Journal of Oncology Research and Treatment
Open Access

Cancer which is a global issue associated with high incidence and high mortality needs effective measures to reduce morbidity and mortality [1]. Yet, the challenge remains in the accurate detection, characterization, treatment and monitoring of cancer makes it difficult to achieve this. However, the radiomics, high-throughput extraction and analysis of large amounts of image features from radiological images have emerged as a promising method that allows for the accurate diagnosis as well proper management strategies [2]. The radiomics features captured from morphometric analysis are important to understand the geometric aspects of a particular tumor Region of Interest (ROI). The appearance of cancer or malignant cells differs from the normal cells due to the abnormality in their size, shape and other features and the increasing abnormalities in morphometry suggest the likelihood of increasing invasiveness [3-5]. Therefore, the morphometric assessment plays a crucial role in cancer detection and management. However, identifying the usefulness, robustness and challenging areas with respect to morphometric features is essential to employ such features in the clinical practice. Therefore, the following review is aimed at providing an overview of the morphometric features with significant findings and challenges encountered in utilization of morphometric features.

Outcomes of morphometric feature analysis

The significant findings related to the individual morphometric features (i.e., volume, surface area, SV, sphericity, asphercity, SD, compactness, maximum 3D diameter, flatness, axis lengths, solidity/ volume density based on convex hull, area density based on minimum volume enclosing ellipsoid and Moran’s I) are described herein.

Tumor volume is the most commonly evaluated morphometric feature since it is considered to be an important predictor in determining the clinical outcomes [6,7]. The volume was among the optimal parameters for differentiating Breast Carcinoma (BC) and breast lymphoma [8]. Nevertheless, it was not a top contributing feature for classifying histological subtypes of NSCLC [9]. Aerts revealed that the volume had a good prognostic performance for patients with NSCLC and Head and Neck Carcinoma (HNC) but combining the radiomics signature with volume had even better prognostic performance than the use of volume alone [10]. According to Carvalho volume of the Lymph Nodes (LNs) was an independent prognostic factor for NSCLC but not the volume of primary tumour [11]. However, volume was not a predictive feature of pathologic Complete Response (pCR) for NSCLC after neoadjuvant Chemo Radiotherapy (nCRT) [12]. Ulrich also revealed that a favorable prognosis was associated with small tumour volume of patients with Head and Neck Squamous Cell Carcinoma (HNSCC) after chemo radiotherapy (CRT) [13]. According to Yang volume failed to predict pCR for esophageal Squamous Cell Carcinoma (OSCC) after nCRT [14]. In addition, Gabryś showed volume as a useful predictor of longterm Xerostomia in HNC patients treated with radiotherapy while normal tissue complication models based on mean radiation dose failed to predict Xerostomia [15].

Surface area exhibited the highest difference between Grade II and Grade III gliomas in terms of mean rank values [16]. Further, the surface area obtained from margin ROI ranked 3rd among the highest ranking five features for distinguishing recurrent and non-recurrent patients with Prostate Carcinoma (PC) after radiotherapy [17]. As shown by Chad dad it was among five radiomics features that moderately correlated with survival time of large cell carcinoma. Moreover, it was significant in all four groups (i.e., large cell carcinoma, primary tumor size (T2), none LN metastasis (N0), and TNM stage I) associated with NSCLC survival [18]. Fang selected surface area to construct the radiomics signature which showed the ability for predicting treatment response in patients with locally advanced cervical carcinoma prior to concurrent CRT [19].

Yang suggested that the SVR may provide more information about pCR than tumour volume. Moreover, they revealed that a lower SVR to be indicative of a more compact shape [14]. In addition, lower SVR was shown to be an independent factor differentiating Invasive Adenocarcinomas (IAs) from Minimally Invasive Adenocarcinomas (MIAs) and Adenocarcinomas In Situ (AISs) that appear as pure ground-glass nodules (pGGNs) [20]. As shown by Park it was chosen to build the radiomics score that showed significant association with Disease-Free Survival (DFS) in patients with invasive BC. Moreover, they stated that the degree of irregularity of the tumour boundary was quantified by SVR and thus the irregular tumour boundary was likely to be associated with poor survival [21]. In addition, SVR was a dominant feature in predicting LN metastasis in patients with PC after prostate-specific membrane antigen radio-guided surgery [22].

