Radiomics: the bridge between medical imaging and personalized medicine

Nature Reviews Clinical Oncology, Journal Year: 2017, Volume and Issue: 14(12), P. 749 - 762

Published: Oct. 4, 2017

Language: Английский

---

Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study

PLoS Medicine, Journal Year: 2019, Volume and Issue: 16(1), P. e1002730 - e1002730

Published: Jan. 24, 2019

Background For virtually every patient with colorectal cancer (CRC), hematoxylin–eosin (HE)–stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can prognosticators directly from these widely available images. Methods and findings We hand-delineated single-tissue regions in 86 CRC slides, yielding more than 100,000 HE image patches, train a CNN by transfer learning, reaching nine-class accuracy of >94% an independent data set 7,180 25 patients. With this tool, performed automated decomposition representative multitissue 862 500 stage I–IV patients The Cancer Genome Atlas (TCGA) cohort, large international multicenter collection tissue. Based on output neuron activations CNN, calculated "deep stroma score," was factor for overall survival (OS) multivariable Cox proportional hazard model (hazard ratio [HR] 95% confidence interval [CI]: 1.99 [1.27–3.12], p = 0.0028), while same manual quantification stromal areas gene expression signature cancer-associated fibroblasts (CAFs) were only specific tumor stages. validated cohort 409 "Darmkrebs: Chancen der Verhütung durch Screening" (DACHS) study who recruited between 2003 2007 multiple institutions Germany. Again, score OS (HR 1.63 [1.14–2.33], 0.008), CRC-specific 2.29 [1.5–3.48], 0.0004), relapse-free (RFS; HR 1.92 [1.34–2.76], 0.0004). A prospective validation required before biomarker be implemented clinical workflows. Conclusions our retrospective show that assess human microenvironment predict prognosis histopathological

Language: Английский

---

The Applications of Radiomics in Precision Diagnosis and Treatment of Oncology: Opportunities and Challenges

Theranostics, Journal Year: 2019, Volume and Issue: 9(5), P. 1303 - 1322

Published: Jan. 1, 2019

Medical imaging can assess the tumor and its environment in their entirety, which makes it suitable for monitoring temporal spatial characteristics of tumor.Progress computational methods, especially artificial intelligence medical image process analysis, has converted these images into quantitative minable data associated with clinical events oncology management.This concept was first described as radiomics 2012.Since then, computer scientists, radiologists, oncologists have gravitated towards this new tool exploited advanced methodologies to mine information behind images.On basis a great quantity radiographic novel technologies, researchers developed validated radiomic models that may improve accuracy diagnoses therapy response assessments.Here, we review recent methodological developments radiomics, including acquisition, segmentation, feature extraction, modelling, well rapidly developing deep learning technology.Moreover, outline main applications diagnosis, treatment planning evaluations field aim personalized medicine.Finally, discuss challenges scope applicability methods.

Language: Английский

---

Promises and challenges for the implementation of computational medical imaging (radiomics) in oncology

Annals of Oncology, Journal Year: 2017, Volume and Issue: 28(6), P. 1191 - 1206

Published: Feb. 7, 2017

Language: Английский

---

Radiomic analysis of contrast-enhanced CT predicts microvascular invasion and outcome in hepatocellular carcinoma

Journal of Hepatology, Journal Year: 2019, Volume and Issue: 70(6), P. 1133 - 1144

Published: March 13, 2019

Language: Английский

---

Radiomics Analysis for Evaluation of Pathological Complete Response to Neoadjuvant Chemoradiotherapy in Locally Advanced Rectal Cancer

Clinical Cancer Research, Journal Year: 2017, Volume and Issue: 23(23), P. 7253 - 7262

Published: Sept. 23, 2017

Purpose: To develop and validate a radiomics model for evaluating pathologic complete response (pCR) to neoadjuvant chemoradiotherapy in patients with locally advanced rectal cancer (LARC).Experimental Design: We enrolled 222 (152 the primary cohort 70 validation cohort) clinicopathologically confirmed LARC who received before surgery. All underwent T2-weighted diffusion-weighted imaging after chemoradiotherapy; 2,252 radiomic features were extracted from each patient treatment imaging. The two-sample t test least absolute shrinkage selection operator regression used feature selection, whereupon signature was built support vector machines. Multivariable logistic analysis then incorporating independent clinicopathologic risk factors. performance of assessed by its calibration, discrimination, clinical usefulness validation.Results: comprised 30 selected showed good discrimination both cohorts. individualized model, which incorporated tumor length, also an area under receiver operating characteristic curve 0.9756 (95% confidence interval, 0.9185-0.9711) cohort, calibration. Decision utility model.Conclusions: Using pre- posttreatment MRI data, we developed excellent individualized, noninvasive prediction pCR. This may be identify can omit surgery chemoradiotherapy. Clin Cancer Res; 23(23); 7253-62. ©2017 AACR.

