANDA, this method seeks a common discriminant space to minimize the differences in distributions among omics, can lead to a more robust latent representations than other methods, and is against the inconsistency of the different omics. We compared PANDA to 10 other state-of-the-art multi-omics data integration methods using both simulated and real-world multi-omics datasets and found that PANDA consistently outperformed them while providing meaningful discriminant latent representations. PANDA is implemented using both R and MATLAB, with codes available at https://github.com/WuLabMDA/PANDA.

Enhancing NSCLC recurrence prediction with PET/CT habitat imaging, ctDNA, and integrative radiogenomics-blood insights.

While we recognize the prognostic importance of clinicopathological measures and circulating tumor DNA (ctDNA), the independent contribution of quantitative image markers to prognosis in non-small cell lung cancer (NSCLC) remains underexplored. In our multi-institutional study of 394 NSCLC patients, we utilize pre-treatment computed tomography (CT) and 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) to establish a habitat imaging framework for assessing regional heterogeneity within individual tumors. This framework identifies three PET/CT subtypes, which maintain prognostic value after adjusting for clinicopathologic risk factors including tumor volume. Additionally, these subtypes complement ctDNA in predicting disease recurrence. Radiogenomics analysis unveil the molecular underpinnings of these imaging subtypes, highlighting downregulation in interferon alpha and gamma pathways in the high-risk subtype. In summary, our study demonstrates that these habitat imaging subtypes effectively stratify NSCLC patients based on their risk levels for disease recurrence after initial curative surgery or radiotherapy, providing valuable insights for personalized treatment approaches.

Synthetic PET from CT improves diagnosis and prognosis for lung cancer: Proof of concept.

[18F]Fluorodeoxyglucose positron emission tomography (FDG-PET) and computed tomography (CT) are indispensable components in modern medicine. Although PET can provide additional diagnostic value, it is costly and not universally accessible, particularly in low-income countries. To bridge this gap, we have developed a conditional generative adversarial network pipeline that can produce FDG-PET from diagnostic CT scans based on multi-center multi-modal lung cancer datasets (n = 1,478). Synthetic PET images are validated across imaging, biological, and clinical aspects. Radiologists confirm comparable imaging quality and tumor contrast between synthetic and actual PET scans. Radiogenomics analysis further proves that the dysregulated cancer hallmark pathways of synthetic PET are consistent with actual PET. We also demonstrate the clinical values of synthetic PET in improving lung cancer diagnosis, staging, risk prediction, and prognosis. Taken together, this proof-of-concept study testifies to the feasibility of applying deep learning to obtain high-fidelity PET translated from CT.

Automated Cellular-Level Dual Global Fusion of Whole-Slide Imaging for Lung Adenocarcinoma Prognosis.

Histopathologic whole-slide images (WSI) are generally considered the gold standard for cancer diagnosis and prognosis. Survival prediction based on WSI has recently attracted substantial attention. Nevertheless, it remains a central challenge owing to the inherent difficulties of predicting patient prognosis and effectively extracting informative survival-specific representations from WSI with highly compounded gigapixels. In this study, we present a fully automated cellular-level dual global fusion pipeline for survival prediction. Specifically, the proposed method first describes the composition of different cell populations on WSI. Then, it generates dimension-reduced WSI-embedded maps, allowing for efficient investigation of the tumor microenvironment. In addition, we introduce a novel dual global fusion network to incorporate global and inter-patch features of cell distribution, which enables the sufficient fusion of different types and locations of cells. We further validate the proposed pipeline using The Cancer Genome Atlas lung adenocarcinoma dataset. Our model achieves a C-index of 0.675 (±0.05) in the five-fold cross-validation setting and surpasses comparable methods. Further, we extensively analyze embedded map features and survival probabilities. These experimental results manifest the potential of our proposed pipeline for applications using WSI in lung adenocarcinoma and other malignancies.

SwarmDeepSurv: swarm intelligence advances deep survival network for prognostic radiomics signatures in four solid cancers.