Sphericity was the most important feature across multiple models for discriminating Glioblastoma (GBM) and Brain Metastasis (BM) [23,24]. Further, it was revealed that the sphericity value of GBM is lower than that of BM [24]. Also, it was beneficial in discriminating PC labeled with different Gleason scores [25]. This feature exhibited its usefulness as an independent parameter in distinguishing IAs and MIAs as well [26]. Likewise, radiomics signature constructed by Jiang uniting sphericity with other non-morphometric features showed good discriminative performance in differentiating IAs from MIA in pGGNs with pleural contact [27]. Sphericity was selected to build the radiomics signature which demonstrated significant differentiation between seminomas and non-seminomas according to Zhang [28].

Coroller identified a rounder tumour which was quantified by sphericity as a feature predicting Gross Residual Disease (GRD) and directly proportional to the probability of GRD in patients with NSCLC after nCRT and before surgical resection [12]. Song suggested that the sphericity may reflect histological peripheral distribution of micro papillary patterns within lung adenocarcinomas [29]. Du identified sphericity as the most important risk factor for predicting disease progression in patients with nasopharyngeal carcinoma. They showed that the risk of 3-year disease progression after radiotherapy is increased with decreasing sphericity [30]. Also, it was one of the two most significant predictors of lymphovascular invasion in OSCC [31]. Morin concluded that sphericity had the potential to predict tumour grade, local failure and Overall Survival (OS) in meningioma patients. Low sphericity was linked to increased local failure and worse OS according to them [32].

Asphercity provided better prognostic values for progression free survival and OS in NSCLS patients compared to standardized uptake value, metabolic tumour volume, total lesion glycolysis and solidity [33]. Also, it was associated with poor survival despite palliative systemic treatment in patients with metastatic colorectal carcinoma [34]. Furthermore, asphercity showed the potential to be an independent predictor of prognosis in patients with invasive ductal BC [35].

High SD was significantly associated with high grade meningiomas which exhibited non-spherical shape compared to low grade meningiomas [36]. Chu revealed that the pancreas in pancreatic ductal carcinoma indicated less SD than normal pancreas [37]. SD of the primary tumour site was associated with pCR and GRD in NSCLC patients. Moreover, this feature reflected that the rounder-shaped tumours were less likely to respond well to nCRT [38]. Bogowicz showed that larger SD linked to worse prognosis in patients with HNC. Furthermore, this larger SD indicated larger LN spread and suggested that the SD should be interpreted as a spread of disease than as complexity of LN shape [39]. Ulrich also showed that a favorable prognosis was associated with lower SD in patients with HNC [13]. Moreover, this feature ranked as the feature with the highest importance for predicting OS as well as DFS in patients with GBM [40].

Compactness 1 was a significant predictor of pCR in patients with Locally Advanced Rectal Carcinomas (LARC) and poor tumour compactness demonstrated close association with lymph vascular space invasion [41,42]. Also, this was a potential predictive feature for assessing the risk in OS of patients with HNC [43]. In addition, this feature was selected to construct the radiomics signature which exhibited significant prognostic power for patients with Oropharyngeal squamous cell carcinoma [44]. Fave revealed that the prognostic potential of the NSCLC patients was improved by selecting Compactness 2 as a pretreatment feature for OS time and time to distance metastasis models. Furthermore, their study reflected that the larger compactness 2 was associated with a higher predicted risk of experiencing the outcome [45]. Compactness 1 and compactness 2 were useful features to differentiate heart from other normal tissues and tumour volumes in patients with Hodgkin disease and Erwin sarcoma [46].

Aerts identified compactness along with four non-morphometric features to achieve significant prognostic performance in lung and HNC patients but it was not among the most dominant features [10]. However, it was among the top five discriminative features between tumour progression and pseudo progression in patents with GBM [47]. In addition, compactness was selected as a top contributing feature from morphometric features for the classification of NSCLC subtypes [9]. Besides, compactness was associated with OS of gastroesophageal junction adenocarcinoma treated with nCRT and high compactness was suggestive of low risk while low compactness was suggestive of high risk [48]. Discrete compactness demonstrated higher predictive performance in discriminating encapsulated Thymoma from invasive Thymoma according to Lee [49] but Yamazaki showed that it did not differentiate high risk and low risk Thymoma [50]. Nevertheless, it was identified as a useful parameter for differentiating subtypes of gliomas [51].

