Estimation of colon cancer grade and metastatic lymph node involvement using DWI/ADC sequences

Abstract

Background

The potential benefit of neoadjuvant chemotherapy (NAC) in colon cancer is under evaluation. There is a need to improve preoperative non-invasive diagnostics using techniques that provide more accurate staging information in assessing patient eligibility for NAC.

Purpose

To investigate the link between the tumor grade (pathohistological confirmed) and the N status (corresponding to lymph node involvement) with apparent diffusion coefficient (ADC) values.

Material and Methods

A total of 17 patients planned for surgical resection had a biopsy confirming colon carcinoma and participated in the study. Abdominal magnetic resonance imaging with diffusion-weighted imaging/ADC sequence was recorded before surgery. The tumor and all visible lymph nodes were manually delineated directly on a grayscale ADC map for every single slice and detected to access the total tumor and summarized lymph node volume. The mean ADC value was further calculated for the mean tumor and mean lymph node values.

Results

Low-grade tumors had a mean ADC equivalent to 1225 ± 170×10⁻⁶ mm²/s, and the coefficient of high-grade tumors was 1444 ± 69×10⁻⁶ mm²/s. The group of patients with positive lymph nodes in operative tissue samples (N+) exhibited lower mean ADC values (1023 ± 142×10⁻⁶ mm²/s) as opposed to the group without metastatic lymph nodes (N–) with ADC values of 1260 ± 231×10⁻⁶ mm²/s.

Conclusion

The mean whole-tumor ADC is associated with the histological tumor grade, and the mean ADC value of whole-volume abdominal lymph nodes could assume real nodal infiltration.

Keywords

Colon cancer diffusion-weighted imaging apparent diffusion coefficient lymph node

Introduction

As the fourth most common form of cancer occurring worldwide, colorectal carcinoma is a central issue requiring better diagnostic and therapeutic options (1,2). Though neoadjuvant chemotherapy (NAC) for rectal cancer has been well established (3), there is still ongoing research in the use of NAC in the setting of colonic cancer (4,5). Several recent studies have shown promising results using NAC in colon cancer before surgery (6 –10). There is a need to improve preoperative non-invasive diagnostics using techniques that provide more accurate staging information in assessing patient eligibility for NAC (11).

Thorax, abdominal, and pelvic multi-slice computed tomography (MSCT) imaging, a standard procedure for the assessment of colon cancer staging, means that the preoperative TNM status is based on morphological changes in the primary tumor and suspected distant infiltration (lymph node, visceral metastasis). However, morphological changes visible on MSCT imaging are not always linked to exact tumor infiltration. For example, lymph node enlargement is not always linked to metastatic lymph node infiltration. Moreover, assessments of the primary tumor grade and aggressive potential based on morphological changes are not necessarily reliable (11).

Magnetic resonance imaging (MRI) in colon cancer staging is not a routine option (12), given that MSCT has several important advantages, i.e. a fast, relatively cheap imaging method with high spatial resolution and powerful postprocessing options (13). However, unlike MSCT, functional MRI sequence offers additional information based on molecular functioning and tissue micro-architecture (14,15). One of the basic functional sequences are diffusion-weighted imaging (DWI) with the corresponding apparent diffusion coefficient (ADC) maps (16). Differences in Brownian movement are detected in this sequence based on differences in tissue cellularity, which is in line with the type of tumor, accumulation of macromolecules, changes in the number of cells, and the ratio of nucleus/cytoplasm. Furthermore, this information can be provided in metric units using ADC, thus facilitating a comparison of differences between tissues. Several studies have shown differences in the ADC corresponding to the type of metastatic tissue or primary tumor type (10,17 –20).

In a colon setting, DWI are the standard sequences in rectal cancer workup and in assessing the stage of inflammatory bowel disease (21 –26). Only a few studies have investigated the advantages of using DWI sequences for colon cancer (22,27).

Looking into the future, the development of molecular diagnostics for cancers and new drug treatments targeting specific types of tumors, and any non-invasive approach to assuming tumor characteristics seems to be an interesting step forward. If excluding rectal carcinoma, studies on investigating grade of colon cancer seem to be lacking, which is an independent factor of prognosis, based on DWI/ADC tumor characteristics.

The aim of the present study was to investigate the link between the tumor grade (pathohistological confirmed) and the N status (corresponding to lymph node involvement) with ADC values.

