Share：
Share this content in WeChat
X
[Chinese][PDF]321
• Special Focu •
Deep learning based on multiparametric magnetic resonance imaging features to predict BRAF gene mutation status in rectal cancer patients
HU Hongbo  ZHAO Sheng  JIANG Hao  ZHANG Ying  JIANG Huijie 

Cite this article as: HU H B, ZHAO S, JIANG H, et al. Deep learning based on multiparametric magnetic resonance imaging features to predict BRAF gene mutation status in rectal cancer patients[J]. Chin J Magn Reson Imaging, 2025, 16(1): 22-28. DOI:10.12015/issn.1674-8034.2025.01.004.


[Abstract] Objective The mutation status of the B-Raf proto-oncogene, serine/threonine kinase (BRAF) gene, a homolog of the murine sarcoma virus oncogene B, is related to the survival of patients with colorectal tumors. This study aims to explore the feasibility of using a radiomics model to predict BRAF gene mutations in colorectal cancer patients.Materials and Methods A retrospective analysis was conducted on the case data of patients diagnosed with rectal cancer at our institution from June 2020 to June 2023, utilizing exon sequencing to identify BRAF gene mutations. Survival analysis was performed to evaluate the relationship between BRAF mutations and prognosis in rectal cancer. From 260 patients with multi-parametric magnetic resonance imaging, 7,388 modules were extracted, including preoperative T1-weighted images (T1WI), T2-weighted images (T2WI), and contrast-enhanced T1-weighted images (CE-T1WI). Finally, a feature-based radiomics model was established using convolutional neural networks (ConvNet). The model's performance was evaluated using receiver operating characteristic (ROC) curves, accuracy, sensitivity, and specificity as metrics.Results The study included 89 patients with BRAF mutations and 171 patients with wild-type BRAF. There were no significant differences in clinical characteristics such as tumor malignancy staging, age, and sex between the two groups (P > 0.05); however, a significant difference was observed in the 5-year survival rates. The survival duration of the BRAF mutation group was significantly lower than that of the wild-type group (P < 0.001). The area under the ROC curve for the predictive model was 0.929, The Kappa statistic for the consistency analysis with pathological results was 0.87, indicating good predictive value.Conclusions The radiomics model constructed using convolutional neural networks can effectively distinguish BRAF mutation status in rectal cancer patients, providing new insights for non-invasive screening of BRAF status in the future.
[Keywords] rectal cancer；magnetic resonance imaging；image features；deep learning；B-Raf proto-oncogene serine/threonine kinase；convolutional neural network；radiomics model

HU Hongbo1   ZHAO Sheng1   JIANG Hao1   ZHANG Ying2   JIANG Huijie1*  

1 Department of Radiology, the Second Affiliated Hospital of Harbin Medical University, Harbin 150086, China

2 Pharmacology Teaching and Research Office, College of Pharmacy Harbin Medical University, Harbin 150086, China

Corresponding author: JIANG H J, E-mail: jhjemail@163.com

Conflicts of interest   None.

Received  2024-08-21
Accepted  2024-12-10
DOI: 10.12015/issn.1674-8034.2025.01.004
Cite this article as: HU H B, ZHAO S, JIANG H, et al. Deep learning based on multiparametric magnetic resonance imaging features to predict BRAF gene mutation status in rectal cancer patients[J]. Chin J Magn Reson Imaging, 2025, 16(1): 22-28. DOI:10.12015/issn.1674-8034.2025.01.004.