Survival models exist to study relationships between biomarkers and treatment effects. Deep learning-powered survival models supersede the classical Cox proportional hazards (CoxPH) model, but substantial performance drops were observed on high-dimensional features because of irrelevant/redundant information. To fill this gap, we proposed SwarmDeepSurv by integrating swarm intelligence algorithms with the deep survival model. Furthermore, four objective functions were designed to optimize prognostic prediction while regularizing selected feature numbers. When testing on multicenter sets (n = 1,058) of four different cancer types, SwarmDeepSurv was less prone to overfitting and achieved optimal patient risk stratification compared with popular survival modeling algorithms. Strikingly, SwarmDeepSurv selected different features compared with classical feature selection algorithms, including the least absolute shrinkage and selection operator (LASSO), with nearly no feature overlapping across these models. Taken together, SwarmDeepSurv offers an alternative approach to model relationships between radiomics features and survival endpoints, which can further extend to study other input data types including genomics.

Predicting benefit from immune checkpoint inhibitors in patients with non-small-cell lung cancer by CT-based ensemble deep learning: a retrospective study.

BACKGROUND: Only around 20-30% of patients with non-small-cell lung cancer (NCSLC) have durable benefit from immune-checkpoint inhibitors. Although tissue-based biomarkers (eg, PD-L1) are limited by suboptimal performance, tissue availability, and tumour heterogeneity, radiographic images might holistically capture the underlying cancer biology. We aimed to investigate the application of deep learning on chest CT scans to derive an imaging signature of response to immune checkpoint inhibitors and evaluate its added value in the clinical context. METHODS: In this retrospective modelling study, 976 patients with metastatic, EGFR/ALK negative NSCLC treated with immune checkpoint inhibitors at MD Anderson and Stanford were enrolled from Jan 1, 2014, to Feb 29, 2020. We built and tested an ensemble deep learning model on pretreatment CTs (Deep-CT) to predict overall survival and progression-free survival after treatment with immune checkpoint inhibitors. We also evaluated the added predictive value of the Deep-CT model in the context of existing clinicopathological and radiological metrics. FINDINGS: Our Deep-CT model demonstrated robust stratification of patient survival of the MD Anderson testing set, which was validated in the external Stanford set. The performance of the Deep-CT model remained significant on subgroup analyses stratified by PD-L1, histology, age, sex, and race. In univariate analysis, Deep-CT outperformed the conventional risk factors, including histology, smoking status, and PD-L1 expression, and remained an independent predictor after multivariate adjustment. Integrating the Deep-CT model with conventional risk factors demonstrated significantly improved prediction performance, with overall survival C-index increases from 0·70 (clinical model) to 0·75 (composite model) during testing. On the other hand, the deep learning risk scores correlated with some radiomics features, but radiomics alone could not reach the performance level of deep learning, indicating that the deep learning model effectively captured additional imaging patterns beyond known radiomics features. INTERPRETATION: This proof-of-concept study shows that automated profiling of radiographic scans through deep learning can provide orthogonal information independent of existing clinicopathological biomarkers, bringing the goal of precision immunotherapy for patients with NSCLC closer. FUNDING: National Institutes of Health, Mark Foundation Damon Runyon Foundation Physician Scientist Award, MD Anderson Strategic Initiative Development Program, MD Anderson Lung Moon Shot Program, Andrea Mugnaini, and Edward L C Smith.

Machine Learning Models for the Identification of Prognostic and Predictive Cancer Biomarkers: A Systematic Review.

The identification of biomarkers plays a crucial role in personalized medicine, both in the clinical and research settings. However, the contrast between predictive and prognostic biomarkers can be challenging due to the overlap between the two. A prognostic biomarker predicts the future outcome of cancer, regardless of treatment, and a predictive biomarker predicts the effectiveness of a therapeutic intervention. Misclassifying a prognostic biomarker as predictive (or vice versa) can have serious financial and personal consequences for patients. To address this issue, various statistical and machine learning approaches have been developed. The aim of this study is to present an in-depth analysis of recent advancements, trends, challenges, and future prospects in biomarker identification. A systematic search was conducted using PubMed to identify relevant studies published between 2017 and 2023. The selected studies were analyzed to better understand the concept of biomarker identification, evaluate machine learning methods, assess the level of research activity, and highlight the application of these methods in cancer research and treatment. Furthermore, existing obstacles and concerns are discussed to identify prospective research areas. We believe that this review will serve as a valuable resource for researchers, providing insights into the methods and approaches used in biomarker discovery and identifying future research opportunities.