Material and Methods

Patients and the MRI protocol

The study complies with the Declaration of Helsinki and was approved by the Ethics Committee at the University Hospital Split (Class: 500-03/20-01/24; Reference no.: 2181-147-01/06/M.S.-20-6). All enrolled patients signed an informed consent form.

A total of 35 patients planned for surgical resection had a biopsy confirming colorectal carcinoma over one year (November 2019 to November 2020) and were considered for participating in the study. In all, 20 patients were registered for participation after excluding those with rectal carcinomas, distant metastases confirmed by preoperative staging MSCT imaging, and pathohistologically confirmed mucinous adenocarcinoma. Since there are significant differences between mucinous and non-mucinous adenocarcinomas (28), only patients with pathohistologically confirmed non-mucinous adenocarcinomas were included in the final analysis. If the pathohistological report confirmed a mucinous adenocarcinoma, patient data were excluded for further analysis.

The MRI scan was scheduled on the day before the planned surgery. The patients were required to fast overnight, and each received an intravenous bolus equivalent to 20 mg of hyoscine butylbromide (Buscopan®; Boehringer Ingelheim BV, Ingelheim, Germany) before scanning. Where appropriate, scanning was performed using the breath-hold technique. Patients were scanned in the supine position using 1.5-T MRI (Avanto; Siemens Healthineers, Erlangen, Germany) and phased array body coil for imaging. The protocol consisted of a localizer, T2-weighted turbo spin echo (TSE) in a 1 axial and 1 coronal stack for anatomic orientation, and DWI sequence acquired in 3 stacks – b0, b800, and b1000. According to Nerad et al. (27), the echo time (TE) and repetition time (TR) for the TSE sequence were 80 ms and 5600 ms, respectively. The slice thickness was 3 mm, with a minimal slice gap of 3 mm and a field of view (FOV) equivalent to 390 × 390 mm. The acquisition matrix was 392 × 392 with an acquisition voxel size of 0.99 × 0.99 × 3.00 mm. For the DWI TE at 65 ms and TR at 3800 ms, the slice thickness was 8 mm, and the minimal slice gap was 0 mm with a FOV of 380 × 280 mm, where the acquisition matrix was 152 × 115 with an acquisition voxel size of 2.50 × 2.51 × 8.00 mm. The number of excitations was 4. The operating system automatically generated ADC maps using a mono-exponential decay model for all three b-values. The total acquisition time was 10 min.

The surgical procedure and intrahospital treatment were performed according to standard clinical routines as set by the Department of Abdominal Surgery.

A pathohistological analysis was performed in line with the standard procedure. All samples were examined by the same experienced gastrointestinal pathologist who was blinded to MRI findings.

Image evaluation and statistic

Statistical analysis was performed using the Statistica 8.0 software package (StatSoft, Tulsa, OK, USA). Where appropriate, descriptive statistics were applied.

Due to pronounced motion artifacts, three patients were excluded. Hence, the data for 17 patients were included in the final statistics.

Two senior residents in radiology were blinded to the postoperative pathohistological report and independently analyzed the recorded MRI scans using a Syngo.via (Siemens Healthcare, Forchheim, Germany) equipped with a monitor (1600 × 1200 pixel resolution). Both were supervised by a senior radiologist with 15 years of experience in abdominal MRI. The tumor and all visible abdominal lymph nodes were manually delineated directly on a grayscale ADC map for every single slice and detected to access the total tumor and summarized lymph node volume. The mean ADC value was further calculated for the mean tumor and mean lymph node values.

After obtaining the raw data, an inter-observer variability analysis was performed. An intraclass correlation coefficient (ICC) was used to assess the agreement between measurements of a continuous variable for two observers as a single measure of the extent of agreement. Subsequently, an analysis of the Bland–Altman plot provided a visual and quantitative reliability estimate.

Finally, after obtaining the definitive pathohistological report, serving as a gold standard, patients were allocated into two subgroups for further analysis: “low” or “high” (based on the tumor grade), and the N– or N+ group (based on pathohistologically discovered malign lymph nodes).

The Mann–Whitney U test compared differences in metric parameters (ADC values) between the two patient groups. Both the U value and Z score were calculated. The two-sided P value <0.05 was considered significant.

