Artificial intelligence: A review of current applications in hepatocellular carcinoma imaging

MED-ChatGPT CoPilot: a ChatGPT medical assistant for case mining and adjunctive therapy

Background

The large-scale language model, GPT-4-1106-preview, supports text of up to 128 k characters, which has enhanced the capability of processing vast quantities of text. This model can perform efficient and accurate text data mining without the need for retraining, aided by prompt engineering.

Method

The research approach includes prompt engineering and text vectorization processing. In this study, prompt engineering is applied to assist ChatGPT in text mining. Subsequently, the mined results are vectorized and incorporated into a local knowledge base. After cleansing 306 medical papers, data extraction was performed using ChatGPT. Following a validation and filtering process, 241 medical case data entries were obtained, leading to the construction of a local medical knowledge base. Additionally, drawing upon the Langchain framework and utilizing the local knowledge base in conjunction with ChatGPT, we successfully developed a fast and reliable chatbot. This chatbot is capable of providing recommended diagnostic and treatment information for various diseases.

Results

The performance of the designed ChatGPT model, which was enhanced by data from the local knowledge base, exceeded that of the original model by 7.90% on a set of medical questions.

Conclusion

ChatGPT, assisted by prompt engineering, demonstrates effective data mining capabilities for large-scale medical texts. In the future, we plan to incorporate a richer array of medical case data, expand the scale of the knowledge base, and enhance ChatGPT's performance in the medical field.

Imaging in France: 2024 Update

Radiology in France has made major advances in recent years through innovations in research and clinical practice. French institutions have developed innovative imaging techniques and artificial intelligence applications in the field of diagnostic imaging and interventional radiology. These include, but are not limited to, a more precise diagnosis of cancer and other diseases, research in dual-energy and photon-counting computed tomography, new applications of artificial intelligence, and advanced treatments in the field of interventional radiology. This article aims to explore the major research initiatives and technological advances that are shaping the landscape of radiology in France. By highlighting key contributions in diagnostic imaging, artificial intelligence, and interventional radiology, we provide a comprehensive overview of how these innovations are improving patient outcomes, enhancing diagnostic accuracy, and expanding the possibilities for minimally invasive therapies. As the field continues to evolve, France's position at the forefront of radiological research ensures that these innovations will play a central role in addressing current healthcare challenges and improving patient care on a global scale.

Artificial intelligence in hepatocellular carcinoma diagnosis: a comprehensive review of current literature

BACKGROUND AND AIM

Hepatocellular carcinoma (HCC) diagnosis mainly relies on its pathognomonic radiological profile, obviating the need for biopsy. The project of incorporating artificial intelligence (AI) techniques in HCC aims to improve the performance of image recognition. Herein, we thoroughly analyze and evaluate proposed AI models in the field of HCC diagnosis.

METHODS

A comprehensive review of the literature was performed utilizing MEDLINE/PubMed and Web of Science databases with the end of search date being the 30th of September 2023. The MESH terms "Artificial Intelligence," "Liver Cancer," "Hepatocellular Carcinoma," "Machine Learning," and "Deep Learning" were searched in the title and/or abstract. All references of the obtained articles were also evaluated for any additional information.

RESULTS

Our search resulted in 183 studies meeting our inclusion criteria. Across all diagnostic modalities, reported area under the curve (AUC) of most developed models surpassed 0.900. A B-mode US and a contrast-enhanced US model achieved AUCs of 0.947 and 0.957, respectively. Regarding the more challenging task of HCC diagnosis, a 2021 deep learning model, trained with CT scans, classified hepatic malignant lesions with an AUC of 0.986. Finally, a MRI machine learning model developed in 2021 displayed an AUC of 0.975 when differentiating small HCCs from benign lesions, while another MRI-based model achieved HCC diagnosis with an AUC of 0.970.