Efficacy and clinicogenomic correlates of response to immune checkpoint inhibitors alone or with chemotherapy in non-small cell lung cancer.

The role of combination chemotherapy with immune checkpoint inhibitors (ICI) (ICI-chemo) over ICI monotherapy (ICI-mono) in non-small cell lung cancer (NSCLC) remains underexplored. In this retrospective study of 1133 NSCLC patients, treatment with ICI-mono vs ICI-chemo associate with higher rates of early progression, but similar long-term progression-free and overall survival. Sequential vs concurrent ICI and chemotherapy have similar long-term survival, suggesting no synergism from combination therapy. Integrative modeling identified PD-L1, disease burden (Stage IVb; liver metastases), and STK11 and JAK2 alterations as features associate with a higher likelihood of early progression on ICI-mono. CDKN2A alterations associate with worse long-term outcomes in ICI-chemo patients. These results are validated in independent external (n = 89) and internal (n = 393) cohorts. This real-world study suggests that ICI-chemo may protect against early progression but does not influence overall survival, and nominates features that identify those patients at risk for early progression who may maximally benefit from ICI-chemo.

Learning-based Cancer Treatment Outcome Prognosis using Multimodal Biomarkers.

Predicting early in treatment whether a tumor is likely to be responsive is a difficult yet important task to support clinical decision-making. Studies have shown that multimodal biomarkers could provide complementary information and lead to more accurate treatment outcome prognosis than unimodal biomarkers. However, the prognosis accuracy could be affected by multimodal data heterogeneity and incompleteness. The small-sized and imbalance datasets also bring additional challenges for training a designed prognosis model. In this study, a modular framework employing multimodal biomarkers for cancer treatment outcome prediction was proposed. It includes four modules of synthetic data generation, deep feature extraction, multimodal feature fusion, and classification to address the challenges described above. The feasibility and advantages of the designed framework were demonstrated through an example study, in which the goal was to stratify oropharyngeal squamous cell carcinoma (OPSCC) patients with low- and high-risks of treatment failures by use of positron emission tomography (PET) image data and microRNA (miRNA) biomarkers. The superior prognosis performance and the comparison with other methods demonstrated the efficiency of the proposed framework and its ability of enabling seamless integration, validation and comparison of various algorithms in each module of the framework. The limitation and future work was discussed as well.

Habitat Imaging Biomarkers for Diagnosis and Prognosis in Cancer Patients Infected with COVID-19.

OBJECTIVES: Cancer patients have worse outcomes from the COVID-19 infection and greater need for ventilator support and elevated mortality rates than the general population. However, previous artificial intelligence (AI) studies focused on patients without cancer to develop diagnosis and severity prediction models. Little is known about how the AI models perform in cancer patients. In this study, we aim to develop a computational framework for COVID-19 diagnosis and severity prediction particularly in a cancer population and further compare it head-to-head to a general population. METHODS: We have enrolled multi-center international cohorts with 531 CT scans from 502 general patients and 420 CT scans from 414 cancer patients. In particular, the habitat imaging pipeline was developed to quantify the complex infection patterns by partitioning the whole lung regions into phenotypically different subregions. Subsequently, various machine learning models nested with feature selection were built for COVID-19 detection and severity prediction. RESULTS: These models showed almost perfect performance in COVID-19 infection diagnosis and predicting its severity during cross validation. Our analysis revealed that models built separately on the cancer population performed significantly better than those built on the general population and locked to test on the cancer population. This may be because of the significant difference among the habitat features across the two different cohorts. CONCLUSIONS: Taken together, our habitat imaging analysis as a proof-of-concept study has highlighted the unique radiologic features of cancer patients and demonstrated effectiveness of CT-based machine learning model in informing COVID-19 management in the cancer population.

Leveraging hybrid biomarkers in clinical endpoint prediction.

