Learning from scanners: Bias reduction and feature correction in radiomics

Zhovannik, Ivan; Bussink, Johan; Traverso, Alberto; Shi, Zhenwei; Kalendralis, Petros; Wee, Leonard; Dekker, André; Fijten, Rianne; Monshouwer, R.

doi:10.1016/j.ctro.2019.07.003

Cited by 66 publications

(69 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Note: Modelling does not have a letter associated with since there is no consensus on the best statistical modelling strategies. Problem area Potential problems Potential solutions Image acquisition A Different scanners and acquisition protocols affect feature reproducibility [ [79] , [80] , [81] , [82] , [83] , [84] , [85] , [86] , [87] , [88] , [89] , [90] , [91] ] Image phantoms on different scanners to provide baseline [ 79 ], establish credibility of scanners and protocols [ 84 ], catalogue reproducible features [ 86 , 90 ], model a correction algorithm [ 89 ], harmonize data [ 91 ]. B Patient motion affects feature reproducibility [ 80 , 92 , 93 ] Set motion tolerances, reduce ROI boundaries [ 80 ], use single phase from 4D images [ 92 ], find robust features using 4DCT data [ 93 ].…”

Section: Reported Methodological Limitations Of Ct Based Radiomics Stmentioning

confidence: 99%

“…Indeed, one study showed that using a controlled protocol across different CT scanners reduced feature variability by over 50 % compared to using local protocols [79]. Other studies used post-extraction deep learning [100] or correction factors [89] to reduce feature variability.…”

Section: Image Acquisitionmentioning

confidence: 99%

“…Image acquisition A Different scanners and acquisition protocols affect feature reproducibility [79][80][81][82][83][84][85][86][87][88][89][90][91] Image phantoms on different scanners to provide baseline [79], establish credibility of scanners and protocols [84], catalogue reproducible features [86,90], model a correction algorithm [89] [98,[110][111][112] Normalization of features to volume [98], bit depth resampling [110], feature redesign [110], more robust statistics to check added value of radiomics signatures [111]. Test re-test H Radiomic features may not be repeatable over multiple measurements [113][114][115], repeatable features are not generalizable to other disease sites [116].…”

Section: Problem Area Potential Problems Potential Solutionsmentioning

confidence: 99%

See 2 more Smart Citations

Radiomics as a personalized medicine tool in lung cancer: Separating the hope from the hype

et al. 2020

View full text Add to dashboard Cite

Radiomics has become a popular image analysis method in the last few years. Its key hypothesis is that medical images harbor biological, prognostic and predictive information that is not revealed upon visual inspection. In contrast to previous work with a priori defined imaging biomarkers, radiomics instead calculates image features at scale and uses statistical methods to identify those most strongly associated to outcome. This builds on years of research into computer aided diagnosis and pattern recognition. While the potential of radiomics to aid personalized medicine is widely recognized, several technical limitations exist which hinder biomarker translation. Aspects of the radiomic workflow lack repeatability or reproducibility under particular circumstances, which is a key requirement for the translation of imaging biomarkers into clinical practice. One of the most commonly studied uses of radiomics is for personalized medicine applications in Non-Small Cell Lung Cancer (NSCLC). In this review, we summarize reported methodological limitations in CT based radiomic analyses together with suggested solutions. We then evaluate the current NSCLC radiomics literature to assess the risk associated with accepting the published conclusions with respect to these limitations. We review different complementary scoring systems and initiatives that can be used to critically appraise data from radiomics studies. Wider awareness should improve the quality of ongoing and future radiomics studies and advance their potential as clinically relevant biomarkers for personalized medicine in patients with NSCLC.

show abstract

Section: Reported Methodological Limitations Of Ct Based Radiomics Stmentioning

confidence: 99%

Section: Image Acquisitionmentioning

confidence: 99%

Section: Problem Area Potential Problems Potential Solutionsmentioning

confidence: 99%

See 1 more Smart Citation

Radiomics as a personalized medicine tool in lung cancer: Separating the hope from the hype

et al. 2020

View full text Add to dashboard Cite

show abstract

“…One proposed solution is to eliminate the dependency of features on a certain factor by modeling the relationship and applying corrections accordingly. This had been explored recently for different CT exposure settings [123]. Another method to eliminate the dependency is to convert images using deep learning, in order to simulate reconstruction with different settings, which was shown to improve CT radiomics' reproducibility for images reconstructed with different kernels [62].…”