[1]
SUTTON T S, HAO S, SUZUKI M, et al. Rectal cancer presentation during the COVID-19 pandemic: are decreasing screening rates leading to an increase in acute presentations?[J/OL]. PLoS One, 2023, 18(9): e0291447 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37708208/. DOI: 10.1371/journal.pone.0291447.
[2]
WANG J W, ZHANG L J, WANG M H, et al. Long-term outcomes in a retrospective cohort of patients with rectal cancer with complete response after total neoadjuvant therapy: a propensity-score weighted analysis[J/OL]. Ther Adv Med Oncol, 2023, 15: 17588359231197955 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37701810/. DOI: 10.1177/17588359231197955.
[3]
ZHANG Z W, CHEN Y, WEN Z Q, et al. MRI for nodal restaging after neoadjuvant therapy in rectal cancer with histopathologic comparison[J/OL]. Cancer Imag, 2023, 23(1): 67 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37443085/. DOI: 10.1186/s40644-023-00589-0.
[4]
CIARDIELLO D, NAPOLITANO S, FAMIGLIETTI V, et al. Pretreatment plasma circulating tumor DNA RAS/BRAF mutational status in refractory metastatic colorectal cancer patients who are candidates for anti-EGFR rechallenge therapy: a pooled analysis of the CAVE and VELO clinical trials[J/OL]. Cancers, 2023, 15(7): 2117 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37046778/. DOI: 10.3390/cancers15072117.
[5]
ERBEN P, STRÖBEL P, HORISBERGER K, et al. KRAS and BRAF mutations and PTEN expression do not predict efficacy of cetuximab-based chemoradiotherapy in locally advanced rectal cancer[J]. Int J Radiat Oncol, 2011, 81(4): 1032-1038. DOI: 10.1016/j.ijrobp.2010.06.043.
[6]
FAN Y Q, YU J J, ZHAO M. Metanephric stromal tumor with BRAF V600E mutation in an adult patient: case report and literature review[J/OL]. Front Oncol, 2022, 12: 993414 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/36276087/. DOI: 10.3389/fonc.2022.993414.
[7]
GORODEZKI D, ZIPFEL J, QUEUDEVILLE M, et al. Resection extent and BRAF V600E mutation status determine postoperative tumor growth velocity in pediatric low-grade glioma: results from a single-center cohort analysis[J]. J Neurooncol, 2022, 160(3): 567-576. DOI: 10.1007/s11060-022-04176-4.
[8]
KONO H, YAMANAKA T, NISHIHARA Y, et al. BRAF mutation heterogeneity detected using circulating tumor DNA sequencing in synchronous colon cancer: a case report[J]. Cancer Diagn Progn, 2023, 3(5): 605-608. DOI: 10.21873/cdp.10262.
[9]
HAMADA K, KURASHIGE T, SHIMAMURA M, et al. MIEAP and ATG5 are tumor suppressors in a mouse model of BRAFV600E-positive thyroid cancer[J/OL]. Front Endocrinol, 2022, 13: 932754 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/36187114/. DOI: 10.3389/fendo.2022.932754.
[10]
HONG S, JEON M, KWON J, et al. Targeting RAF isoforms and tumor microenvironments in RAS or BRAF mutant colorectal cancers with SJ-C1044 for anti-tumor activity[J]. Curr News Mol Biol, 2023, 45(7): 5865-5878. DOI: 10.3390/cimb45070371.
[11]
CRISTOFANI R, PICCOLELLA M, MONTAGNANI MARELLI M, et al. HSPB8 counteracts tumor activity of BRAF- and NRAS-mutant melanoma cells by modulation of RAS-prenylation and autophagy[J/OL]. Cell Death Dis, 2022, 13(11): 973 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/36400750/. DOI: 10.1038/s41419-022-05365-9.
[12]
KARBHARI A, BAHETI A D, ANKATHI S K, et al. MRI in rectal cancer patients on 'watch and wait': patterns of response and their evolution[J]. Abdom Radiol (NY), 2023, 48(11): 3287-3296. DOI: 10.1007/s00261-023-04003-y.
[13]
KE J, JIN C, TANG J H, et al. A longitudinal MRI-based artificial intelligence system to predict pathological complete response after neoadjuvant therapy in rectal cancer: a multicenter validation study[J/OL]. Dis Colon Rectum, 2023, 66(12): e1195-e1206 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37682775/. DOI: 10.1097/DCR.0000000000002931.
[14]
KIKANO E G, MATALON S A, ESKIAN M, et al. Concordance of MRI with pathology for primary staging of rectal cancer in routine clinical practice: a single institution experience[J]. Curr Probl Diagn Radiol, 2024, 53(1): 68-72. DOI: 10.1067/j.cpradiol.2023.08.016.
[15]
CHEN Y, DING L, ZHANG Z W, et al. Role of dynamic contrast-enhanced MRI in predicting severe acute radiation-induced rectal injury in patients with rectal cancer[J]. Eur Radiol, 2024, 34(3): 1471-1480. DOI: 10.1007/s00330-023-10194-1.
[16]
PIKŪNIENĖ I, SALADŽINSKAS Ž, BASEVIČIUS A, et al. MRI evaluation of rectal cancer lymph node staging using apparent diffusion coefficient[J/OL]. Cureus, 2023, 15(9): e45002 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37701166/. DOI: 10.7759/cureus.45002.
[17]