Results

General observation

There were 17 patients (11 women, 6 men; mean age = 73 ± 9 years; age range = 57–85 years; median age = 73 years) in total. In all, 10 (59%) tumors were identified in the ascending colon, 3 (18%) in the transversal colon, and 4 (23%) on the descending colon. All tumors were non-mucinous adenocarcinomas, with 10 (59%) low grades and 7 (41%) high grades. The mean number of isolated lymph nodes identified in the pathohistological examination was 17 ± 6 (median = 18, range = 9–37). A total of 9 (53%) patients had a positive lymph node, whereas 8 (47%) were without tumor infiltration, as was evident at the pathohistological exam.

A representative figure of MRI analysis is given in Fig. 1.

Inter-observer agreement

For the primary tumor ADC value, the consistency of the measurements ICC was 0.863 (95% confidence interval [CI] = 0.621–0.950), whereas the absolute agreement for ICC between the two raters was 0.836 (95% CI = 0.519–0.942). The coefficient of variation was 8.04% (95% CI = 5.04–11.12) (Supplementary Material, Fig. 1(s)).

The Bland–Altman analysis failed to detect a significant proportional difference between the raters. The proportional difference was 4.76% (95% CI = −0.44 to 9.97; P = 0.0700) (Supplementary Material, Fig. 2(s)).

For the lymph node ADC value, the consistency of the measurements was 0.840 (95% CI = 0.502–0.949), whereas the absolute agreement between the two raters was 0.850 (95% CI = 0.521–0.952). The coefficient of variation was 13.52% (95% CI = 7.79–19.55) (Supplementary Material, Fig. 3(s)).

Based on proportional measurements, the Bland–Altman analysis for the lymph node ADC value failed to detect a significant proportional difference between the raters. The proportional difference was −2.26% (95% CI = –12.8 to 8.31; P = 0.652) (Supplementary Material, Fig. 4(s)).

ADC values and the tumor grade

Low-grade tumors had a mean ADC equivalent to 1225 ± 170×10⁻⁶ mm²/s, and the coefficient of high-grade tumors was 1444 ± 69×10⁻⁶ mm²/s (Fig. 2). The U value is 7, while the Z score is −2.68373. The P value is 0.007.

Lymph node infiltration and the ADC value

The group of patients with positive lymph nodes in operative tissue samples (N+) exhibited lower mean ADC values (1023 ± 142×10⁻⁶ mm²/s) as opposed to the group without metastatic lymph nodes (N–) with ADC values of 1260 ± 231×10⁻⁶ mm²/s (Fig. 3). The U value is 12, and Z score is −2.01117. The P value is 0.04.

Discussion

The main finding of this study suggests that the ADC value is associated with the histologic grade and nodal stage of colon cancer.

Numerous studies have pointed out the advantages of using MRI in the preoperative assessment of T status due to better soft tissue discrimination (12,15,27,29). Most research has investigated DWI/ADC presentation of rectal cancer (30 –33). Some have shown that ADC values can serve as a predictor of NAC response in rectal cancer (34) or a predictor of biological features of tumors (33). To the best of our knowledge, this is the first study investigating the tumor histological grade and ADC values in the setting of colon cancer. Our results suggest that the mean whole volume primary tumor ADC coefficient is lower for low-grade tumors and higher for high-grade tumors. Although the study provided by Nerad et al. (27) showed that a locally advanced tumor, expressed as a T3–T4 status, has a lower ADC value, the histological tumor grade is independent of the T status. A non-invasive assessment of tumor grade can be an important stage-independent prognostic variable, and a high grade is associated with poor patient survival (35). Another study conducted by Metin et al. (36) investigated the relationship between the histopathologic grade of colorectal adenocarcinomas in liver metastases and ADC values. They showed that poorly differentiated adenocarcinomas have the lowest ADC values among other groups. This is contrary to our findings. Tissue differences that cause differences in cellularity as a consequence of Brownian movement between the primary colon tumor in our case and visceral (especially hepatic) metastases in the latter case could be an explanation of the observed variability. In addition, the level of movement disruption of the presence of water at the cellular level correlates with tumor aggressiveness (37) and is different between original tumorous and metastatic tissue. The recently published study provided by Rafaelsen et al. (5) has shown that the diffusion restriction was significantly lower in mucin-containing colonic tumors, where the mean primary tumor ADC values tended to be lower in patients with distant metastases than in those without it. Though the data in our study relate to non–mucinous adenocarcinomas, differences in the amount of the mucin component (range = 0%–50%) (38), is a possible clue for unexpected findings related to the tumor grade. This finding should be addressed in further studies.