BACKGROUND: Clinical endpoint prediction remains challenging for health providers. Although predictors such as age, gender, and disease staging are of considerable predictive value, the accuracy often ranges between 60 and 80%. An accurate prognosis assessment is required for making effective clinical decisions. METHODS: We proposed an extended prognostic model based on clinical covariates with adjustment for additional variables that were radio-graphically induced, termed imaging biomarkers. Eight imaging biomarkers were introduced and investigated in a cohort of 68 non-small cell lung cancer subjects with tumor internal characteristic. The subjects comprised of 40 males and 28 females with mean age at 68.7 years. The imaging biomarkers used to quantify the solid component and non-solid component of a tumor. The extended model comprises of additional frameworks that correlate these markers to the survival ends through uni- and multi-variable analysis to determine the most informative predictors, before combining them with existing clinical predictors. Performance was compared between traditional and extended approaches using Receiver Operating Characteristic (ROC) curves, Area under the ROC curves (AUC), Kaplan-Meier (KM) curves, Cox Proportional Hazard, and log-rank tests (p-value). RESULTS: The proposed hybrid model exhibited an impressive boosting pattern over the traditional approach of prognostic modelling in the survival prediction (AUC ranging from 77 to 97%). Four developed imaging markers were found to be significant in distinguishing between subjects having more and less dense components: (P = 0.002-0.006). The correlation to survival analysis revealed that patients with denser composition of tumor (solid dominant) lived 1.6-2.2 years longer (mean survival) and 0.5-2.0 years longer (median survival), than those with less dense composition (non-solid dominant). CONCLUSION: The present study provides crucial evidence that there is an added value for incorporating additional image-based predictors while predicting clinical endpoints. Though the hypotheses were confirmed in a customized case study, we believe the proposed model is easily adapted to various clinical cases, such as predictions of complications, treatment response, and disease evolution.

Are shape morphologies associated with survival? A potential shape-based biomarker predicting survival in lung cancer.

PURPOSE: Imaging biomarkers (IBMs) are increasingly investigated as prognostic indicators. IBMs might be capable of assisting treatment selection by providing useful insights into tumor-specific factors in a non-invasive manner. METHODS: We investigated six three-dimensional shape-based IBMs: eccentricities between (I) intermediate-major axis (Eimaj), (II) intermediate-minor axis (Eimin), (III) major-minor axis (Emj-mn) and volumetric index of (I) sphericity (VioS), (II) flattening (VioF), (III) elongating (VioE). Additionally, we investigated previously established two-dimensional shape IBMs: eccentricity (E), index of sphericity (IoS), and minor-to-major axis length (Mn_Mj). IBMs were compared in terms of their predictive performance for 5-year overall survival in two independent cohorts of patients with lung cancer. Cohort 1 received surgical excision, while cohort 2 received radiation therapy alone or chemo-radiation therapy. Univariate and multivariate survival analyses were performed. Correlations with clinical parameters were evaluated using analysis of variance. IBM reproducibility was assessed using concordance correlation coefficients (CCCs). RESULTS: E was associated with reduced survival in cohort 1 (hazard ratio [HR]: 0.664). Eimin and VioF were associated with reduced survival in cohort 2 (HR 1.477 and 1.701). VioS was associated with reduced survival in cohorts 1 and 2 (HR 1.758 and 1.472). Spherical tumors correlated with shorter survival durations than did irregular tumors (median survival difference: 1.21 and 0.35 years in cohorts 1 and 2, respectively). VioS was a significant predictor of survival in multivariate analyses of both cohorts. All IBMs showed good reproducibility (CCC ranged between 0.86-0.98). CONCLUSIONS: In both investigated cohorts, VioS successfully linked shape morphology to patient survival.

Automated integrated system for stained neuron detection: An end-to-end framework with a high negative predictive rate.

BACKGROUND AND OBJECTIVE: Mapping the architecture of the brain is essential for identifying the neural computations that affect behavior. Traditionally in histology, stained objects in tissue slices are hand-marked under a microscope in a manually intensive, time-consuming process. An integrated hardware and software system is needed to automate image acquisition, image processing, and object detection. Such a system would enable high throughput tissue analysis to rapidly map an entire brain. METHODS: We demonstrate an automated system to detect neurons using a monkey brain slice immunohistochemically stained for retrogradely labeled neurons. The proposed system obtains a reconstructed image of the sample, and stained neurons are detected in three steps. First, the reconstructed image is pre-processed using adaptive histogram equalization. Second, candidates for stained neurons are segmented from each region via marker-controlled watershed transformation (MCWT) using maximally stable extremal regions (MSERs). Third, the candidates are categorized as neurons or non-neurons using deep transfer learning via pre-trained convolutional neural networks (CNN). RESULTS: The proposed MCWT algorithm was compared qualitatively against MorphLibJ and an IHC analysis tool, while our unified classification algorithm was evaluated quantitatively using ROC analyses. The proposed classification system was first compared with five previously developed layers (AlexNet, VGG-16, VGG-19, GoogleNet, and ResNet). A comparison with conventional multi-stage frameworks followed using six off-the-shelf classifiers [Bayesian network (BN), support vector machines (SVM), decision tree (DT), bagging (BAG), AdaBoost (ADA), and logistic regression (LR)] and two descriptors (LBP and HOG). The system achieved a 0.918 F1-score with an 86.6% negative prediction value. Remarkably, other metrics such as precision, recall, and F-scores surpassed the 90% threshold compared to traditional methods. CONCLUSIONS: We demonstrate a fully automated, integrated hardware and software system for rapidly acquiring focused images and identifying neurons from a stained brain slice. This system could be adapted for the identification of stained features of any biological tissue.

Computer-assisted subtyping and prognosis for non-small cell lung cancer patients with unresectable tumor.

BACKGROUND: The histological classification or subtyping of non-small cell lung cancer is essential for systematic therapy decisions. Differentiating between the two main subtypes of pulmonary adenocarcinoma and squamous cell carcinoma highlights the considerable differences that exist in the prognosis of patient outcomes. Physicians rely on a pathological analysis to reveal these phenotypic variations that requires invasive methods, such as biopsy and resection sample, but almost 70% of tumors are unresectable at the point of diagnosis. METHOD: A computational method that fuses two frameworks of computerized subtyping and prognosis was proposed, and it was validated against publicly available dataset in The Cancer Imaging Archive that consisted of 82 curated patients with CT scans. The accuracy of the proposed method was compared with the gold standard of pathological analysis, as defined by theInternational Classification of Disease for Oncology (ICD-O). A series of survival outcome test cases were evaluated using the Kaplan-Meier estimator and log-rank test (p-value) between the computational method and ICD-O. RESULTS: The computational method demonstrated high accuracy in subtyping (96.2%) and good consistency in the statistical significance of overall survival prediction for adenocarcinoma and squamous cell carcinoma patients (p < 0.03) with respect to its counterpart pathological subtyping (p < 0.02). The degree of reproducibility between prognosis taken on computational and pathological subtyping was substantial with an averaged concordance correlation coefficient (CCC) of 0.9910. CONCLUSION: The findings in this study support the idea that quantitative analysis is capable of representing tissue characteristics, as offered by a qualitative analysis.

Deciphering unclassified tumors of non-small-cell lung cancer through radiomics.

BACKGROUND: Tumors are highly heterogeneous at the phenotypic, physiologic, and genomic levels. They are categorized in terms of a differentiated appearance under a microscope. Non-small-cell lung cancer tumors are categorized into three main subgroups: adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. In approximately 20% of pathology reports, they are returned unclassified or classified as not-otherwise-specified (NOS) owing to scant materials or poor tumor differentiation. METHOD: We present a radiomic interrogation of molecular spatial variations to decode unclassified NOS tumor architecture quantitatively. Twelve spatial descriptors with various displacements and directions were extracted and profiled with respect to the subgroups. The profiled descriptors were used to decipher the NOS tumor morphologic clues from the imaging phenotype perspective. This profiler was built as an extended version of a conventional support-vector-machine classifier, wherein a genetic algorithm and correlation analysis were embedded to define the molecular signatures of poorly differentiated tumors using well-differentiated-tumor information. RESULTS: Sixteen multi-class classifier models with an 81% average accuracy and descriptor subset size ranging from 12 to 144 were reported. The average area under the curve was 86.3% at a 95% confidence interval and a 0.03-0.08 standard error. Correlation analysis returned an unclassified NOS membership matrix with respect to the cell-architecture similarity score for the subgroups. The best model demonstrated 53% NOS reduction. CONCLUSION: The membership matrix is expected to assist pathologists and oncologists in cases of unresectable tumors or scant biopsy materials for histological subtyping and cancer therapy.