Larger maximum 3D diameter was an independent differentiator of lung adenocarcinoma [20]. In meningiomas, it was higher in brain invasion group compared to non-invasion group [52]. Yet, it was not a statistically significant feature for distinguishing histological subtypes of renal carcinomas [53]. As shown by Zhuang maximum 3D diameter was a contributing feature of CT-based radiomics score that differentiated pCR and non-pCR patients with LARC after neoadjuvant treatment compared to clinical variables [54].

Flatness was a contributing feature of the biomarker exhibiting strong and significant performance for discrimination of benign and malignant lung lesions [55]. Similarly, Palumbo found flatness to be significantly differentiating lung nodules and higher values were indicative of malignancy. However, this was applicable for PET-based features whereas CT-based features did not exhibit such significant differentiation [56]. Besides, a favorable discrimination was exhibited between lymphosvascular space invasion and non-lymphovascular space invasion in cervical carcinoma by using the radiomics nomogram in which flatness was the selected morphometric feature [57]. It was identified as a potential biomarker for predicting tumour response after radiotherapy in NSCLC patients as well [58].

The major axis length was identified as an independent prognostic factor for patients with nasopharyngeal carcinoma and its prediction of OS was better than N stage according to Zhai [59]. Except for another non-morphometric feature least axis length was recognized as the most important independent prognostic radiomics feature for nodal control in HNSCC. Furthermore, a larger least axis length of LN was likely to indicate a round-shaped LN rather than an oval-shaped LN with similar volume [60]. Also, it was among the two most significant radiomics features to discriminate patients achieved pCR and nonpCR in locally advanced rectal adenocarcinoma after nCRT [61].

Solidity was a useful feature for the classification of endometrial carcinoma patients with and without LN metastasis [62]. Also, it was a useful predictor for OS of patients with stage III NSCLC by Fried. They presented that the lower solidity was an indication for more dispersion of the tumour [63]. Higher area density based on the minimum volume enclosing ellipsoid was found to be associated with worse OS in patients with Renal Cell Carcinoma (RCC) [64]. Moran’s I was an independent prognostic factor for predicting progression free survival and OS in patients with invasive squamous cell carcinoma of vulva [65].

Promises and challenges

The morphometric features were insensitive to normalization as well as to pixel space resampling or interpolation [66,67]. Also, they were less affected by the noise which is favorable for their utilization in radiation oncology [68]. Apart from the usefulness of morphometric features to perform a given task such as discriminating benign and malignant lesions or classifying histological subtypes or predicting prognosis or assessing response to therapy, the robustness and repeatability of these features are important for achieving the optimum benefit in clinical applications. Even though morphometric features had exhibited highest repeatability and robustness [67,69] there are factors affecting their repeatability and robustness. For example, the image sequence or image contrast may impact the robustness and repeatability of a morphometric feature for a particular study [66,70]. In addition to the type of image [71,72], image acquisition parameters [73] and software platform [74] could also affect the reliability of these features. Moreover, the robustness of extracted morphometric features may vary depending on the method of segmentation [75,76]. Lack of standardization and harmonization methods is also a problem in obtaining reliable results [77]. Therefore, it is necessary to take the above factors into consideration when incorporating morphometric features into a radiomics model which would be clinically implementable and acceptable.

The morphometric features have emerged as promising biomarkers with discriminative and predictive capabilities and their exploitation with careful consideration would enhance the clinical benefit in radiation oncology. It was confirmed that the radiomics models with significant findings and challenges encountered in utilization of morphometric features. In the early detection and treatment of cancer,the morphometric evaluation is essential. To apply such features in practice while treating the patients, it is necessary to understand the usefulness, robustness, and problematic areas with respect to morphometric features.

1. Sung H, Ferlay J, Siegel RL, Eser S, Mathers C, et al. (2021) Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: Cancer J Clin 71: 209-249.

   [CrossReff] [Google Scholar] [PubMed]

2. Lambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, Van Stiphout RGPM, et al. (2012) Radiomics: Extracting more information from medical images using advanced feature analysis. Eur J Cancer; 48: 441-446.

   [CrossReff] [Google Scholar] [PubMed]

3. Lodish H, Berk A, Zipursky S, Matsudaira P, Baltimore D, et al. (2000) Cancer. In: Molecular cell biology. W. H. Freeman, New York.
4. Stephens FO, Aigner KR (2009) What is malignancy? In: Basics of oncology.
5. Snoeckx A, Reyntiens P, Desbuquoit D, Spinhoven MJ, Van Schil PE, et al. (2018) Evaluation of the solitary pulmonary nodule: Size matters, but do not ignore the power of morphology. Insights Imaging 9: 73-86.

   [CrossReff] [Google Scholar] [PubMed]

6. Strongin A, Yovino S, Taylor R, Wolf J, Cullen K, et al. (2012) Primary tumor volume is an important predictor of clinical outcomes among patients with locally advanced squamous cell cancer of the head and neck treated with definitive chemoradiotherapy. Int J Radiat Oncol Biol Phys 82: 1823-1830.

   [CrossReff] [Google Scholar [PubMed]

7. Tang X, He Q, Sun G, Qu H, Liu J, et al. (2020) Total tumor volume should be considered as an important prognostic factor for synchronous multiple gastric cancer patients with curative gastrectomy. Pathol Oncol Res 26: 2169-2175.

   [CrossReff] [Google Scholar] [PubMed]

8. Ou X, Wang J, Zhou R, Zhu S, Pang F, et al (2019) Ability of 18 F-FDG PET/CT radiomic features to distinguish breast carcinoma from breast lymphoma. Contrast Media Mol Imaging 2019.

   [CrossReff] [Google Scholar] [PubMed]

9. Patil R, Mahadevaiah G, Dekker A (2016) An Approach toward automatic classification of tumor histopathology of non-small cell lung cancer based on radiomic features. Tomography 2: 374-377.

   [CrossReff] [Google Scholar] [PubMed]

10. Aerts HJWL, Velazquez ER, Leijenaar RTH, Parmar C, Grossmann P, et al. (2014) Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun 5: 4006.

    [CrossReff] [Google Scholar] [PubMed]

11. Carvalho S, Leijenaar RTH, Troost EGC, Van Timmeren JE, Oberije C, et al. (2018) 18F-fluorodeoxyglucose positron-emission tomography (FDG- PET)-Radiomics of metastatic lymph nodes and primary tumor in non-small cell lung cancer (NSCLC)-A prospective externally validated study. PLoS One 13: e0192859.

    [CrossReff] [Google Scholar] [PubMed]

12. Coroller TP, Agrawal V, Huynh E, Narayan V, Lee SW, et al. (2017) Radiomic-Based pathological response prediction from primary tumors and lymph nodes in NSCLC. J Thorac Oncol 12: 467-476.

    [CrossReff] [Google Scholar] [PubMed]

13. Ulrich EJ, Menda Y, Ponto LLB, Anderson CM, Smith BJ, et al. (2019) FLT PET radiomics for response prediction to chemoradiation therapy in head and neck squamous cell cancer. Tomography 5: 161-169.

    [CrossReff] [Google Scholar] [PubMed]

14. Yang Z, He B, Zhuang X, Gao X, Wang D, et al. (2019) CT-based radiomic signatures for prediction of pathologic complete response in esophageal squamous cell carcinoma after neoadjuvant chemoradiotherapy. J Radiat Res 60: 538-545.

    [CrossReff] [Google Scholar] [PubMed]

15. Gabryś HS, Buettner F, Sterzing F, Hauswald H, Bangert M, et al. (2018) Design and selection of machine learning methods using radiomics and dosiomics for normal tissue complication probability modeling of xerostomia. Front Oncol 8: 35.

    [CrossReff] [Google Scholar] [PubMed]

16. Cinarer G, Gursel B. (2020) Statistical analysis of radiomic features in differentiation of glioma grades. New Trends and Issues Proceedings on Advances in Pure and Applied Sciences 12: 68-79.

    [Google Scholar]

17. Fernandes CD, Dinh CV, Walraven I, Heijmink SW, Smolic M, et al. (2018) Biochemical recurrence prediction after radiotherapy for prostate cancer with T2w magnetic resonance imaging radiomic features. Phys Imaging Radiat Oncol 7: 9-15.

    [CrossReff] [Google Scholar] [PubMed]

18. Chaddad A, Desrosiers C, Toews M, Abdulkarim B (2017) Predicting survival time of lung cancer patients using radiomic analysis. Oncotarget 8: 104393-104407.

    [CrossReff] [Google Scholar] [PubMed]

19. Fang M, Kan Y, Dong D, Yu T, Zhao N et al. (2020) Multi-habitat based radiomics for the prediction of treatment response to concurrent chemotherapy and radiation therapy in locally advanced cervical cancer. Front Oncol 10.

    [CrossReff] [Google Scholar] [PubMed]

20. Yang Y, Wang WW, Ren Y, Jin XQ, Zhu QD, et al. (2019) Computerized texture analysis predicts histological invasiveness within lung adenocarcinoma manifesting as pure ground-glass nodules. Acta Radiol 60: 1258-1264.

    [CrossReff] [Google Scholar] [PubMed]

21. Park H, Lim Y, Ko ES, Cho HH, Lee JE, et al. (2018) Radiomics signature on magnetic resonance imaging: Association with disease-free survival in patients with invasive breast cancer. Clin Cancer Res 24: 4705-4714.

    [CrossReff] [Google Scholar] [PubMed]

22. Peeken JC, Shouman MA, Kroenke M, Rauscher I, Maurer T, et al. (2020) A CT-based radiomics model to detect prostate cancer lymph node metastases in PSMA radioguided surgery patients. Eur J Nucl Med 47: 2968-2977.

    [CrossReff] [Google Scholar] [PubMed]

23. Priya S, Liu Y, Ward C, Li NH, Soni N et al (2021) Machine learning based differentiation of glioblastoma from brain metastasis using MRI derived radiomics. Sci Rep 11: 10478.

    [CrossReff] [Google Scholar] [PubMed]

24. de Causans A, Carré A, Roux A, Tauziède-Espariat A, Ammari A, et al. (2021) Development of a machine learning classifier based on radiomic features extracted from post-contrast 3D T1-weighted MR images to distinguish glioblastoma from solitary brain metastasis. Front Oncol 11: 638262.

    [CrossReff] [Google Scholar] [PubMed]

25. Brunese L, Mercaldo F, Reginelli A, Santone A (2020) Radiomics for Gleason score detection through deep learning. Sensors (Switzerland) 20: 1-23.

    [CrossReff] [Google Scholar] [PubMed]

26. Xiong Z, Jiang Y, Che S, Zhao W, Guo Y, et al. (2021) Use of CT radiomics to differentiate minimally invasive adenocarcinomas and invasive adenocarcinomas presenting as pure ground-glass nodules larger than 10 mm. Eur J Radiol 141: 109772.

    [CrossReff] [Google Scholar] [PubMed]

27. Jiang Y, Che S, Ma S, Liu X, Guo Y, et al. (2021) Radiomic signature based on CT imaging to distinguish invasive adenocarcinoma from minimally invasive adenocarcinoma in pure ground-glass nodules with pleural contact. Cancer Imaging 21: 1.

    [CrossReff] [Google Scholar] [PubMed]

28. Zhang P, Feng Z, Cai W, You H, Fan C, et al. (2019) T2-weighted image-based radiomics signature for discriminating between seminomas and nonseminoma. Front Oncol 9.

    [CrossReff] [Google Scholar] [PubMed]

29. Song SH, Park H, Lee G, Lee HY, Sohn I, et al. (2017) Imaging phenotyping using radiomics to predict micropapillary pattern within lung adenocarcinoma. J Thorac Oncol 12: 624-632.

    [CrossReff]

30. Du R, Lee VH, Yuan H, Lam K, Pang HH, et al. (2019) Radiomics model to predict early progression of nonmetastatic nasopharyngeal carcinoma after intensity modulation radiation therapy: A Multicenter Study. Radiol Artif Intell 1: e180075..

    [CrossReff] [Google Scholar] [PubMed]

31. Li Y, Yu M, Wang G, Yang L, Ma C, et al (2021) Contrast-Enhanced CT-based radiomics analysis in predicting lymphovascular invasion in esophageal squamous cell carcinoma. Front Oncol 11: 644165.

    [CrossReff] [Google Scholar] [PubMed]

32. Morin O, Chen WC, Nassiri F, Susko M, Magill ST, et al. (2019) Integrated models incorporating radiologic and radiomic features predict meningioma grade, local failure, and overall survival. Neuro-Oncol Adv 1: vdz011.

    [CrossReff] [Google Scholar] [PubMed]

33. Apostolova I, Rogasch J, Buchert R, Wertzel H, Achenbach HJ, et al. (2014) Quantitative assessment of the asphericity of pretherapeutic FDG uptake as an independent predictor of outcome in NSCLC. BMC Cancer 14: 896.

    [CrossReff] [Google Scholar] [PubMed]

34. van Helden EJ, Vacher YJL, van Wieringen WN, van Velden FHP, Verheul HMW, et al. (2018) Radiomics analysis of pre-treatment [18F] FDG PET/CT for patients with metastatic colorectal cancer undergoing palliative systemic treatment. Eur J Nucl Med Mol Imaging 45: 2307-2317.

    [CrossReff] [Google Scholar] [PubMed]

35. Jung JH, Son SH, Kim DH, Lee J, Jeong SY, et al. (2017) CONSORT-Independent prognostic value of asphericity of pretherapeutic F-18 FDG uptake by primary tumors in patients with breast cancer. Medicine (United States) 96.

    [CrossReff] [Google Scholar] [PubMed]

36. Coroller TP, Bi WL, Huynh E, Abedalthagafi M, Aizer AA, et al. (2017) Radiographic prediction of meningioma grade by semantic and radiomic features. PLoS One 12.

    [CrossReff] [Google Scholar] [PubMed]

37. Chu LC, Park S, Kawamoto S, Fouladi DF, Shayesteh S, et al. (2019) Utility of CT radiomics features in differentiation of pancreatic ductal adenocarcinoma from normal pancreatic tissue. Am J Roentgenol 213: 349-357.

    [CrossReff] [Google Scholar] [PubMed]

38. Coroller TP, Agrawal V, Narayan V, Hou Y, Grossmann P, et al. (2016) Radiomic phenotype features predict pathological response in non-small cell radiomic predicts pathological response lung cancer. Radiother Oncol 119:480-486.

    [CrossReff] [Google Scholar] [PubMed]

39. Bogowicz M, Tanadini-Lang S, Guckenberger M, Riesterer O (2019) Combined CT radiomics of primary tumor and metastatic lymph nodes improves prediction of loco-regional control in head and neck cancer. Sci Rep 9: 15198.

    [CrossReff] [Google Scholar] [PubMed]

40. Bae S, Choi YS, Ahn SS, Chang JH, Kang SG, et al. (2018) Radiomic MRI phenotyping of glioblastoma: Improving survival prediction. Radiology 289: 797-806.

    [CrossReff] [Google Scholar] [PubMed]

41. Hsu CY, Wang CW, Kuo CC, Chen YH, Lan KH, et al. (2016) Tumor compactness improves the preoperative volumetry-based prediction of the pathological complete response of rectal cancer after preoperative concurrent chemoradiotherapy. Oncotarget 8: 7921-7934.

    [CrossReff] [Google Scholar] [PubMed]

42. Wang G, Wu F, Wang J, Yang C, Zhou C, et al. (2019) Volumetric imaging parameters are significant for predicting the pathological complete response of preoperative concurrent chemoradiotherapy in local advanced rectal cancer. J Radiat Res 60: 666-676.

    [CrossReff] [Google Scholar] [PubMed]

43. Vallières M, Kay-Rivest E, Perrin LJ, Liem X, Furstoss C, et al. (2017) Radiomics strategies for risk assessment of tumour failure in head-and-neck cancer. Sci Rep 7: 10117.

    [CrossReff] [Google Scholar] [PubMed]

44. Leijenaar RTH, Carvalho S, Hoebers FJP, Aerts HJWL, van Elmpt WJC, et al. (2015) External validation of a prognostic CT-based radiomic signature in oropharyngeal squamous cell carcinoma. Acta Oncol 54: 1423-1429.

    [CrossReff] [Google Scholar] [PubMed]

45. Fave X, Zhang L, Yang J, MacKin D, Balter P, et al. (2017) Delta-radiomics features for the prediction of patient outcomes in non-small cell lung cancer. Sci Rep 7: 588.

    [CrossReff] [Google Scholar] [PubMed]

46. Hsu CY, Doubrovin M, Hua CH, Mohammed O, Shulkin BL, et al. (2018) Radiomics features differentiate between normal and tumoral high-FDG uptake. Sci Rep 8: 3913.

    [CrossReff] [Google Scholar] [PubMed]

47. Ismail M, Hill V, Statsevych V, Huang R, Prasanna P, et al. (2018) Shape features of the lesion habitat to differentiate brain tumor progression from pseudoprogression on routine multiparametric MRI: A Multisite study. Am J Neuroradiol 39: 2187-2193.

    [CrossReff] [Google Scholar] [PubMed]

48. Wang Q, Zhou S, Court LE, Verma V, Koay EJ, et al. (2017) Radiomics predicts clinical outcome in primary gastroesophageal junction adenocarcinoma treated by chemo/radiotherapy and surgery. Phys Imaging Radiat Oncol 3: 37-42.

    [CrossReff] [Google Scholar]

49. Lee JH, Park CM, Park SJ, Bae JS, Lee SM, et al. (2015) Value of computerized 3D shape analysis in differentiating encapsulated from invasive Thymomas. PLoS One 10.

    [CrossReff] [Google Scholar] [PubMed]

50. Yamazaki M, Oyanagi K, Umezu H, Yagi T, Ishikawa H, et al. (2020) Quantitative 3D shape analysis of CT images of Thymoma: A Comparison with histological types. Am J Roentgenol 214: 341-347.

    [CrossReff] [Google Scholar] [PubMed]

51. Hevia-Montiel N, Rodriguez-Perez PI, Lamothe-Molina PJ, Arellano-Reynoso A, Bribiesca E, et al. (2015) Neuro morphometry of primary brain tumors by magnetic resonance imaging. J Med Imaging 2:024503.

    [CrossReff] [Google Scholar] [PubMed]

52. Zhang J, Yao K, Liu P, Liu Z, Han T, et al. (2020) A Radiomics model for preoperative prediction of brain invasion in meningioma non-invasively based on MRI: A Multi-centre study. EBioMedicine 58: 102933.

    [CrossReff] [Google Scholar] [PubMed]

53. Uhlig J, Leha A, Delonge LM, Haack AM, Shuch B et al. (2020) Radiomic features and machine learning for the discrimination of renal tumor histological subtypes: A pragmatic study using clinical-routine computed tomography. Cancers 12: 1-14.

    [CrossReff] [Google Scholar] [PubMed]

54. Zhuang Z, Liu Z, Li J, Wang X, Xie P, et al. (2021) Radiomics signature of the FOWARC trial predicts pathological response to neoadjuvant treatment in rectal cancer. J Transl Med 19: 256.

    [CrossReff] [Google Scholar] [PubMed]

55. van Griethuysen JJM, Fedorov A, Parmar C, Hosny A, Aucoin N, et al. (2017) Computational radiomics system to decode the radiographic phenotype. Cancer Res 77: e104-e107.

    [CrossReff], [Google Scholar], [PubMed]

56. Palumbo B, Bianconi F, Palumbo I, Fravolini ML, Minestrini M, et al. (2020) Value of shape and texture features from 18F-FDG PET/CT to discriminate between benign and malignant solitary pulmonary nodules: an experimental evaluation. Diagnostics 10.

    [CrossReff] [Google Scholar] [PubMed]

57. Li Z, Li H, Wang S, Dong D, Yin F, et al. (2019) MR-Based Radiomics Nomogram of cervical cancer in prediction of the lymph-vascular space invasion preoperatively. J Magn Reson Imaging 49: 1420-1426.

    [CrossReff] [Google Scholar] [PubMed]

58. Yan M, Wang W (2020) Radiomic analysis of CT predicts tumor response in human lung cancer with radiotherapy. J Digit Imaging 33: 1401-1403.

    [CrossReff] [Google Scholar] [PubMed]

59. Zhai TT, van Dijk Lv, Huang BT, Lin ZX, Ribeiro CO, et al. (2017) Improving the prediction of overall survival for head and neck cancer patients using image biomarkers in combination with clinical parameters. Radiother Oncol 124: 256-262.

    [CrossReff] [Google Scholar] [PubMed]

60. Zhai TT, Wesseling F, Langendijk JA, Shi Z, Kalendralis P et al. (2021) External validation of nodal failure prediction models including radiomics in head and neck cancer. Oral Oncol 112: 105083.

    [CrossReff] [Google Scholar] [PubMed]

61. Boldrini L, Cusumano D, Chiloiro G, Casà C, Masciocchi C, et al. (2019) Delta radiomics for rectal cancer response prediction with hybrid 0.35 T magnetic resonance-guided radiotherapy (MRgRT): a hypothesis-generating study for an innovative personalized medicine approach. Radiol Med 124: 145-153.

    [CrossReff] [Google Scholar] [PubMed]

62. de Bernardi E, Buda A, Guerra L, Vicini D, Elisei F, et al. (2018) Radiomics of the primary tumour as a tool to improve 18F-FDG-PET sensitivity in detecting nodal metastases in endometrial cancer. EJNMMI Res 8: 86.