CONCLUSIONS

AI tools may lead to significant improvement in diagnostic management of HCC. Many models fared better or comparable to experienced radiologists while proving capable of elevating radiologists' accuracy, demonstrating promising results for AI implementation in HCC-related diagnostic tasks.

MRI for hepatocellular carcinoma and the role of abbreviated MRI for surveillance of hepatocellular carcinoma

INTRODUCTION

Hepatocellular carcinoma (HCC) constitutes the majority of liver cancers and significantly impacts global cancer mortality. While ultrasound (US) with or without alpha-fetoprotein is the mainstay for HCC surveillance, its limitations highlight the necessity for more effective surveillance tools. Therefore, this review explores evolving imaging modalities and abbreviated magnetic resonance imaging (MRI) (AMRI) protocols as promising alternatives, addressing challenges in HCC surveillance.

AREAS COVERED

This comprehensive review delves into the evaluation and challenges of HCC surveillance tools, focusing on non-contrast abbreviated MRI (NC-AMRI) and contrast-enhanced abbreviated MRI protocols. It covers the implementation of AMRI for HCC surveillance, patient preferences, adherence, and strategies for optimizing cost-effectiveness. Additionally, the article provides insights into prospects for HCC surveillance by summarizing meta-analyses, prospective studies, and ongoing clinical trials evaluating AMRI protocols.

EXPERT OPINION

The opinions underscore the transformative impact of AMRI on HCC surveillance, especially in overcoming US limitations. Promising results from NC-AMRI protocols indicate its potential for high-risk patient surveillance, though prospective studies in true surveillance settings are essential for validation. Future research should prioritize risk-stratified AMRI protocols and address cost-effectiveness for broader clinical implementation, alongside comparative analyses with US for optimal surveillance strategies.

Artificial intelligence in interventional radiology: Current concepts and future trends

While artificial intelligence (AI) is already well established in diagnostic radiology, it is beginning to make its mark in interventional radiology. AI has the potential to dramatically change the daily practice of interventional radiology at several levels. In the preoperative setting, recent advances in deep learning models, particularly foundation models, enable effective management of multimodality and increased autonomy through their ability to function minimally without supervision. Multimodality is at the heart of patient-tailored management and in interventional radiology, this translates into the development of innovative models for patient selection and outcome prediction. In the perioperative setting, AI is manifesting itself in applications that assist radiologists in image analysis and real-time decision making, thereby improving the efficiency, accuracy, and safety of interventions. In synergy with advances in robotic technologies, AI is laying the groundwork for an increased autonomy. From a research perspective, the development of artificial health data, such as AI-based data augmentation, offers an innovative solution to this central issue and promises to stimulate research in this area. This review aims to provide the medical community with the most important current and future applications of AI in interventional radiology.

An Introductory Guide to Artificial Intelligence in Interventional Radiology: Part 1 Foundational Knowledge

Artificial intelligence (AI) is rapidly evolving and has transformative potential for interventional radiology (IR) clinical practice. However, formal training in AI may be limited for many clinicians and therefore presents a challenge for initial implementation and trust in AI. An understanding of the foundational concepts in AI may help familiarize the interventional radiologist with the field of AI, thus facilitating understanding and participation in the development and deployment of AI. A pragmatic classification system of AI based on the complexity of the model may guide clinicians in the assessment of AI. Finally, the current state of AI in IR and the patterns of implementation are explored (pre-procedural, intra-procedural, and post-procedural).

Assessment of a multivariable model using MRI-radiomics, age and sex for the classification of hepatocellular adenoma subtypes

Objectives

Non-invasive subtyping of hepatocellular adenomas (HCA) remains challenging for several subtypes, thus carrying different levels of risks and management. The goal of this study is to devise a multivariable diagnostic model based on basic clinical features (age and sex) combined with MRI-radiomics and to evaluate its diagnostic performance.

