Machine learning
Credit: Shutterstock

A problem that physicians face today is that the world of medicine, inspiring and innovative though it is, can be rather limiting. Physicians and medical researchers talk to other physicians and medical researchers but may have little appetite for wandering too far from the medical fraternity. Therefore, advancements in science and technology in nonmedical fields can be missed entirely. 

In recent years, the astounding advancements in the technology of artificial intelligence (AI) are slowly being incorporated into every facet of human life, including home, work, entertainment, and travel. An immediate application of AI in medicine is the processing of large amounts of data at lightning speed, opening the door to faster and more efficient medical research. 

To what extent will AI infiltrate the practice of medicine in the future? Will it eventually be incorporated into bedside assessment, serum analysis, and histopathological diagnosis? The possible applications of AI conjured up by the imagination are truly endless. Two researchers who investigated the use of AI in the evaluation of cholangiocarcinoma (CCA) have published their findings in Oncology. We delve more deeply into their work in this article. 

Continue Reading

CCA Medical Imaging 

CCA is a rare cancer of the biliary epithelium and peribiliary glands that carries a very poor prognosis. Currently, the 5-year survival rate after surgical intervention is usually less than 35%. Surgical intervention is the only curative treatment for CCA but is feasible only when lesions are well localized. However, in most cases, the disease is not amenable to resection; approximately 20% to 50% of patients who undergo surgical exploration are found to have unresectable CCA. 

Read more about CCA etiology

Physicians currently do have an arsenal of medical imaging methods that can guide diagnosis and management. For example, ultrasound can be used for the initial assessment of patients who present with symptoms of biliary obstruction or hepatomegaly. Ultrasound is excellent for revealing bile duct dilatation but less useful for determining the pathogenesis of biliary obstruction. 

Computed tomography (CT) is useful for characterizing liver masses and detecting lymph node involvement or distant metastases. Studies have found that CT can be used to differentiate CCA from other liver diseases, such as hepatocellular carcinoma and liver abscess. CT can also be used to calculate the liver volume and residual liver volume, which are useful parameters for surgeons planning a course of surgical intervention. 

Magnetic resonance imaging (MRI) provides excellent soft-tissue contrast and accurately visualizes the bile ducts and adjacent structures. Currently, MRI is considered to be superior to CT in the assessment of CCA. “MRI techniques excel at visualizing malignant stricture with or without an associated mass, permit accurate lesion characterization, improve sensitivity in the evaluation of tumor extent along the bile duct and liver invasion, and provide staging information regarding vascular, ductal and lymph node status,” the researchers wrote.

The Role of AI in Radiology

The ability of AI to process large amounts of data makes it incredibly promising in the field of radiology. This idea makes perfect sense, given that radiology is a discipline that builds on data. As a result, a new field—radiomics—has emerged. Radiopaedia defines radiomics as “a field of medical study that aims to extract a large number of quantitative features from medical images using data characterization algorithms. The data is assessed for improved decision support.” 

The beauty of AI is that it can mimic the human process of “learning”; it actually gets better at processing data when more data is fed into it. In CCA imaging, AI can be used to process and integrate images obtained with different imaging modalities (ultrasound, CT, and MRI) and to enrich information about a tumor, such as its shape, size, and volume. In a published study, researchers created an artificial neural network (ANN) to differentiate between various hepatic masses. The results showed an average appropriate use criteria (AUC) rating of 0.961 for ANN alone.  

Read more about CCA epidemiology

AI can also play a role in identifying risk factors for CCA, as well as in diagnosing and grading the disease. We have a wealth of data in terms of gene expression profiles, disease recurrence patterns, and overall survival rates. AI can help us to analyze this information and guide us in evaluating prognosis and predicting treatment response.

“A novel radiomics nomogram based on radiomics signatures and clinical characteristics can be used to preoperatively predict early recurrence of [intrahepatic cholangiocarcinoma] after partial hepatectomy, resulting in devising appropriate strategies, which benefits patients with a high risk of early recurrence who need additional chemotherapy,” the authors of this study wrote.

An Open Door of Medical Opportunities 

If the use of AI in the aspects of medicine described in this article continues to increase, the possibilities for enhancing patient care are endless. Al is currently playing “an important role in the medical field, can be used for drug discovery, remote patient monitoring, medical diagnostics and imaging, risk management, virtual assistance, and hospital management,” wrote the authors of the study. 

In the future, perhaps we will see more AI scientists at our medical conferences and more physicians at tech conferences. If we have the courage to combine our expertise across different disciplines, there is no telling what we can achieve together. 


Yang CM, Shu J. Cholangiocarcinoma evaluation via imaging and artificial intelligence. Oncology. 2021;99(2):72-83. doi:10.1159/00050744

Feger J, Idris M. Radiomics. Radiopaedia. Last revised August 2, 2021. Accessed September 30, 2021.