ABSTRACT
Objective@#To measure inter-reader agreement and identify associated factors in interpreting complete response (CR) on magnetic resonance imaging (MRI) following chemoradiotherapy (CRT) for rectal cancer. @*Materials and Methods@#This retrospective study involved 10 readers from seven hospitals with experience of 80–10210 cases, and 149 patients who underwent surgery after CRT for rectal cancer. Using MRI-based tumor regression grading (mrTRG) and methods employed in daily practice, the readers independently assessed mrTRG, CR on T2-weighted images (T2WI) denoted as mrCR T2W, and CR on all images including diffusion-weighted images (DWI) denoted as mrCRoverall. The readers described their interpretation patterns and how they utilized DWI. Inter-reader agreement was measured using multi-rater kappa, and associated factors were analyzed using multivariable regression. Correlation between sensitivity and specificity of each reader was analyzed using Spearman coefficient. @*Results@#The mrCR T2W and mrCRoverall rates varied widely among the readers, ranging 18.8%–40.3% and 18.1%–34.9%, respectively. Nine readers used DWI as a supplement sequence, which modified interpretations on T2WI in 2.7% of cases (36/1341 [149 patients x 9 readers]) and mostly (33/36) changed mrCR T2W to non-mrCRoverall. The kappa values for mrTRG, mrCR T2W, and mrCRoverall were 0.56 (95% confidence interval: 0.49, 0.62), 0.55 (0.52, 0.57), and 0.54 (0.51, 0.57), respectively.No use of rectal gel, larger initial tumor size, and higher initial cT stage exhibited significant association with a higher interreader agreement for assessing mrCRoverall (P ≤ 0.042). Strong negative correlations were observed between the sensitivity and specificity of individual readers (coefficient, -0.718 to -0.963; P ≤ 0.019). @*Conclusion@#Inter-reader agreement was moderate for assessing CR on post-CRT MRI. Readers’ varying standards on MRI interpretation (i.e., threshold effect), along with the use of rectal gel, initial tumor size, and initial cT stage, were significant factors associated with inter-reader agreement.
ABSTRACT
Objective@#This study aims to explore the opinions on the insurance coverage of artificial intelligence (AI), as categorized based on the distinct value elements offered by AI, with a specific focus on patient-centered outcomes (PCOs). PCOs are distinguished from traditional clinical outcomes and focus on patient-reported experiences and values such as quality of life, functionality, well-being, physical or emotional status, and convenience. @*Materials and Methods@#We classified the value elements provided by AI into four dimensions: clinical outcomes, economic aspects, organizational aspects, and non-clinical PCOs. The survey comprised three sections: 1) experiences with PCOs in evaluating AI, 2) opinions on the coverage of AI by the National Health Insurance of the Republic of Korea when AI demonstrated benefits across the four value elements, and 3) respondent characteristics. The opinions regarding AI insurance coverage were assessed dichotomously and semi-quantitatively: non-approval (0) vs. approval (on a 1–10 weight scale, with 10 indicating the strongest approval). The survey was conducted from July 4 to 26, 2023, using a web-based method. Responses to PCOs and other value elements were compared. @*Results@#Among 200 respondents, 44 (22%) were patients/patient representatives, 64 (32%) were industry/developers, 60 (30%) were medical practitioners/doctors, and 32 (16%) were government health personnel. The level of experience with PCOs regarding AI was low, with only 7% (14/200) having direct experience and 10% (20/200) having any experience (either direct or indirect). The approval rate for insurance coverage for PCOs was 74% (148/200), significantly lower than the corresponding rates for other value elements (82.5%–93.5%; P ≤ 0.034). The approval strength was significantly lower for PCOs, with a mean weight ± standard deviation of 5.1 ± 3.5, compared to other value elements (P ≤ 0.036). @*Conclusion@#There is currently limited demand for insurance coverage for AI that demonstrates benefits in terms of nonclinical PCOs.