Methods

This single-center retrospective case-control study included all consecutive patients with HCA identified within the pathological database from our institution from January 2003 to April 2018 with MRI examination (T2, T1-no injection/injection-arterial-portal); volumes of interest were manually delineated in adenomas and 38 textural features were extracted (LIFEx, v5.10). Qualitative (i.e., visual on MRI) and automatic (computer-assisted) analysis were compared. The prognostic scores of a multivariable diagnostic model based on basic clinical features (age and sex) combined with MRI-radiomics (tumor volume and texture features) were assessed using a cross-validated Random Forest algorithm.

Results

Via visual MR-analysis, HCA subgroups could be classified with balanced accuracies of 80.8 % (I-HCA or ß-I-HCA, the two being indistinguishable), 81.8 % (H-HCA) and 74.4 % (sh-HCA or ß-HCA also indistinguishable). Using a model including age, sex, volume and texture variables, HCA subgroups were predicted (multivariate classification) with an averaged balanced accuracy of 58.6 %, best=73.8 % (sh-HCA) and 71.9 % (ß-HCA). I-HCA and ß-I-HCA could be also distinguished (binary classification) with a balanced accuracy of 73 %.

Conclusion

Multiple HCA subtyping could be improved using machine-learning algorithms including two clinical features, i.e., age and sex, combined with MRI-radiomics. Future HCA studies enrolling more patients will further test the validity of the model.

Evaluation of a new beads reflux control microcatheter in drug-eluting bead transarterial chemoembolization

Rationale and objectives

A new microcatheter was recently developed claiming to reduce beads reflux in drug-eluting bead transarterial chemoembolization (DEB-TACE). The aim of this study was to compare the reflux control microcatheter ability versus a standard microcatheter for TACE treatment in patients with hepatocellular carcinoma.

Material and methods

Patients were prospectively included between November 2017 and February 2022. They received a DEB-TACE treatment with charged radiopaque beads using standard microcatheters or the SeQure reflux control microcatheter (Guerbet, France) and were assigned respectively to a control and a test group. Beads distribution mismatch was evaluated between the targeted territory on treatment planning CBCT and beads' spontaneous opacities on the post-intervention CBCT and the 1-month CT scanner.

Results

Twenty-three patients (21 men, median age 64 years [12.5 years]) with 37 hepatocellular carcinoma nodules were treated. The control group consisted of 13 patients - 19 nodules, while the test group included ten patients - 18 nodules. Non target embolization (NTE) was found in 20 % (2/10) of patients in the test group and 85 % (11/13) in the control group. NTE involved only an adjacent segment in the test group while it affected the adjacent biliary sector or even the contralateral liver lobe in the control group. No complication linked to NTE was found in the test group, while it led to one case of ischemic cholangitis and another case of biloma in the control group.

Conclusion

The reflux control microcatheter may be efficient in reducing NTE and thus eventual adverse events in comparison to standard of care end-hole microcatheters.

Application of artificial intelligence radiomics in the diagnosis, treatment, and prognosis of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer, with an increasing incidence and poor prognosis. In the past decade, artificial intelligence (AI) technology has undergone rapid development in the field of clinical medicine, bringing the advantages of efficient data processing and accurate model construction. Promisingly, AI-based radiomics has played an increasingly important role in the clinical decision-making of HCC patients, providing new technical guarantees for prediction, diagnosis, and prognostication. In this review, we evaluated the current landscape of AI radiomics in the management of HCC, including its diagnosis, individual treatment, and survival prognosis. Furthermore, we discussed remaining challenges and future perspectives regarding the application of AI radiomics in HCC.

Systemic therapies in hepatocellular carcinoma: A revolution?

The evolution in systemic therapies in hepatocellular carcinoma (HCC) signifies a strategy of high-cost, high-gain innovation that originated with sorafenib, despite its limited impact on tumor response. This strategic approach paved the way for the emergence of a second wave of the short-lived competitive advantage, exemplified by the incorporation of atezolizumab plus bevacizumab and tremelimumab plus durvalumab. In the context of safety concerns within the liver cancer domain, the IMBRAVE150 and HIMALAYA trials boldly incorporated bevacizumab and tremelimumab, respectively, demonstrating the continuation of the high-risk, high-reward innovation paradigm. This review delves into the strengths, weaknesses, opportunities, and threats analysis of systemic therapies in the field of HCC.

Addition of contrast in ultrasound screening for hepatocellular carcinoma

Objective

Screening ultrasound for hepatocellular carcinoma (HCC) identifies lesions which require further characterization by a contrast-enhanced exam to non-invasively diagnose HCC. While ultrasound is recommended in screening, some HCC can be occult on grayscale imaging. The purpose of this study was to determine if the addition of ultrasound contrast (sulfahexafluoride) to screening ultrasound for HCC can identify more HCC lesions than grayscale sonographic imaging alone.

Methods

All HCC screening ultrasounds that also had contrast were evaluated in this retrospective study. Patients with a focal lesion seen only after administration of contrast (OAC) were noted, as well as any follow-up imaging or pathology results. Additional variables collected included patient demographics, cirrhosis type, and laboratory values.

Results

230 unique patients were included, of which 160 had imaging or pathology follow-up. 18 of these patients had an OAC lesion, of which 17 had follow-up. Among these OACs, there was one LIRADS M lesion (1/18, 5.6 %) and one bland portal vein thrombus identified, which were both confirmed on follow-up imaging. All LIRADS 4 OAC lesions were downgraded. No additional HCC were identified on follow-up imaging or pathology of these patients.

Conclusion

Addition of contrast to screening ultrasound did identify additional lesions, portal vein thrombus, and high grade malignancy. However, as the incidence of OAC lesions was low (7.8 %, 18/230) and most of the lesions were not malignant, addition of post contrast sweeps through the liver is of low value in the low to medium at-risk cirrhotic population in identifying occult HCC.

CT and MRI of abdominal cancers: current trends and perspectives in the era of radiomics and artificial intelligence

Abdominal cancers continue to pose daily challenges to clinicians, radiologists and researchers. These challenges are faced at each stage of abdominal cancer management, including early detection, accurate characterization, precise assessment of tumor spread, preoperative planning when surgery is anticipated, prediction of tumor aggressiveness, response to therapy, and detection of recurrence. Technical advances in medical imaging, often in combination with imaging biomarkers, show great promise in addressing such challenges. Information extracted from imaging datasets owing to the application of radiomics can be used to further improve the diagnostic capabilities of imaging. However, the analysis of the huge amount of data provided by these advances is a difficult task in daily practice. Artificial intelligence has the potential to help radiologists in all these challenges. Notably, the applications of AI in the field of abdominal cancers are expanding and now include diverse approaches for cancer detection, diagnosis and classification, genomics and detection of genetic alterations, analysis of tumor microenvironment, identification of predictive biomarkers and follow-up. However, AI currently has some limitations that need further refinement for implementation in the clinical setting. This review article sums up recent advances in imaging of abdominal cancers in the field of image/data acquisition, tumor detection, tumor characterization, prognosis, and treatment response evaluation.

Accurate non-invasive grading of nonfunctional pancreatic neuroendocrine tumors with a CT derived radiomics signature

PURPOSE

The purpose of this study was to develop a radiomics-signature using computed tomography (CT) data for the preoperative prediction of grade of nonfunctional pancreatic neuroendocrine tumors (NF-PNETs).

MATERIALS AND METHODS

A retrospective study was performed on patients undergoing resection for NF-PNETs between 2010 and 2019. A total of 2436 radiomic features were extracted from arterial and venous phases of pancreas-protocol CT examinations. Radiomic features that were associated with final pathologic grade observed in the surgical specimens were subjected to joint mutual information maximization for hierarchical feature selection and the development of the radiomic-signature. Youden-index was used to identify optimal cutoff for determining tumor grade. A random forest prediction model was trained and validated internally. The performance of this tool in predicting tumor grade was compared to that of EUS-FNA sampling that was used as the standard of reference.

RESULTS

A total of 270 patients were included and a fusion radiomic-signature based on 10 selected features was developed using the development cohort (n = 201). There were 149 men and 121 women with a mean age of 59.4 ± 12.3 (standard deviation) years (range: 23.3-85.0 years). Upon internal validation in a new set of 69 patients, a strong discrimination was observed with an area under the curve (AUC) of 0.80 (95% confidence interval [CI]: 0.71-0.90) with corresponding sensitivity and specificity of 87.5% (95% CI: 79.7-95.3) and 73.3% (95% CI: 62.9-83.8) respectively. Of the study population, 143 patients (52.9%) underwent EUS-FNA. Biopsies were non-diagnostic in 26 patients (18.2%) and could not be graded due to insufficient sample in 42 patients (29.4%). In the cohort of 75 patients (52.4%) in whom biopsies were graded the radiomic-signature demonstrated not different AUC as compared to EUS-FNA (AUC: 0.69 vs. 0.67; P = 0.723), however greater sensitivity (i.e., ability to accurately identify G2/3 lesion was observed (80.8% vs. 42.3%; P < 0.001).

CONCLUSION

Non-invasive assessment of tumor grade in patients with PNETs using the proposed radiomic-signature demonstrated high accuracy. Prospective validation and optimization could overcome the commonly experienced diagnostic uncertainty in the assessment of tumor grade in patients with PNETs and could facilitate clinical decision-making.

SADSNet: A robust 3D synchronous segmentation network for liver and liver tumors based on spatial attention mechanism and deep supervision

Highlights

• Introduce a data augmentation strategy to expand the required different morphological data during the training and learning phase, and improve the algorithm's feature learning ability for complex and diverse tumor morphology CT images.• Design attention mechanisms for encoding and decoding paths to extract fine pixel level features, improve feature extraction capabilities, and achieve efficient spatial channel feature fusion.• The deep supervision layer is used to correct and decode the final image data to provide high accuracy of results.• The effectiveness of this method has been affirmed through validation on the LITS, 3DIRCADb, and SLIVER datasets.

BACKGROUND

Accurately extracting liver and liver tumors from medical images is an important step in lesion localization and diagnosis, surgical planning, and postoperative monitoring. However, the limited number of radiation therapists and a great number of images make this work time-consuming.

OBJECTIVE

This study designs a spatial attention deep supervised network (SADSNet) for simultaneous automatic segmentation of liver and tumors.

METHOD

Firstly, self-designed spatial attention modules are introduced at each layer of the encoder and decoder to extract image features at different scales and resolutions, helping the model better capture liver tumors and fine structures. The designed spatial attention module is implemented through two gate signals related to liver and tumors, as well as changing the size of convolutional kernels; Secondly, deep supervision is added behind the three layers of the decoder to assist the backbone network in feature learning and improve gradient propagation, enhancing robustness.

RESULTS

The method was testing on LITS, 3DIRCADb, and SLIVER datasets. For the liver, it obtained dice similarity coefficients of 97.03%, 96.11%, and 97.40%, surface dice of 81.98%, 82.53%, and 86.29%, 95% hausdorff distances of 8.96 mm, 8.26 mm, and 3.79 mm, and average surface distances of 1.54 mm, 1.19 mm, and 0.81 mm. Additionally, it also achieved precise tumor segmentation, which with dice scores of 87.81% and 87.50%, surface dice of 89.63% and 84.26%, 95% hausdorff distance of 12.96 mm and 16.55 mm, and average surface distances of 1.11 mm and 3.04 mm on LITS and 3DIRCADb, respectively.

CONCLUSION

The experimental results show that the proposed method is effective and superior to some other methods. Therefore, this method can provide technical support for liver and liver tumor segmentation in clinical practice.