MEYER H J, HÖHN A, SUROV A. Histogram analysis of ADC in rectal cancer: associations with different histopathological findings including expression of EGFR, Hif1-alpha, VEGF, p53, PD1, and KI 67. A preliminary study[J]. Oncotarget, 2018, 9(26): 18510-18517. DOI: 10.18632/oncotarget.24905.
[18]
XIANG S, ZHENG L B, ZHU L, et al. Radiomics-based prediction of microsatellite instability in stage Ⅱ and Ⅲ rectal cancer patients based on T2WI MRI and diffusion-weighted imaging[J]. Chin J Surg, 2023, 61(9): 782-787. DOI: 10.3760/cma.j.cn112139-20230315-00106.
[19]
ZHANG G W, CHEN L, LIU A E, et al. Comparable performance of deep learning-based to manual-based tumor segmentation in KRAS/NRAS/BRAF mutation prediction with MR-based radiomics in rectal cancer[J/OL]. Front Oncol, 2021, 11: 696706 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/34395262/. DOI: 10.3389/fonc.2021.696706.
[20]
ZHAO M N, FENG L L, ZHAO K, et al. An MRI-based scoring system for pretreatment risk stratification in locally advanced rectal cancer[J]. Br J Cancer, 2023, 129(7): 1095-1104. DOI: 10.1038/s41416-023-02384-x.
[21]
ZHOU L Q, YU G Y, WEN R B, et al. Neoadjuvant chemoradiation therapy combined with immunotherapy for microsatellite stable ultra-low rectal cancer (CHOICE II): study protocol of a multicentre prospective randomised clinical trial[J/OL]. BMJ Open, 2023, 13(9): e069793 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37709314/. DOI: 10.1136/bmjopen-2022-069793.
[22]
AMINTAS S, GIRAUD N, FERNANDEZ B, et al. The crying need for a better response assessment in rectal cancer[J]. Curr Treat Options Oncol, 2023, 24(11): 1507-1523. DOI: 10.1007/s11864-023-01125-9.
[23]
BOYLE J M, VAN DER MEULEN J, KURYBA A, et al. What is the impact of hospital and surgeon volumes on outcomes in rectal cancer surgery?[J]. Colorectal Dis, 2023, 25(10): 1981-1993. DOI: 10.1111/codi.16745.
[24]
HE W M, SUN Y J, GE J W, et al. STRA6 regulates tumor immune microenvironment and is a prognostic marker in BRAF-mutant papillary thyroid carcinoma[J/OL]. Front Endocrinol, 2023, 14: 1076640 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/36843593/. DOI: 10.3389/fendo.2023.1076640.
[25]
LI H, ZHANG Y C, XU Y J, et al. Tumor immune microenvironment and immunotherapy efficacy in BRAF mutation non-small-cell lung cancer[J/OL]. Cell Death Dis, 2022, 13(12): 1064 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/36543792/. DOI: 10.1038/s41419-022-05510-4.
[26]
SCLAFANI F, CHAU I, CUNNINGHAM D, et al. KRAS and BRAF mutations in circulating tumour DNA from locally advanced rectal cancer[J/OL]. Sci Rep, 2018, 8(1): 1445 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/29362371/. DOI: 10.1038/s41598-018-19212-5.
[27]
BI Q, QIN K, ZHANG H, et al. Local semantic enhanced ConvNet for aerial scene recognition[J]. IEEE Trans Image Process, 2021, 30: 6498-6511. DOI: 10.1109/TIP.2021.3092816.
[28]
RUDD-ORTHNER R N M, MIHAYLOVA L. Deep ConvNet: non-random weight initialization for repeatable determinism, examined with FSGM[J/OL]. Sensors, 2021, 21(14): 4772 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/34300512/. DOI: 10.3390/s21144772.
[29]
YANG T, ZHU S J, MENDIETA M, et al. MutualNet: adaptive ConvNet via mutual learning from different model configurations[J]. IEEE Trans Pattern Anal Mach Intell, 2023, 45(1): 811-827. DOI: 10.1109/TPAMI.2021.3138389.
[30]
ZHAO W, XU H, ZHAO R, et al. MRI-based radiomics model for preoperative prediction of lateral pelvic lymph node metastasis in locally advanced rectal cancer[J]. Acad Radiol, 2024, 31(7): 2753-2772. DOI: 10.1016/j.acra.2023.07.016.
[31]
YANG C C, LIN L C, LIN Y W, et al. Higher nuclear EGFR expression is a better predictor of survival in rectal cancer patients following neoadjuvant chemoradiotherapy than cytoplasmic EGFR expression[J]. Oncol Lett, 2019, 17(2): 1551-1558. DOI: 10.3892/ol.2018.9756.
[32]
YASUDA H, TANAKA K, SAIGUSA S, et al. Elevated CD133, but not VEGF or EGFR, as a predictive marker of distant recurrence after preoperative chemoradiotherapy in rectal cancer[J]. Oncol Rep, 2009, 22(4): 709-717. DOI: 10.3892/or_00000491.
[33]
ZOU M J, AL-YAHYA S, AL-ALWAN M, et al. β-catenin attenuation leads to up-regulation of activating NKG2D ligands and tumor regression in BrafV600E-driven thyroid cancer cells[J/OL]. Front Immunol, 2023, 14: 1171816 [2024-08-20]. https://pubmed.ncbi.nlm.nih.gov/37483610/. DOI: 10.3389/fimmu.2023.1171816.

PREV Radiology in assessment of neoadjuvant treatment efficacy in rectal cancer
NEXT Comparative study on the application of CT grayscale inversion and MRI in the MRF of rectal cancer
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn