The Application Status of Radiomics-Based Machine Learning in Intrahepatic Cholangiocarcinoma: Systematic Review and Meta-Analysis

Journal of Medical Internet Research, Год журнала: 2025, Номер 27, С. e69906 - e69906

Опубликована: Май 5, 2025

Background Over the past few years, radiomics for detection of intrahepatic cholangiocarcinoma (ICC) has been extensively studied. However, systematic evidence is lacking in use this domain, which hinders its further development. Objective To address gap, our study delved into status quo and application value ICC aimed to offer evidence-based support promote field. Methods PubMed, Web Science, Cochrane Library, Embase were comprehensively retrieved determine relevant original studies. The quality was appraised through Radiomics Quality Score. In addition, subgroup analyses undertaken according datasets (training validation sets), imaging sources, model types. Results Fifty-eight studies encompassing 12,903 patients eligible, with an average Score 9.21. Radiomics-based machine learning (ML) mainly used diagnose (n=30), microvascular invasion (n=8), gene mutations (n=5), perineural (PNI; n=2), lymph node (LN) positivity (n=2), tertiary lymphoid structures (TLSs; predict overall survival (n=6) recurrence (n=9). C-index, sensitivity (SEN), specificity (SPC) ML developed using clinical features (CFs) 0.762 (95% CI 0.728-0.796), 0.72 0.66-0.77), 0.66-0.78), respectively, dataset. contrast, SEN, SPC radiomics-based detecting 0.853 0.824-0.882), 0.80 0.73-0.85), 0.88 0.83-0.92), respectively. constructed both CFs diagnosing 0.912 0.889-0.935), 0.77 0.72-0.81), 0.90 0.86-0.92). deep learning–based that integrated yielded a notably higher C-index 0.924 (0.863-0.984) task ICC. Additional showed demonstrated promising accuracy predicting recurrence, as well invasion, mutations, PNI, LN positivity, TLSs. Conclusions demonstrates excellent diagnosis involving specific tasks, such PNI TLSs, are still scarce. limited research on hindered analysis development across various models. Furthermore, challenges data heterogeneity interpretability caused by segmentation parameter variations require optimization refinement. Future should delve enhance use. Its integration practice holds great promise improving decision-making, boosting diagnostic treatment accuracy, minimizing unnecessary tests, optimizing health care resource usage.

Язык: Английский

---

The predictive accuracy of machine learning for the risk of death in HIV patients: a systematic review and meta-analysis

BMC Infectious Diseases, Год журнала: 2024, Номер 24(1)

Опубликована: Май 6, 2024

Abstract Background Early prediction of mortality in individuals with HIV (PWH) has perpetually posed a formidable challenge. With the widespread integration machine learning into clinical practice, some researchers endeavor to formulate models predicting risk for PWH. Nevertheless, diverse timeframes among PWH and potential multitude modeling variables have cast doubt on efficacy current predictive model HIV-related deaths. To address this, we undertook systematic review meta-analysis, aiming comprehensively assess utilization early deaths furnish evidence-based support advancement artificial intelligence this domain. Methods We systematically combed through PubMed, Cochrane, Embase, Web Science databases November 25, 2023. evaluate bias original studies included, employed Predictive Model Bias Risk Assessment Tool (PROBAST). During conducted subgroup analysis based survival non-survival models. Additionally, utilized meta-regression explore influence death time value Results After our comprehensive review, analyzed total 24 pieces literature, encompassing data from 401,389 diagnosed HIV. Within dataset, 23 articles specifically delved during long-term follow-ups outside hospital settings. The applied these comprised (COX regression) other outcomes meta-analysis unveiled that within training set, c-index people using stands at 0.83 (95% CI: 0.75–0.91). In validation is slightly lower 0.81 0.78–0.85). Notably, demonstrated neither follow-up nor occurrence events significantly impacted performance Conclusions study suggests viable approach developing non-time-based predictions regarding limited inclusion necessitates additional multicenter thorough validation.

Язык: Английский

---

Radiomics and machine learning based on preoperative MRI for predicting extrahepatic metastasis in hepatocellular carcinoma patients treated with transarterial chemoembolization

European Journal of Radiology Open, Год журнала: 2024, Номер 12, С. 100551 - 100551

Опубликована: Фев. 6, 2024

Язык: Английский

---

Artificial intelligence in radiology

Journal of the Mexican Federation of Radiology and Imaging, Год журнала: 2024, Номер 3(2)

Опубликована: Июль 9, 2024

Artificial intelligence (AI) is revolutionizing clinical medicine, particularly radiology, by enhancing diagnostic accuracy and streamlining operational efficiency.Radiology benefits from AI's prowess in image pattern recognition, which not only augments radiologists' capabilities but also optimizes tasks such as scheduling radiation monitoring.AI's applications span interventional enabling the interpretation of complex imaging data through advanced technologies convolutional neural networks radiomics.These tools help detect subtle disease indicators often missed human eye.AI improves radiology department management automating routine prioritizing urgent cases to ensure timely medical interventions.Educational programs must evolve prepare next generation radiologists for a future where AI ubiquitous their professional landscape.However, integrating into brings challenges, including ethical legal concerns about patient privacy, security, potential bias algorithms.Ethical be addressed developing robust guidelines that keep pace with technological advancements.Addressing these issues requires rigorous validation across various settings demographics.Undoubtedly, will empower radiologists, enhance accuracy, contribute precision personalized medicine.