ABSTRACT
Plants belonging to the genus Juncus are widely distributed across North America, which are known to make a diverse array of bioactive natural products. In 2022, Juncus torreyi, a species of Juncus, was firstly found in Korea. Morphological and ecological characteristics of the species have been previously investigated;however, bioactive chemical potentials still remain to be explored. In the present work, we focused on the isolation and characterization of metabolites that harbor growth inhibitory activity against a fungal indicator, Candida albicans. Using activity-guided discovery method, we subsequently isolated and purified three metabolites from the most active methylene chloride-soluble fraction. Through NMR and high-resolution ESIOrbitrap-MS data analysis, the metabolites were structurally determined to be juncatrin B (1), ensifolin I (2), and juncusol (3). Metabolites 1–3 were evaluated for their C. albicans growth inhibitory activity and revealed inhibition with an IC 50 value of 74.3, 31.5, and 64.6 μg/mL, respectively.
ABSTRACT
Objective@#Chronic enteropathy associated with SLCO2A1 gene (CEAS) is a recently recognized disease. We aimed to evaluate the enterographic findings of CEAS. @*Materials and Methods@#Altogether, 14 patients with CEAS were confirmed based on known SLCO2A1 mutations. They were registered in a multicenter Korean registry between July 2018 and July 2021. Nine of the patients (37.2 ± 13 years; all female) who underwent surgery-naïve-state computed tomography enterography (CTE) or magnetic resonance enterography (MRE) were identified. Two experienced radiologists reviewed 25 and 2 sets of CTE and MRE examinations, respectively, regarding the small bowel findings. @*Results@#In initial evaluation, eight patients showed a total of 37 areas with mural abnormalities in the ileum on CTE, including 1–4 segments in six and > 10 segments in two patients. One patient showed unremarkable CTE. The involved segments were 10–85 mm (median, 20 mm) in length, 3–14 mm (median, 7 mm) in mural thickness, circumferential in 86.5% (32/37), and showed stratified enhancement in the enteric and portal phases in 91.9% (34/37) and 81.8% (9/11), respectively. Perienteric infiltration and prominent vasa recta were noted in 2.7% (1/37) and 13.5% (5/37), respectively. Bowel strictures were identified in six patients (66.7%), with a maximum upstream diameter of 31–48 mm. Two patients underwent surgery for strictures immediately after the initial enterography. Follow-up CTE and MRE in the remaining patients showed minimal-to-mild changes in the extent and thickness of the mural involvement for 17–138 months (median, 47.5 months) after initial enterography. Two patients required surgery for bowel stricture at 19 and 38 months of follow-up, respectively. @*Conclusion@#CEAS of the small bowel typically manifested on enterography in varying numbers and lengths of abnormal ileal segments that showed circumferential mural thickening with layered enhancement without perienteric abnormalities. The lesions caused bowel strictures that required surgery in some patients.
ABSTRACT
The editors of the Korean Journal of Radiology thank manuscript reviewers who completed their reviews for the journal from Nov. 2021 to Oct. 2022. We sincerely express our gratitude to all the reviewers listed below for their time and expertise.Please note that editors listed in the current masthead for the Korean Journal of Radiology were not included.
ABSTRACT
Glossopharyngeal neuralgia is a condition characterized by lancinating pain in the tongue, soft palate, and pharynx. This condition can be caused by the combination of traction and compression of the glossopharyngeal nerve by tortuous vertebral and posterior inferior cerebellar arteries, which tug down and exert pressure on the nerve. Medical treatments including carbamazepine and gabapentin have been found to effectively manage glossopharyngeal neuralgia, even in cases with overt compression and traction of the nerve.
ABSTRACT
Objective@#Computed tomography enterography (CTE) and magnetic resonance enterography (MRE) are considered substitutes for each other for evaluating Crohn’s disease (CD). However, the adequacy of mixing them for routine periodic follow-up for CD has not been established. This study aimed to compare MRE alone with the mixed use of CTE and MRE for the periodic follow-up of small bowel inflammation in patients with CD. @*Materials and Methods@#We retrospectively compared two non-randomized groups, each comprising 96 patients with CD. One group underwent CTE and MRE (MRE followed by CTE or vice versa) for the follow-up of CD (interval, 13–27 months [median, 22 months]), and the other group underwent MRE alone (interval, 15–26 months [median, 21 months]). However, these two groups were similar in clinical characteristics. Three independent readers from three different institutions determined whether inflammation had decreased, remained unchanged, or increased within the entire small bowel and the terminal ileum based on sequential enterography of the patients after appropriate blinding. We compared the two groups for inter-reader agreement and accuracy (terminal ileum only) using endoscopy as the reference standard for enterographic interpretation. @*Results@#The inter-reader agreement was greater in the MRE alone group for the entire small bowel (intraclass correlation coefficient [ICC]: 0.683 vs. 0.473; p = 0.005) and the terminal ileum (ICC: 0.656 vs. 0.490; p = 0.030). The interpretation accuracy was higher in the MRE alone group without statistical significance (70.9%–74.5% vs. 57.9%–64.9% in individual readers; adjusted odds ratio = 3.21; p = 0.077). @*Conclusion@#The mixed use of CTE and MRE was inferior to MRE alone in terms of inter-reader reliability and could probably be less accurate than MRE alone for routine monitoring of small bowel inflammation in patients with CD. Therefore, the consistent use of MRE is favored for this purpose.
ABSTRACT
The importance of ethics in research and the use of artificial intelligence (AI) is increasingly recognized not only in the field of healthcare but throughout society. This article intends to provide domestic readers with practical points regarding the ethical issues of using radiological images for AI research, focusing on data security and privacy protection and the right to data. Therefore, this article refers to related domestic laws and government policies. Data security and privacy protection is a key ethical principle for AI, in which proper de-identification of data is crucial. Sharing healthcare data to develop AI in a way that minimizes business interests is another ethical point to be highlighted. The need for data sharing makes the data security and privacy protection even more important as data sharing increases the risk of data breach.
ABSTRACT
We are pleased to announce the appointment of Dong Hyun Yang, MD, PhD as the new section editor of cardiovascular section for the Korean Journal of Radiology (KJR). Dr. Yang is an Associate Professor in the Department of Radiology at Asan Medical Center (AMC) and the University of Ulsan College of Medicine. He graduated from College of Medicine, Pusan National University in 2000. At AMC, he completed his residency in the Department of Radiology. He secured fellowships in Pediatric (2008) and Cardiothoracic (2011) Radiology during this period. He became a faculty member at AMC in March 2012, with his research mainly focusing on cardiovascular imaging. He has over 150 scientific publications in peer-reviewed academic journals and six patents in the United States. Dr. Yang has served on the editorial board of KJR since 2018. Furthermore, he was an active reviewer for other journals such as Heart and JACC cardiovascular imaging. His major clinical and research interests include quantitative imaging, four-dimensional flow MRI, three-dimensional printing, and artificial intelligence in cardiovascular imaging.
ABSTRACT
Artificial intelligence (AI) will likely affect various fields of medicine. This article aims to explain the fundamental principles of clinical validation, device approval, and insurance coverage decisions of AI algorithms for medical diagnosis and prediction.Discrimination accuracy of AI algorithms is often evaluated with the Dice similarity coefficient, sensitivity, specificity, and traditional or free-response receiver operating characteristic curves. Calibration accuracy should also be assessed, especially for algorithms that provide probabilities to users. As current AI algorithms have limited generalizability to realworld practice, clinical validation of AI should put it to proper external testing and assisting roles. External testing could adopt diagnostic case-control or diagnostic cohort designs. A diagnostic case-control study evaluates the technical validity/ accuracy of AI while the latter tests the clinical validity/accuracy of AI in samples representing target patients in realworld clinical scenarios. Ultimate clinical validation of AI requires evaluations of its impact on patient outcomes, referred to as clinical utility, and for which randomized clinical trials are ideal. Device approval of AI is typically granted with proof of technical validity/accuracy and thus does not intend to directly indicate if AI is beneficial for patient care or if it improves patient outcomes. Neither can it categorically address the issue of limited generalizability of AI. After achieving device approval, it is up to medical professionals to determine if the approved AI algorithms are beneficial for real-world patient care. Insurance coverage decisions generally require a demonstration of clinical utility that the use of AI has improved patient outcomes.
ABSTRACT
Interest in health insurance coverage for artificial intelligence (AI)–based medical technologies is growing. This article provides a review of the current developments in the sphere and provides future perspectives, focusing on AI application in radiology.Current Concepts: In December 2019, the Health Insurance Review and Assessment Service under the Korean Ministry of Health and Welfare released its first guidelines for determining the National Health Insurance coverage for AI–based medical technologies. Additionally, in 2020, the largest US health insurance provider, the Centers for Medicare and Medicaid Services, approved payment for AI technologies using two different systems. First, in September 2020, it granted New Technology Add-on Payments for AI algorithms that facilitate the diagnosis and treatment of large vessel occlusion strokes. Second, in December 2020, the Centers for Medicare and Medicaid Services finalized the provision of reimbursements for IDx-DR through a Current Procedural Terminology code. The AI system screens for more than mild diabetic retinopathy, which requires further evaluation by an ophthalmologist.Discussion and Conclusion: An in-depth look at the three events suggests the importance of demonstrating the added clinical value of AI technologies through improved patient outcomes in enabling insurance coverage. Therefore, it is critical to create clinically meaningful collaboration between healthcare professionals and AI by understanding and combining their unique strengths, thus actualizing new forms of patient care instead of having AI merely copy the professionals. Furthermore, if National Health Insurance coverage is granted for AI technologies in radiology, add-on payments would be the most appropriate method.
ABSTRACT
Objective@#Flattening in the anteroposterior direction (AP flattening) of the terminal ileum (TI) or sigmoid colon (SC) lying across the psoas muscle, on magnetic resonance enterography (MRE), might mimic bowel inflammation in the coronal view.This study investigated the prevalence of AP flattening and the factors associated with its development. @*Materials and Methods@#A total of 364 surgery-naïve patients with Crohn’s disease (CD) who had undergone MRE were retrospectively reviewed. AP flattening was defined as a luminal collapse in the anteroposterior direction, with a bowel width in the axial plane < 1/4 of the normal diameter without reduction of bowel width in coronal images. The prevalence of AP flattening of the TI and SC on MRE in patients with bowel segments lying across the psoas muscle was determined. We further compared the rate of AP flattening between MRE and computed tomography enterography (CTE) in a subcohort of patients with prior CTE. The factors associated with AP flattening were analyzed using multivariable logistic regression in a subcohort of patients with endoscopic findings of TI. @*Results@#Three hundred and twenty-two and 363 patients, respectively, had TI and SC lying across the psoas muscle. The prevalence of AP flattening on MRE was 7.5% (24/322) in TI and 5.2% (19/363) in SC. The prevalences were significantly higher on MRE than on CTE in both the TI (7.3% [12/164] vs. 0.6% [1/164]; p = 0.003) and SC (5.8% [11/190] vs. 1.6% [3/ 190]; p = 0.039). AP flattening of the TI was independently and strongly associated with the absence of CD inflammation on endoscopy, with an adjusted odds ratio of 0.066 (p = 0.003) for the presence versus the absence (reference) of inflammation. @*Conclusion@#AP flattening of the TI or SC lying across the psoas muscle was uncommon and predominantly observed on MRE of the bowel without CD inflammation.
ABSTRACT
The recent introduction of various high-dimensional modeling methods, such as radiomics and deep learning, has created a much greater diversity in modeling approaches for survival prediction (or, more generally, time-to-event prediction). The newness of the recent modeling approaches and unfamiliarity with the model outputs may confuse some researchers and practitioners about the evaluation of the performance of such models. Methodological literacy to critically appraise the performance evaluation of the models and, ideally, the ability to conduct such an evaluation would be needed for those who want to develop models or apply them in practice. This article intends to provide intuitive, conceptual, and practical explanations of the statistical methods for evaluating the performance of survival prediction models with minimal usage of mathematical descriptions. It covers from conventional to deep learning methods, and emphasis has been placed on recent modeling approaches. This review article includes straightforward explanations of C indices (Harrell’s C index, etc.), timedependent receiver operating characteristic curve analysis, calibration plot, other methods for evaluating the calibration performance, and Brier score.
ABSTRACT
Objective@#Adequate methods of combining T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) to assess complete response (CR) to chemoradiotherapy (CRT) for rectal cancer are obscure. We aimed to determine an algorithm for combining T2WI and DWI to optimally suggest CR on MRI using visual assessment. @*Materials and Methods@#We included 376 patients (male:female, 256:120; mean age ± standard deviation, 59.7 ± 11.1 years) who had undergone long-course CRT for rectal cancer and both pre- and post-CRT high-resolution rectal MRI during 2017– 2018. Two experienced radiologists independently evaluated whether a tumor signal was absent, representing CR, on both post-CRT T2WI and DWI, and whether the pre-treatment DWI showed homogeneous hyperintensity throughout the lesion. Algorithms for combining T2WI and DWI were as follows: ‘AND,’ if both showed CR; ‘OR,’ if any one showed CR; and ‘conditional OR,’ if T2WI showed CR or DWI showed CR after the pre-treatment DWI showed homogeneous hyperintensity. Their efficacies for diagnosing pathologic CR (pCR) were determined in comparison with T2WI alone. @*Results@#Sixty-nine patients (18.4%) had pCR. AND had a lower sensitivity without statistical significance (vs. 62.3% [43/69]; 59.4% [41/69], p = 0.500) and a significantly higher specificity (vs. 87.0% [267/307]; 90.2% [277/307], p = 0.002) than those of T2WI. Both OR and conditional OR combinations resulted in a large increase in sensitivity (vs. 62.3% [43/69]; 81.2% [56/69], p < 0.001; and 73.9% [51/69], p = 0.008, respectively) and a large decrease in specificity (vs. 87.0% [267/307]; 57.0% [175/307], p < 0.001; and 69.1% [212/307], p < 0.001, respectively) as compared with T2WI, ultimately creating additional false interpretations of CR more frequently than additional identification of patients with pCR. @*Conclusion@#AND combination of T2WI and DWI is an appropriate strategy for suggesting CR using visual assessment of MRI after CRT for rectal cancer.
ABSTRACT
Few radiation-induced bowel perforations have been reported to date. Furthermore, perforation after ileal restoration in asymptomatic patients is rare. We report the case of a 61-year-old man who was administered preoperative chemoradiotherapy for advanced rectal cancer. The patient underwent ultra-low anterior resection with ileal diversion, followed by ileal restoration. Perforation was detected 9 days after restoration, and he underwent a right hemicolectomy. The histologic evaluation indicated ileal perforation caused by acute radiation enteritis.
ABSTRACT
Objective@#Flattening in the anteroposterior direction (AP flattening) of the terminal ileum (TI) or sigmoid colon (SC) lying across the psoas muscle, on magnetic resonance enterography (MRE), might mimic bowel inflammation in the coronal view.This study investigated the prevalence of AP flattening and the factors associated with its development. @*Materials and Methods@#A total of 364 surgery-naïve patients with Crohn’s disease (CD) who had undergone MRE were retrospectively reviewed. AP flattening was defined as a luminal collapse in the anteroposterior direction, with a bowel width in the axial plane < 1/4 of the normal diameter without reduction of bowel width in coronal images. The prevalence of AP flattening of the TI and SC on MRE in patients with bowel segments lying across the psoas muscle was determined. We further compared the rate of AP flattening between MRE and computed tomography enterography (CTE) in a subcohort of patients with prior CTE. The factors associated with AP flattening were analyzed using multivariable logistic regression in a subcohort of patients with endoscopic findings of TI. @*Results@#Three hundred and twenty-two and 363 patients, respectively, had TI and SC lying across the psoas muscle. The prevalence of AP flattening on MRE was 7.5% (24/322) in TI and 5.2% (19/363) in SC. The prevalences were significantly higher on MRE than on CTE in both the TI (7.3% [12/164] vs. 0.6% [1/164]; p = 0.003) and SC (5.8% [11/190] vs. 1.6% [3/ 190]; p = 0.039). AP flattening of the TI was independently and strongly associated with the absence of CD inflammation on endoscopy, with an adjusted odds ratio of 0.066 (p = 0.003) for the presence versus the absence (reference) of inflammation. @*Conclusion@#AP flattening of the TI or SC lying across the psoas muscle was uncommon and predominantly observed on MRE of the bowel without CD inflammation.
ABSTRACT
The recent introduction of various high-dimensional modeling methods, such as radiomics and deep learning, has created a much greater diversity in modeling approaches for survival prediction (or, more generally, time-to-event prediction). The newness of the recent modeling approaches and unfamiliarity with the model outputs may confuse some researchers and practitioners about the evaluation of the performance of such models. Methodological literacy to critically appraise the performance evaluation of the models and, ideally, the ability to conduct such an evaluation would be needed for those who want to develop models or apply them in practice. This article intends to provide intuitive, conceptual, and practical explanations of the statistical methods for evaluating the performance of survival prediction models with minimal usage of mathematical descriptions. It covers from conventional to deep learning methods, and emphasis has been placed on recent modeling approaches. This review article includes straightforward explanations of C indices (Harrell’s C index, etc.), timedependent receiver operating characteristic curve analysis, calibration plot, other methods for evaluating the calibration performance, and Brier score.