    [CrossReff] [Google Scholar] [PubMed]

63. Fried D v., Mawlawi O, Zhang L, Fave X, Zhou S, et al. (2016) Stage III non-small cell lung cancer: Prognostic value of FDG PET quantitative imaging features combined with clinical prognostic factors. Radiology 278: 214-222.

    [CrossReff] [Google Scholar] [PubMed]

64. Khodabakhshi Z, Amini M, Mostafaei S, Avval AH, Nazari M, et al. (2021) Overall survival prediction in renal cell carcinoma patients using computed tomography radiomic and clinical information. J Digit Imaging 34: 1086-1098.

    [CrossReff] [Google Scholar] [PubMed]

65. Collarino A, Garganese G, Fragomeni SM, Arias-Bouda LMP, Ieria FP, et al. (2019) Radiomics in vulvar cancer: First clinical experience using 18 F-FDG PET/CT images. J Nucl Med 60: 199-206.

    [CrossReff] [Google Scholar] [PubMed]

66. McHugh DJ, Porta N, Little RA, Cheung S, Watson Y et al. (2021) Image contrast, image pre-processing, and T1 mapping affect MRI radiomic feature repeatability in patients with colorectal cancer liver metastases. Cancers 13: 1-21.

    [CrossReff] [Google Scholar] [PubMed]

67. Park SH, Lim H, Bae BK, Hahm MH, Chong GO, et al. (2021) Robustness of magnetic resonance radiomic features to pixel size resampling and interpolation in patients with cervical cancer. Cancer Imaging 21: 19.

    [CrossReff] [Google Scholar] [PubMed]

68. Oliver JA, Budzevich M, Hunt D, Moros EG, Latifi K, et al. (2017) Sensitivity of image features to noise in conventional and respiratory-gated PET/CT images of lung cancer: Uncorrelated noise effects. Technol Cancer Res Treat 16: 595-608.

    [CrossReff] [Google Scholar] [PubMed]

69. Merisaari H, Taimen P, Shiradkar R, Ettala O, Pesola M, et al. (2020) Repeatability of radiomics and machine learning for DWI: Short-term repeatability study of 112 patients with prostate cancer. Magn Reson Med 83: 2293-2309.

    [CrossReff] [Google Scholar] [PubMed]

70. Baeßler B, Weiss K, dos Santos DP. (2019) Robustness and reproducibility of radiomics in magnetic resonance imaging: A phantom study. Investig Radiol 54: 221-228.

    [CrossReff] [Google Scholar] [PubMed]

71. Huynh E, Coroller TP, Narayan V, Agrawal V, Romano J, et al. (2017) Associations of radiomic data extracted from static and respiratory-gated CT scans with disease recurrence in lung cancer patients treated with SBRT. PLoS One 12: e0169172.

    [CrossReff] [Google Scholar] [PubMed]

72. Sha X, Gong G, Qiu Q, Duan J, Li D et al. (2020) Discrimination of mediastinal metastatic lymph nodes in NSCLC based on radiomic features in different phases of CT imaging. BMC Med Imaging 20:

    [CrossReff] [Google Scholar] [PubMed].

73. Lu L, Ehmke RC, Schwartz LH, Zhao B (2016) Assessing agreement between radiomic features computed for multiple CT imaging settings. PLoS One 11: e0166550.

    [CrossReff] [Google Scholar] [PubMed]

74. Fornacon-Wood I, Mistry H, Ackermann CJ, Blackhall F, McPartlin A, et al. (2020) Reliability and prognostic value of radiomic features are highly dependent on choice of feature extraction platform. Eur Radiol 30: 6241-6250.

    [CrossReff] [Google Scholar] [PubMed]

75. Kalpathy-Cramer J, Mamomov A, Zhao B, Lu L, Cherezov D, et al. (2016) Radiomics of lung nodules: A Multi-institutional study of robustness and agreement of quantitative imaging features. Tomography 2: 430-437.

    [CrossReff] [Google Scholar] [PubMed]

76. Zhao B, Tan Y, Tsai WY, Qi J, Xie C et al. (2016) Reproducibility of radiomics for deciphering tumor phenotype with imaging. Sci Rep 6: 23428.

    [CrossReff] [Google Scholar] [PubMed]

77. Papadimitroulas P, Brocki L, Chung NC, Marchadour W, Vermet F, et al. (2021) Artificial intelligence: Deep learning in oncological radiomics and challenges of interpretability and data harmonization. Phys Med 83: 108-121.

    [CrossReff] [Google Scholar] [PubMed]