Section: Reduce Radiomics' Dependencymentioning

confidence: 99%

Radiomics in medical imaging—“how-to” guide and critical reflection

et al. 2020

View full text Add to dashboard Cite

Radiomics is a quantitative approach to medical imaging, which aims at enhancing the existing data available to clinicians by means of advanced mathematical analysis. Through mathematical extraction of the spatial distribution of signal intensities and pixel interrelationships, radiomics quantifies textural information by using analysis methods from the field of artificial intelligence. Various studies from different fields in imaging have been published so far, highlighting the potential of radiomics to enhance clinical decision-making. However, the field faces several important challenges, which are mainly caused by the various technical factors influencing the extracted radiomic features. The aim of the present review is twofold: first, we present the typical workflow of a radiomics analysis and deliver a practical "how-to" guide for a typical radiomics analysis. Second, we discuss the current limitations of radiomics, suggest potential improvements, and summarize relevant literature on the subject.

show abstract

“…The correct and robust application of radiomics analysis has to be investigated, since the algorithm-based analysis of quantitative date is not standardized within different institutions, or even within single institutions, and can easily be performed differently. Independently of software-and hardware-induced variability [116][117][118][119][120][121], texture analysis is also hindered by uncertainties in patient immobilization and organ movements, especially with regard to MRI examinations [122,123]. On top of this, its usage for treatment decisions and treatment planning has to be investigated in prospective trials including ex-vivo, in-vivo volunteer and in-vivo patient examinations allowing for founded conclusions about its usage.…”

Section: Current Limitations Of Radiomicsmentioning

confidence: 99%

Radiomics for liver tumours

et al. 2020

View full text Add to dashboard Cite

Current research, especially in oncology, increasingly focuses on the integration of quantitative, multiparametric and functional imaging data. In this fast-growing field of research, radiomics may allow for a more sophisticated analysis of imaging data, far beyond the qualitative evaluation of visible tissue changes. Through use of quantitative imaging data, more tailored and tumour-specific diagnostic work-up and individualized treatment concepts may be applied for oncologic patients in the future. This is of special importance in cross-sectional disciplines such as radiology and radiation oncology, with already high and still further increasing use of imaging data in daily clinical practice. Liver targets are generally treated with stereotactic body radiotherapy (SBRT), allowing for local dose escalation while preserving surrounding normal tissue. With the introduction of online target surveillance with implanted markers, 3D-ultrasound on conventional linacs and hybrid magnetic resonance imaging (MRI)-linear accelerators, individualized adaptive radiotherapy is heading towards realization. The use of big data such as radiomics and the integration of artificial intelligence techniques have the potential to further improve image-based treatment planning and structured follow-up, with outcome/toxicity prediction and immediate detection of (oligo)progression. The scope of current research in this innovative field is to identify and critically discuss possible application forms of radiomics, which is why this review tries to summarize current knowledge about interdisciplinary integration of radiomics in oncologic patients, with a focus on investigations of radiotherapy in patients with liver cancer or oligometastases including multiparametric, quantitative data into (radio)-oncologic workflow from disease diagnosis, treatment planning, delivery and patient follow-up.

show abstract

Learning from scanners: Bias reduction and feature correction in radiomics

Abstract: Highlights Typically encountered CT exposure variations result in radiomics values distortion. Proposed correction method suppresses the non-tumor related variation in radiomics. The clinical relevance was shown using a 221 lung cancer patient cohort.

Cited by 66 publications

References 13 publications

Radiomics as a personalized medicine tool in lung cancer: Separating the hope from the hype

Radiomics as a personalized medicine tool in lung cancer: Separating the hope from the hype

Radiomics in medical imaging—“how-to” guide and critical reflection

Radiomics for liver tumours

Contact Info

Product

Resources

About