Язык: Английский

---

Enhanced ISUP grade prediction in prostate cancer using multi-center radiomics data

Abdominal Radiology, Год журнала: 2025, Номер unknown

Опубликована: Март 6, 2025

To explore the predictive value of radiomics features extracted from anatomical ROIs in differentiating International Society Urological Pathology (ISUP) grading prostate cancer patients. This study included 1,500 patients a multi-center study. The peripheral zone (PZ) and central gland (CG, transition + zone) were segmented using deep learning algorithms defined as regions interest (ROI) this A total 12,918 image-based T2-weighted imaging (T2WI), apparent diffusion coefficient (ADC), diffusion-weighted (DWI) images these two ROIs. Synthetic minority over-sampling technique (SMOTE) algorithm was used to address class imbalance problem. Feature selection performed Pearson correlation analysis random forest regression. prediction model built classification algorithm. Kruskal-Wallis H test, ANOVA, Chi-Square Test for statistical analysis. 20 ISUP grading-related selected, including 10 PZ ROI CG ROI. On test set, combined exhibited better performance, with an AUC 0.928 (95% CI: 0.872, 0.966), compared alone (AUC: 0.838; 95% 0.722, 0.920) 0.904; 0.851, 0.945). demonstrates that radiomic based on sub-region can contribute enhanced grade prediction. combination GG provide more comprehensive information improved accuracy. Further validation strategy future will enhance its prospects improving decision-making clinical settings.

Язык: Английский

---

A Machine Learning Approach for Breast Cancer Risk Prediction in Digital Mammography

Applied Sciences, Год журнала: 2024, Номер 14(22), С. 10315 - 10315

Опубликована: Ноя. 9, 2024

Breast cancer is among the most prevalent cancers in female population globally. Therefore, screening campaigns as well approaches to identify patients at risk are particularly important for early detection of suspect lesions. This study aims propose a workflow automatic classification based on one relevant factors breast cancer, which represented by density. The proposed methodology takes advantage features automatically extracted from mammographic images, digital mammography represents major tool women. Textural were parenchyma through radiomics approach, and they used train different machine learning algorithms neural network models classify density according standard Imaging Reporting Data System (BI-RADS) guidelines. Both binary multiclass tasks have been carried out compared terms performance metrics. Preliminary results show interesting accuracy (93.55% task 82.14% task), promising current literature. As relies straightforward computationally efficient algorithms, it could serve basis fast-track protocol mammograms reduce radiologists’ workload.

Язык: Английский

---

Artificial intelligence in predicting recurrence after first-line treatment of liver cancer: a systematic review and meta-analysis

BMC Medical Imaging, Год журнала: 2024, Номер 24(1)

Опубликована: Окт. 7, 2024

The aim of this study was to conduct a systematic review and meta-analysis comprehensively evaluate the performance methodological quality artificial intelligence (AI) in predicting recurrence after single first-line treatment for liver cancer. A rigorous evaluation conducted on AI studies related cancer, retrieved from PubMed, Embase, Web Science, Cochrane Library, CNKI databases. area under curve (AUC), sensitivity (SENC), specificity (SPEC) each were extracted meta-analysis. Six percutaneous ablation (PA) studies, 16 surgical resection (SR) 5 transarterial chemoembolization (TACE) included hepatocellular carcinoma (HCC) treatment, respectively. Four SR 2 PA intrahepatic cholangiocarcinoma (ICC) colorectal cancer metastasis (CRLM) treatment. pooled SENC, SEPC, AUC primary HCC via PA, SR, TACE 0.78, 0.90, 0.92; 0.81, 0.77, 0.86; 0.73, 0.79, values ICC treated with CRLM 0.85, 0.71, 0.86 0.69, 0.63,0.74, This demonstrates comprehensive application value satisfactory results, indicating clinical translation potential

Язык: Английский

---

State-of-the-art imaging of hepatocellular carcinoma

Journal of Gastrointestinal Surgery, Год журнала: 2024, Номер 28(10), С. 1717 - 1725

Опубликована: Авг. 6, 2024

Язык: Английский

---

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

BMC Medical Imaging, Год журнала: 2024, Номер 24(1)

Опубликована: Сен. 16, 2024

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

Язык: Английский

---

Computed tomography-based radiomic model for the prediction of neoadjuvant immunochemotherapy response in patients with advanced gastric cancer

World Journal of Gastrointestinal Oncology, Год журнала: 2024, Номер 16(10), С. 4115 - 4128

Опубликована: Сен. 25, 2024

BACKGROUND Neoadjuvant immunochemotherapy (nICT) has emerged as a popular treatment approach for advanced gastric cancer (AGC) in clinical practice worldwide. However, the response of AGC patients to nICT displays significant heterogeneity, and no existing radiomic model utilizes baseline computed tomography predict outcomes. AIM To establish nICT. METHODS Patients with who received (n = 60) were randomly assigned training cohort 42) or test 18). Various machine learning models developed using selected features risk factors An individual nomogram was established based on chosen signature signature. The performance all assessed through receiver operating characteristic curve analysis, decision analysis (DCA) Hosmer-Lemeshow goodness-of-fit test. RESULTS could accurately In cohort, area under 0.893, 95% confidence interval 0.803-0.991. DCA indicated that application yielded greater net benefit than alternative models. CONCLUSION A combining designed efficacy AGC. This tool can assist clinicians treatment-related decision-making.

Язык: Английский

---