ABSTRACT
Objective@#Adequate methods of combining T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) to assess complete response (CR) to chemoradiotherapy (CRT) for rectal cancer are obscure. We aimed to determine an algorithm for combining T2WI and DWI to optimally suggest CR on MRI using visual assessment. @*Materials and Methods@#We included 376 patients (male:female, 256:120; mean age ± standard deviation, 59.7 ± 11.1 years) who had undergone long-course CRT for rectal cancer and both pre- and post-CRT high-resolution rectal MRI during 2017– 2018. Two experienced radiologists independently evaluated whether a tumor signal was absent, representing CR, on both post-CRT T2WI and DWI, and whether the pre-treatment DWI showed homogeneous hyperintensity throughout the lesion. Algorithms for combining T2WI and DWI were as follows: ‘AND,’ if both showed CR; ‘OR,’ if any one showed CR; and ‘conditional OR,’ if T2WI showed CR or DWI showed CR after the pre-treatment DWI showed homogeneous hyperintensity. Their efficacies for diagnosing pathologic CR (pCR) were determined in comparison with T2WI alone. @*Results@#Sixty-nine patients (18.4%) had pCR. AND had a lower sensitivity without statistical significance (vs. 62.3% [43/69]; 59.4% [41/69], p = 0.500) and a significantly higher specificity (vs. 87.0% [267/307]; 90.2% [277/307], p = 0.002) than those of T2WI. Both OR and conditional OR combinations resulted in a large increase in sensitivity (vs. 62.3% [43/69]; 81.2% [56/69], p < 0.001; and 73.9% [51/69], p = 0.008, respectively) and a large decrease in specificity (vs. 87.0% [267/307]; 57.0% [175/307], p < 0.001; and 69.1% [212/307], p < 0.001, respectively) as compared with T2WI, ultimately creating additional false interpretations of CR more frequently than additional identification of patients with pCR. @*Conclusion@#AND combination of T2WI and DWI is an appropriate strategy for suggesting CR using visual assessment of MRI after CRT for rectal cancer.
ABSTRACT
Few radiation-induced bowel perforations have been reported to date. Furthermore, perforation after ileal restoration in asymptomatic patients is rare. We report the case of a 61-year-old man who was administered preoperative chemoradiotherapy for advanced rectal cancer. The patient underwent ultra-low anterior resection with ileal diversion, followed by ileal restoration. Perforation was detected 9 days after restoration, and he underwent a right hemicolectomy. The histologic evaluation indicated ileal perforation caused by acute radiation enteritis.
ABSTRACT
Artificial intelligence (AI) will likely affect various fields of medicine. This article aims to explain the fundamental principles of clinical validation, device approval, and insurance coverage decisions of AI algorithms for medical diagnosis and prediction. Discrimination accuracy of AI algorithms is often evaluated with the Dice similarity coefficient, sensitivity, specificity, and traditional or free-response receiver operating characteristic curves. Calibration accuracy should also be assessed, especially for algorithms that provide probabilities to users. As current AI algorithms have limited generalizability to real-world practice, clinical validation of AI should put it to proper external testing and assisting roles. External testing could adopt diagnostic case-control or diagnostic cohort designs. A diagnostic case-control study evaluates the technical validity/accuracy of AI while the latter tests the clinical validity/accuracy of AI in samples representing target patients in real-world clinical scenarios. Ultimate clinical validation of AI requires evaluations of its impact on patient outcomes, referred to as clinical utility, and for which randomized clinical trials are ideal. Device approval of AI is typically granted with proof of technical validity/accuracy and thus does not intend to directly indicate if AI is beneficial for patient care or if it improves patient outcomes. Neither can it categorically address the issue of limited generalizability of AI. After achieving device approval, it is up to medical professionals to determine if the approved AI algorithms are beneficial for real-world patient care. Insurance coverage decisions generally require a demonstration of clinical utility that the use of AI has improved patient outcomes.