Preoperative Differentiation of HER2‐Zero and HER2‐Low from HER2‐Positive Invasive Ductal Breast Cancers Using BI‐RADS MRI Features and Machine Learning Modeling

Journal of Magnetic Resonance Imaging, Год журнала: 2024, Номер unknown

Опубликована: Май 10, 2024

Background Accurate determination of human epidermal growth factor receptor 2 (HER2) is important for choosing optimal HER2 targeting treatment strategies. HER2‐low currently considered HER2‐negative, but patients may be eligible to receive new anti‐HER2 drug conjugates. Purpose To use breast MRI BI‐RADS features classifying three levels, first distinguish HER2‐zero from HER2‐low/positive (Task‐1), and then HER2‐positive (Task‐2). Study Type Retrospective. Population 621 invasive ductal cancer, 245 HER2‐zero, 191 HER2‐low, 185 HER2‐positive. For Task‐1, 488 cases training 133 testing. Task‐2, 294 82 Field Strength/Sequence 3.0 T; 3D T1‐weighted DCE, short time inversion recovery T2, single‐shot EPI DWI. Assessment Pathological information were compared. Random Forest was used select features, four machine learning (ML) algorithms: decision tree (DT), support vector (SVM), k ‐nearest neighbors ( ‐NN), artificial neural nets (ANN), applied build models. Statistical Tests Chi‐square test, one‐way analysis variance, Kruskal–Wallis test performed. The P values <0.05 statistically significant. ML models, the generated probability construct ROC curves. Results Peritumoral edema, presence multiple lesions non‐mass enhancement (NME) showed significant differences. distinguishing non‐zero (low + positive), lesions, margin, tumor size selected, ‐NN model achieved highest AUC 0.86 in set 0.79 testing set. differentiating HER2‐positive, margin DT 0.69 Data Conclusion read by radiologists preoperative can analyzed using more sophisticated feature selection algorithms models classification status identify HER2‐low. Level Evidence 4. Technical Efficacy Stage 2.

Язык: Английский

---

Ultrasound Radiomics for the Prediction of Breast Cancers with HER2-Zero, -Low, and -Positive Status: A Dual-Center Study

Technology in Cancer Research & Treatment, Год журнала: 2024, Номер 23

Опубликована: Янв. 1, 2024

Purpose To assess whether gray-scale ultrasound (US) based radiomic features can help distinguish HER2 expressions (ie, HER2-overexpressing, HER2-low-expressing, and HER2-zero-expressing) in breast cancer. Materials Methods This retrospective study encompassed female cancer patients who underwent US examinations at two distinct centers from February 2021 to July 2023. Tumor segmentation feature extraction were performed on grayscale images. Decision Tree analysis was employed simultaneously evaluate importance, the Least Absolute Shrinkage Selection Operator technique utilized for selection construct signature. The Area Under Curve (AUC) of Receiver Operating Characteristic curve performance features. Multivariate logistic regression used identify independent predictors distinguishing expression dataset. Results training set comprised 292 Center 1 (median, 51 years; interquartile range [IQR]: 45-61), while external validation included 131 2 IQR: 45-62). In dataset, achieved AUC 0.76 between HER2-low positive tumors versus HER2-zero tumors. differentiating (1+) 0.74, (2+) tumors, 0.77. multivariate assessing HER2-positive internal echoes (P = .029) margins < .001) emerged as predictive factors. Conclusion signature tumor descriptors may predict cancers with therapeutic implications.

Язык: Английский

---

The Clinical Study of Intratumoral and Peritumoral Radiomics Based on DCE-MRI for HER-2 Positive and Low Expression Prediction in Breast Cancer

Breast Cancer Targets and Therapy, Год журнала: 2024, Номер Volume 16, С. 957 - 972

Опубликована: Дек. 1, 2024

Core biopsy sampling may not fully capture tumor heterogeneity. Radiomics provides a non-invasive method to assess characteristics, including both the core and surrounding tissue, with potential improve accuracy of HER-2 status prediction.

Язык: Английский

---

A radiomics model development via the associations with genomics features in predicting axillary lymph node metastasis of breast cancer: a study based on a public database and single-centre verification

Clinical Radiology, Год журнала: 2022, Номер 78(3), С. e279 - e287

Опубликована: Дек. 26, 2022

Язык: Английский

---

Integrating clinical data and ultrasonographic imaging for non-invasive prediction of HER2 status in breast cancer

Research Square (Research Square), Год журнала: 2024, Номер unknown

Опубликована: Март 14, 2024

Background The most common cancer in the world, breast (BC), poses serious problems to healthcare. Making an accurate diagnosis of these patients' HER2 status is essential for therapy planning.Methods A prospective cohort patients with BC was enrolled between June 2020 and october 2023. patient's clinical data features from their ultrasonography were gathered. Postoperative tumor pathology specimens subjected immunohistochemistry fluorescence situ hybridization examinations ascertain status. Lasso regression used choose characteristic variables. Univariate multivariate logistic analysis find status-independent factors. performance nomogram model then assessed using calibration curves decision curve (DCA).Result 97 (22.25%) 436 research had positive results. Progesterone receptor expression, Ki-67 levels, estrogen expression differed statistically amongst different statuses. identified six ultrasonographic variables closely associated a pool 786 features, leading generation radiomic score each patient. Multivariate revealed that PR (OR = 0.15, 95%CI 0.06–0.36, p < 0.001), 1.02, 1.00-1.03, 0.012), Radiomic 5.89, 2.58–13.45, 0.001) independent predictors demonstrated areas under (AUC) 0.823 (95% CI 0.772–0.874) 0.812 0.717–0.906) training validation cohort, respectively.Conclusions methodology integrates data, cutting-edge imaging, machine learning provide individualized treatment plans presented non-invasive prediction cancer.

Язык: Английский

---