Optimizing breast cancer diagnosis with photoacoustic imaging: An analysis of intratumoral and peritumoral radiomics

Photoacoustics, Journal Year: 2024, Volume and Issue: 38, P. 100606 - 100606

Published: April 9, 2024

The differentiation between benign and malignant breast tumors extends beyond morphological structures to encompass functional alterations within the nodules. combination of photoacoustic (PA) imaging radiomics unveils insights intricate details that are imperceptible naked eye. This study aims assess efficacy PA in cancer radiomics, focusing on impact peritumoral region size radiomic model accuracy. From January 2022 November 2023, data were collected from 358 patients with nodules, diagnosed via PA/US examination classified as BI-RADS 3-5. used largest lesion dimension images define interest, expanded by 2 mm, 5 8 for extracting features. Techniques statistics machine learning applied feature selection, logistic regression classifiers build models. These models integrated both intratumoral data, regressions identifying key predictive developed nomogram, combining mm clinical features, showed superior diagnostic performance, achieving an AUC 0.950 training cohort 0.899 validation. outperformed those based solely features or other methods, proving most effective research demonstrates significant potential especially advantage integrating approach not only surpasses but also underscores importance comprehensive analysis accurately characterizing

Language: Английский

---

Development of a nomogram-based model combining intra- and peritumoral ultrasound radiomics with clinical features for differentiating benign from malignant in Breast Imaging Reporting and Data System category 3–5 nodules

Quantitative Imaging in Medicine and Surgery, Journal Year: 2023, Volume and Issue: 13(10), P. 6899 - 6910

Published: Sept. 25, 2023

The differences in benign and malignant breast tumors are not only within the nodules but also involve changes surrounding tissues. Radiomics can reveal many details that discernible to naked eye. This study aimed distinguish between using an ultrasound-based intra- peritumoral radiomics model.This retrospectively collected information from 379 patients with Breast Imaging Reporting Data System (BI-RADS) category 3-5 clear pathological diagnosis of screened by routine ultrasound examination Sixth People's Hospital Affiliated Medical College Shanghai Jiao Tong University January 2017 December 2022. largest dimension lesion on 2D image was selected outline area interest which conformally outwardly expanded automatically 5 mm extract peritumor features. included cases were randomly divided into training sets test a ratio 7:3. optimal features models retained statistical machine learning methods dimensionality reduction, logistic regression used as classifier build intratumoral model combined intratumoral-peritumoral model, respectively; through single-factor multifactor regression, could predict screened. clinical imaging established selecting independent risk factors univariate multifactorial regression.Among BI-RADS nodules, there 124 255 nodules; aged 14 88 (46.22±15.51) years, age differences, score, mass diameter statistically significant (P>0.05). had under curve (AUC) 0.840 [95% confidence interval (CI): 0.766-0.914] set. AUC value 0.960 (95% CI: 0.920-0.999).The nomogram, developed features, demonstrated superior performance distinguishing lesions.

Language: Английский

---

Machine-learning models for differentiating benign and malignant breast masses: Integrating automated breast volume scanning intra-tumoral, peri-tumoral features, and clinical information

Digital Health, Journal Year: 2025, Volume and Issue: 11

Published: April 1, 2025

Background Differentiating between benign and malignant breast masses is critical for clinical decision-making. Automated volume scanning (ABVS) provides high-resolution three-dimensional imaging, addressing the limitations of conventional ultrasound. However, impact peritumoral region size on predictive performance has not been systematically studied. This study aims to optimize diagnostic by integrating radiomics features data using multiple machine-learning models. Methods retrospective included ABVS images from 250 patients with masses. Radiomics were extracted both intratumoral regions (5, 10, 20 mm). These features, combined data, used develop models based four algorithms: Support vector machine, random forest, extreme gradient boosting, light boosting machine (LGBM). Model was evaluated area under receiver operating characteristic curve (AUC), calibration curves, decision SHapley Additive exPlanations (SHAP) analysis employed interpretability. Results The inclusion improved varying degrees, model incorporating a 10 mm achieving highest overall accuracy. Combining further enhanced performance. LGBM outperformed other algorithms across subgroups, maximum AUC 0.909, an accuracy 0.878, F1-score 0.971. SHAP revealed contribution key improving Conclusion demonstrates value mass diagnosis, optimized enhancing emerged as preferred algorithm due its superior findings provide strong support application imaging future multicenter studies, highlighting importance microenvironmental in diagnosis.

Language: Английский

---

Multi-center strain elastography radiomics for breast cancer diagnosis: integrating intratumoral and peritumoral regions

Discover Oncology, Journal Year: 2025, Volume and Issue: 16(1)

Published: May 2, 2025

Language: Английский

---

Attention gate and dilation U-shaped network (GDUNet): an efficient breast ultrasound image segmentation network with multiscale information extraction

Quantitative Imaging in Medicine and Surgery, Journal Year: 2024, Volume and Issue: 14(2), P. 2034 - 2048

Published: Jan. 26, 2024

Background: In recent years, computer-aided diagnosis (CAD) systems have played an important role in breast cancer screening and diagnosis. The image segmentation task is the key step a CAD system for rapid identification of lesions. Therefore, efficient network necessary improving diagnostic accuracy screening. However, due to characteristics blurred boundaries, low contrast, speckle noise ultrasound images, lesion challenging. addition, many proposed tumor networks are too complex be applied practice. Methods: We developed attention gate dilation U-shaped (GDUNet), lightweight, model. This model improves inverted bottleneck, integrating it with tokenized multilayer perceptron (MLP) construct encoder. Additionally, we introduce lightweight (AG) within skip connection, which effectively filters low-level semantic information across spatial channel dimensions, thus attenuating irrelevant features. To further improve performance, innovated AG (AGDT) block embedded between encoder decoder order capture critical multiscale contextual information. Results: conducted experiments on two datasets. experiment’s results show that compared UNet, GDUNet could reduce number parameters by 10 times computational complexity 58 while providing double inference speed. Moreover, achieved better performance than did state-of-the-art medical architecture. Conclusions: Our method can achieve advanced different datasets high efficiency.

Language: Английский

---

A radiomic and clinical data-based risk model for malignancy prediction of breast BI-RADS 4A microcalcifications

Clinical Breast Cancer, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

Language: Английский

---

Integrating intratumoral and peritumoral radiomics with deep transfer learning for DCE-MRI breast lesion differentiation: A multicenter study comparing performance with radiologists

European Journal of Radiology, Journal Year: 2024, Volume and Issue: 177, P. 111556 - 111556

Published: June 12, 2024

Language: Английский

---

A radiomics model of contrast-enhanced computed tomography for predicting post-acute pancreatitis diabetes mellitus

Quantitative Imaging in Medicine and Surgery, Journal Year: 2024, Volume and Issue: 14(3), P. 2267 - 2279

Published: Feb. 6, 2024

Diabetes mellitus can occur after acute pancreatitis (AP), but the accurate quantitative methods to predict post-acute diabetes (PPDM-A) are lacking. This retrospective study aimed establish a radiomics model based on contrast-enhanced computed tomography (CECT) for predicting PPDM-A.

Language: Английский

---

Ultrasomics differentiation of malignant and benign focal liver lesions based on contrast-enhanced ultrasound

BMC Medical Imaging, Journal Year: 2024, Volume and Issue: 24(1)

Published: Sept. 16, 2024

To establish a nomogram for differentiating malignant and benign focal liver lesions (FLLs) using ultrasomics features derived from contrast-enhanced ultrasound (CEUS).

Language: Английский

---

Development of a Nomogram-Integrated Model Incorporating Intra-tumoral and Peri-tumoral Ultrasound Radiomics Alongside Clinical Parameters for the Prediction of Histological Grading in Invasive Breast Cancer

Ultrasound in Medicine & Biology, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 1, 2024

Language: Английский

---