Finally, the observed differences may be due to different measurement methods. We measured the whole-volume mean ADC, representing the overall tumor profile and not just a single slice (2-D) ROI used in other studies (22,33).

Although the MRI staging of colon cancer has several advantages and a higher specificity and sensitivity in assessing the progression of local tumors than MSCT staging, the assessment of nodal infiltration (N status) is not sufficiently reliable (15,16). Given that lymph node enlargement is not a reliable indicator of nodal metastasis in patients with colon cancer, the challenge remains in devising an imaging method to more accurately predict nodal infiltration (22,39,40). A reactive lymph node can also be presented by enlargement as a node infiltrated with malign cells. In contrast, a normal-sized lymph node can contain microscopic metastasis (22). There is a lack of studies directly measuring the lymph node ADC in a colon cancer setting. A recent study reported that the ADC value for primary colon tumors can predict a positive nodal status (27). The prediction of lymph node involvement was based on morphological criteria in the same study. Our study measured the whole-volume mean ADC value of all visible abdominal lymph nodes. This approach seems reasonable because lymph node enlargement is not always a result of tumor infiltration. Another study investigating the ADC value of reactive, metastatic, and lymphomatous neck lymph nodes showed that metastatic infiltration produced lower ADC values than benign or reactive lymph nodes (41). This outcome agrees with our results, i.e. the patient group with nodal metastasis had a lower mean ADC value than patients without lymph node metastasis. Similar results in a head and neck lesion setting were reported by Wang et al. (42) and Chen et al. (43).

A small number of studies have assessed the reproducibility of measured ADC values in extracranial organs (22). Previous studies have shown that ADC values obtained from the whole-tumor volume provide the most reproducible results compared to single region of interest-based tumor assessment (44,45). Using the whole-volume measurement as a methodological approach, we have shown good inter-observer compatibility for primary tumor and whole-volume lymph node ADC measurements. The coefficient of variation measurement was used to express a measure of agreement when repeated measurements are performed on the same patients by two different observers. The study provided by Anvary et al. (46) shows that the coefficient of variation <10% means excellent inter-observer agreement and a coefficient of variation in the range of 10%–20% good inter-observer agreement. With the coefficient of variation of 8.04% for primary tumors and 13% for lymph nodes, this result is acceptable. A recent study provided by Rosa et al. (47) also showed that the DWI/ADC results had a high level of agreement between observers in rectal tumors.

The main limitation of this study is the small number of participants. An optimal result requires a larger sample size. Accordingly, one of the ideas in planning the study was to measure the ADC value of a single lymph node with clear anatomic references, extracted during surgery and individually analyzed by a pathologist for signs of metastatic involvement. Unfortunately, this is difficult to apply in practice. It was achieved in only four patients, among whom two had malignant lymph nodes with ADC values equivalent to 590 and 852×10⁻⁶ mm²/s, respectively, while two other isolated lymph nodes were reactive with ADC values of 1140 and 1179×10⁻⁶ mm²/s, respectively.

In conclusion, the mean whole-tumor ADC is associated with the histological tumor grade, and the mean ADC value of whole-volume abdominal lymph nodes could assume real nodal infiltration. Nonetheless, further research is necessary.

Supplemental Material

sj-docx-1-acr-10.1177_02841851221130008 - Supplemental material for Estimation of colon cancer grade and metastatic lymph node involvement using DWI/ADC sequences

Supplemental material, sj-docx-1-acr-10.1177_02841851221130008 for Estimation of colon cancer grade and metastatic lymph node involvement using DWI/ADC sequences by Cavar Marija, Dolic Kresimir, Barcot Ognjen, Peric Iva, Kunac Nenad and Boric Matija in Acta Radiologica

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Marija Cavar

Supplemental material

Supplemental material for this article is available online.

References

Boyle

Langman

. ABC of colorectal cancer: epidemiology. BMJ 2000;321:805–808.

Rawla

Sunkara

Barsouk

. Epidemiology of colorectal cancer: incidence, mortality, survival, and risk factors. Prz Gastroenterol 2019;14:89–103.

Sninsky

Shore

Lupu

, et al. Risk factors for colorectal polyps and cancer. Gastrointest Endosc Clin N Am 2022;32:195–213.

Quezada-Diaz

Smith

. Neoadjuvant therapy for rectal cancer. Surg Oncol Clin N Am 2022;31:279–291.