Advances in Histological and Molecular Classification of Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is a primary liver cancer characterized by hepatocellular differentiation. HCC is molecularly heterogeneous with a wide spectrum of histopathology. The prognosis of patients with HCC is generally poor, especially in those with advanced stages. HCC remains a diagnostic challenge for pathologists because of its morphological and phenotypic diversity. However, recent advances have enhanced our understanding of the molecular genetics and histological subtypes of HCC. Accurate diagnosis of HCC is important for patient management and prognosis. This review provides an update on HCC pathology, focusing on molecular genetics, histological subtypes, and diagnostic approaches.

Pretreatment CT Texture Parameters as Predictive Biomarkers of Progression-Free Survival in Follicular Lymphoma Treated with Immunochemotherapy and Rituximab Maintenance

The purpose of this study was to determine whether texture analysis features present on pretreatment unenhanced computed tomography (CT) images, derived from 18F-fluorodeoxyglucose positron emission/computed tomography (18-FDG PET/CT), can predict progression-free survival (PFS), progression-free survival at 24 months (PFS 24), time to next treatment (TTNT), and overall survival in patients with high-tumor-burden follicular lymphoma treated with immunochemotherapy and rituximab maintenance. Seventy-two patients with follicular lymphoma were retrospectively included. Texture analysis was performed on unenhanced CT images extracted from 18-FDG PET/CT examinations that were obtained within one month before treatment. Skewness at a fine texture scale (SSF = 2) was an independent predictor of PFS (hazard ratio = 3.72 (95% CI: 1.15, 12.11), p = 0.028), PFS 24 (hazard ratio = 13.38; 95% CI: 1.29, 138.13; p = 0.029), and TTNT (hazard ratio = 5.11; 95% CI: 1.18, 22.13; p = 0.029). Skewness values above -0.015 at SSF = 2 were significantly associated with lower PFS, PFS 24, and TTNT. Kurtosis without filtration was an independent predictor of PFS (SSF = 0; HR = 1.22 (95% CI: 1.04, 1.44), p = 0.013), and TTNT (SSF = 0; hazard ratio = 1.23; 95% CI: 1.04, 1.46; p = 0.013). This study shows that pretreatment unenhanced CT texture analysis-derived tumor skewness and kurtosis may be used as predictive biomarkers of PFS and TTNT in patients with high-tumor-burden follicular lymphoma treated with immunochemotherapy and rituximab maintenance.

An Overview of the Use of Precision Population Medicine in Cancer Care: First of a Series

Advances in science and technology in the past century and a half have helped improve disease management, prevention, and early diagnosis and better health maintenance. These have led to a longer life expectancy in most developed and middle-income countries. However, resource- and infrastructure-scarce countries and populations have not enjoyed these benefits. Furthermore, in every society, including in developed nations, the lag time from new advances, either in the laboratory or from clinical trials, to using those findings in day-to-day medical practice often takes many years and sometimes close to or longer than a decade. A similar trend is seen in the application of "precision medicine" (PM) in terms of improving population health (PH). One of the reasons for such lack of application of precision medicine in population health is the misunderstanding of equating precision medicine with genomic medicine (GM) as if they are the same. Precision medicine needs to be recognized as encompassing genomic medicine in addition to other new developments such as big data analytics, electronic health records (EHR), telemedicine, and information communication technology. By leveraging these new developments together and applying well-tested epidemiological concepts, it can be posited that population/public health can be improved. In this paper, we take cancer as an example of the benefits of recognizing the potential of precision medicine in applying it to population/public health. Breast cancer and cervical cancer are taken as examples to demonstrate these hypotheses. There exists significant evidence already to show the importance of recognizing "precision population medicine" (PPM) in improving cancer outcomes not only in individual patients but also for its applications in early detection and cancer screening (especially in high-risk populations) and achieving those goals in a more cost-efficient manner that can reach resource- and infrastructure-scarce societies and populations. This is the first report of a series that will focus on individual cancer sites in the future.