Language: Английский

---

Radiomics Features of Multiparametric MRI as Novel Prognostic Factors in Advanced Nasopharyngeal Carcinoma

Clinical Cancer Research, Journal Year: 2017, Volume and Issue: 23(15), P. 4259 - 4269

Published: March 10, 2017

Purpose: To identify MRI-based radiomics as prognostic factors in patients with advanced nasopharyngeal carcinoma (NPC).Experimental Design: One-hundred and eighteen (training cohort: n = 88; validation 30) NPC were enrolled. A total of 970 features extracted from T2-weighted (T2-w) contrast-enhanced T1-weighted (CET1-w) MRI. Least absolute shrinkage selection operator (LASSO) regression was applied to select for progression-free survival (PFS) nomograms. Nomogram discrimination calibration evaluated. Associations between clinical data investigated using heatmaps.Results: The signatures significantly associated PFS. signature derived joint CET1-w T2-w images showed better performance than or alone. One nomogram combined a the TNM staging system. This significant improvement over system terms evaluating PFS training cohort (C-index, 0.761 vs. 0.514; P < 2.68 × 10-9). Another integrated all data, thereby outperformed based on alone 0.776 0.649; 1.60 10-7). Calibration curves good agreement. Findings confirmed cohort. Heatmaps revealed associations tumor stages.Conclusions: Multiparametric nomograms provided improved ability NPC. These results provide an illustrative example precision medicine may affect treatment strategies. Clin Cancer Res; 23(15); 4259-69. ©2017 AACR.

Language: Английский

---

Deep learning radiomics can predict axillary lymph node status in early-stage breast cancer

Nature Communications, Journal Year: 2020, Volume and Issue: 11(1)

Published: March 6, 2020

Abstract Accurate identification of axillary lymph node (ALN) involvement in patients with early-stage breast cancer is important for determining appropriate treatment options and therefore avoiding unnecessary surgery complications. Here, we report deep learning radiomics (DLR) conventional ultrasound shear wave elastography predicting ALN status preoperatively cancer. Clinical parameter combined DLR yields the best diagnostic performance between disease-free axilla any metastasis areas under receiver operating characteristic curve (AUC) 0.902 (95% confidence interval [CI]: 0.843, 0.961) test cohort. This clinical can also discriminate low heavy metastatic burden disease AUC 0.905 CI: 0.814, 0.996) Our study offers a noninvasive imaging biomarker to predict extent

Language: Английский

---

Beyond imaging: The promise of radiomics

Physica Medica, Journal Year: 2017, Volume and Issue: 38, P. 122 - 139

Published: June 1, 2017

Language: Английский

---

Deep learning Radiomics of shear wave elastography significantly improved diagnostic performance for assessing liver fibrosis in chronic hepatitis B: a prospective multicentre study

Gut, Journal Year: 2018, Volume and Issue: 68(4), P. 729 - 741

Published: May 5, 2018

We aimed to evaluate the performance of newly developed deep learning Radiomics elastography (DLRE) for assessing liver fibrosis stages. DLRE adopts radiomic strategy quantitative analysis heterogeneity in two-dimensional shear wave (2D-SWE) images.A prospective multicentre study was conducted assess its accuracy patients with chronic hepatitis B, comparison 2D-SWE, aspartate transaminase-to-platelet ratio index and based on four factors, by using biopsy as reference standard. Its robustness were also investigated applying different number acquisitions training cohorts, respectively. Data 654 potentially eligible prospectively enrolled from 12 hospitals, finally 398 1990 images included. Analysis receiver operating characteristic (ROC) curves performed calculate optimal area under ROC curve (AUC) cirrhosis (F4), advanced (≥F3) significance (≥F2).AUCs 0.97 F4 (95% CI 0.94 0.99), 0.98 ≥F3 0.96 1.00) 0.85 0.81 0.89) ≥F2, which significantly better than other methods except 2D-SWE ≥F2. diagnostic improved more (especially ≥3 images) acquired each individual. No significant variation found if cohorts applied.DLRE shows best overall predicting stages compared biomarkers. It is valuable practical non-invasive accurate diagnosis HBV-infected patients.NCT02313649; Post-results.

Language: Английский

---