Rafaelsen

Dam

Vagn-Hansen

, et al. CT and 3 Tesla MRI in the TN staging of colon cancer: a prospective, blind study. Curr Oncol 2022;29:1069–1079.

Arredondo

Baixauli

Pastor

, et al. Mid-term oncologic outcome of a novel approach for locally advanced colon cancer with neoadjuvant chemotherapy and surgery. Clin Transl Oncol 2017;19:379–385.

Cheong

Nistala

KRY

, et al. Neoadjuvant therapy in locally advanced colon cancer: a meta-analysis and systematic review. J Gastrointest Oncol 2020;11:847–857.

Jakobsen

Andersen

Fischer

, et al. Neoadjuvant chemotherapy in locally advanced colon cancer. A phase II trial. Acta Oncol 2015;54:1747–1753.

Saltz

. Top advances of the year in colorectal cancer. Cancer 2022;128:2236–2239.

10.

Jacobs

Macura

Zaheer

, et al. Multiparametric whole-body MRI with diffusion-weighted imaging and ADC mapping for the identification of visceral and osseous metastases from solid tumors. Acad Radiol 2018;25:1405–1414.

11.

Kim

Huh

Lee

, et al. Oncologic outcomes of pathologic T4 and T3 colon cancer patients diagnosed with clinical T4 stage disease using preoperative computed tomography scan. Surg Oncol 2022;41:101749.

12.

Nerad

Lambregts

Kersten

, et al.

MRI for local staging of colon cancer: can MRI become the optimal staging modality for patients with colon cancer?

Dis Colon Rectum 2017;60:385–392.

13.

Bai

Ren

Jiang

, et al. Accuracy of multi-slice spiral computed tomography for preoperative tumor node metastasis (TNM) staging of colorectal carcinoma. Med Sci Monit 2017;23:3470–3479.

14.

Baliyan

Das

. Sharma

, et al. Diffusion weighted imaging: technique and applications. World J Radiol 2016;8:785–798.

15.

Rollven

Holm

Glimelius

, et al. Potentials of high resolution magnetic resonance imaging versus computed tomography for preoperative local staging of colon cancer. Acta Radiol 2013;54:722–730.

16.

Hunter

Blake

Jeyadevan

, et al. Local staging and assessment of colon cancer with 1.5-T magnetic resonance imaging. Br J Radiol 2016;89:20160257.

17.

Duygulu

Ovali

Calli

, et al. Intracerebral metastasis showing restricted diffusion: correlation with histopathologic findings. Eur J Radiol 2010;74:117–120.

18.

Koyama

Ohno

Nishio

, et al. Diffusion-weighted imaging vs STIR turbo SE imaging: capability for quantitative differentiation of small-cell lung cancer from non-small-cell lung cancer. Br J Radiol 2014;87:20130307.

19.

Liu

, et al. Usefulness of diffusion-weighted MR imaging in the evaluation of pulmonary lesions. Eur Radiol 2010;20:807–815.

20.

Ulu

Ozturk

Atalay

, et al. Diffusion-weighted imaging of brain metastasis: correlation of MRI parameters with histologic type. Turk Neurosurg 2022;32:58–68.

21.

Caruso

Angriman

Scarpa

, et al. Diffusion-weighted magnetic resonance for assessing fibrosis in Crohn's disease. Abdom Radiol (NY) 2020;45:2327–2335.

22.

Dam

Lindebjerg

Jakobsen

, et al. Local staging of sigmoid colon cancer using MRI. Acta Radiol Open 2017;6:2058460117720957.

23.

Hordonneau

Buisson

Scanzi

, et al. Diffusion-weighted magnetic resonance imaging in ileocolonic Crohn’s disease: validation of quantitative index of activity. Am J Gastroenterol 2014;109:89–98.

24.

Sun

Mao

, et al. Diffusion-weighted MRI enables to accurately grade inflammatory activity in patients of ileocolonic Crohn's disease: results from an observational study. Inflamm Bowel Dis 2017;23:244–253.

25.

Sun

Mao

, et al. Assessment of activity of Crohn disease by diffusion-weighted magnetic resonance imaging. Medicine (Baltimore) 2015;94:e1819.

26.

Yang

Zhang

, et al. Diffusion-weighted magnetic resonance imaging without bowel preparation for detection of ulcerative colitis. World J Gastroenterol 2015;21:9785–9792.

27.

Nerad

Delli Pizzi

Lambregts

DMJ

, et al. The apparent diffusion coefficient (ADC) is a useful biomarker in predicting metastatic colon cancer using the ADC-value of the primary tumor. PLoS One 2019;14:e0211830.

28.

Nasu

Kuroki

Minami

. Diffusion-weighted imaging findings of mucinous carcinoma arising in the ano-rectal region: comparison of apparent diffusion coefficient with that of tubular adenocarcinoma. Jpn J Radiol 2012;30:120–127.

29.

Lambregts

DMJ

Maas

Boellaard

, et al. Long-term imaging characteristics of clinical complete responders during watch-and-wait for rectal cancer-an evaluation of over 1500 MRIs. Eur Radiol 2020;30:272–280.

30.

Curvo-Semedo

Lambregts

Maas

, et al. Diffusion-weighted MRI in rectal cancer: apparent diffusion coefficient as a potential noninvasive marker of tumor aggressiveness. J Magn Reson Imaging 2012;35:1365–1371.

31.

Delli Pizzi

Chiarelli

Chiacchiaretta

, et al. MRI-based clinical-radiomics model predicts tumor response before treatment in locally advanced rectal cancer. Sci Rep 2021;11:5379.

32.

Haak

Maas

Lahaye

, et al. Selection of patients for organ preservation after chemoradiotherapy: MRI identifies poor responders who can go straight to surgery. Ann Surg Oncol 2020;27:2732–2739.

33.

Sun

Tong

Cai

, et al. Apparent diffusion coefficient (ADC) value: a potential imaging biomarker that reflects the biological features of rectal cancer. PLoS One 2014;9:e109371.

34.

Engin

Sharifov

Gural

, et al.

Can diffusion-weighted MRI determine complete responders after neoadjuvant chemoradiation for locally advanced rectal cancer?

Diagn Interv Radiol 2012;18:574–581.

35.

Fleming

Ravula

Tatishchev

, et al. Colorectal carcinoma: pathologic aspects. J Gastrointest Oncol 2012;3:153–173.

36.

Metin

Aydin

Cetin

, et al. The efficacy of diffusion weighted imaging and apparent diffusion coefficients mapping for liver metastasis of colonic adenocarcinomas. Saudi Med J 2016;37:379–385.

37.

Kargol

Rudnicki

Kenig

, et al. Diffusion-weighted magnetic resonance imaging of 103 patients with rectal adenocarcinoma identifies the apparent diffusion coefficient as an imaging marker for tumor invasion and regional lymph node involvement. Med Sci Monit 2021;27:e934941.

38.

Luo

Cen

Ding

, et al. Mucinous colorectal adenocarcinoma: clinical pathology and treatment options. Cancer Commun (Lond) 2019;39:13.

39.

Monig

Baldus

Zirbes

, et al. Lymph node size and metastatic infiltration in colon cancer. Ann Surg Oncol 1999;6:579–581.

40.

Monig

Schroder

Baldus

, et al. Preoperative lymph-node staging in gastrointestinal cancer--correlation between size and tumor stage. Onkologie 2002;25:342–344.

41.

Abdel Razek

Soliman

Elkhamary

, et al. Role of diffusion-weighted MR imaging in cervical lymphadenopathy. Eur Radiol 2006;16:1468–1477.

42.

Wang

Takashima

Takayama

, et al. Head and neck lesions: characterization with diffusion-weighted echo-planar MR imaging. Radiology 2001;220:621–630.

43.

Chen

Lin

Xiao

, et al. Role of diffusion-weighted imaging in the discrimination of benign and metastatic parotid area lymph nodes in patients with nasopharyngeal carcinoma. Sci Rep 2018;8:281.

44.

Jafar

Parsai

Miquel

. Diffusion-weighted magnetic resonance imaging in cancer: reported apparent diffusion coefficients, in-vitro and in-vivo reproducibility. World J Radiol 2016;8:21–49.

45.

Lambregts

Beets

Maas

, et al. Tumour ADC measurements in rectal cancer: effect of ROI methods on ADC values and interobserver variability. Eur Radiol 2011;21:2567–2574.

46.

Anvari

Halpern

Samir

. Essentials of statistical methods for assessing reliability and agreement in quantitative imaging. Acad Radiol 2018;25:391–396.

47.

Rosa

Caravatta

Delli Pizzi

, et al. Reproducibility of rectal tumor volume delineation using diffusion-weighted MRI: agreement on volumes between observers. Cancer Radiother 2019;23:216–221.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB