Andreas Mittermeier, Matthias Aßenmacher, Balthasar Schachtner, Sergio Grosu, Vladana Dakovic, Viktar Kandratovich, Bastian Sabel, Michael Ingrisch
{"title":"[Automatic ICD-10 coding : Natural language processing for German MRI reports].","authors":"Andreas Mittermeier, Matthias Aßenmacher, Balthasar Schachtner, Sergio Grosu, Vladana Dakovic, Viktar Kandratovich, Bastian Sabel, Michael Ingrisch","doi":"10.1007/s00117-024-01349-2","DOIUrl":"10.1007/s00117-024-01349-2","url":null,"abstract":"<p><strong>Background: </strong>The medical coding of radiology reports is essential for a good quality of care and correct billing, but at the same time a complex and error-prone task.</p><p><strong>Objective: </strong>To assess the performance of natural language processing (NLP) for ICD-10 coding of German radiology reports using fine tuning of suitable language models.</p><p><strong>Material and methods: </strong>This retrospective study included all magnetic resonance imaging (MRI) radiology reports acquired at our institution between 2010 and 2020. The codes on discharge ICD-10 were matched to the corresponding reports to construct a dataset for multiclass classification. Fine tuning of GermanBERT and flanT5 was carried out on the total dataset (ds<sub>total</sub>) containing 1035 different ICD-10 codes and 2 reduced subsets containing the 100 (ds<sub>100</sub>) and 50 (ds<sub>50</sub>) most frequent codes. The performance of the model was assessed using top‑k accuracy for k = 1, 3 and 5. In an ablation study both models were trained on the accompanying metadata and the radiology report alone.</p><p><strong>Results: </strong>The total dataset consisted of 100,672 radiology reports, the reduced subsets ds<sub>100</sub> of 68,103 and ds<sub>50</sub> of 52,293 reports. The performance of the model increased when several of the best predictions of the model were taken into consideration, when the number of target classes was reduced and the metadata were combined with the report. The flanT5 outperformed GermanBERT across all datasets and metrics and was is suited as a medical coding assistant, achieving a top 3 accuracy of nearly 70% in the real-world dataset ds<sub>total</sub>.</p><p><strong>Conclusion: </strong>Finely tuned language models can reliably predict ICD-10 codes of German magnetic resonance imaging (MRI) radiology reports across various settings. As a coding assistant flanT5 can guide medical coders to make informed decisions and potentially reduce the workload.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"793-800"},"PeriodicalIF":0.0,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141908595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R Morant, A Gräwingholt, J Subelack, D Kuklinski, J Vogel, M Blum, A Eichenberger, A Geissler
{"title":"[The possible benefit of artificial intelligence in an organized population-related screening program : Initial results and perspective].","authors":"R Morant, A Gräwingholt, J Subelack, D Kuklinski, J Vogel, M Blum, A Eichenberger, A Geissler","doi":"10.1007/s00117-024-01345-6","DOIUrl":"10.1007/s00117-024-01345-6","url":null,"abstract":"<p><strong>Background: </strong>Mammography screening programs (MSP) have shown that breast cancer can be detected at an earlier stage enabling less invasive treatment and leading to a better survival rate. The considerable numbers of interval breast cancer (IBC) and the additional examinations required, the majority of which turn out not to be cancer, are critically assessed.</p><p><strong>Objective: </strong>In recent years companies and universities have used machine learning (ML) to develop powerful algorithms that demonstrate astonishing abilities to read mammograms. Can such algorithms be used to improve the quality of MSP?</p><p><strong>Method: </strong>The original screening mammographies of 251 cases with IBC were retrospectively analyzed using the software ProFound AI® (iCAD) and the results were compared (case score, risk score) with a control group. The relevant current literature was also studied.</p><p><strong>Results: </strong>The distributions of the case scores and the risk scores were markedly shifted to higher risks compared to the control group, comparable to the results of other studies.</p><p><strong>Conclusion: </strong>Retrospective studies as well as our own data show that artificial intelligence (AI) could change our approach to MSP in the future in the direction of personalized screening and could enable a significant reduction in the workload of radiologists, fewer additional examinations and a reduced number of IBCs; however, the results of prospective studies are needed before implementation.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"773-778"},"PeriodicalIF":0.0,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11422457/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141629522","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Artificial intelligence in diagnostic radiology for dose management : Advances and perspectives using the example of computed tomography].","authors":"Laura Garajová, Stephan Garbe, Alois M Sprinkart","doi":"10.1007/s00117-024-01330-z","DOIUrl":"10.1007/s00117-024-01330-z","url":null,"abstract":"<p><strong>Clinical-methodological problem: </strong>Imaging procedures employing ionizing radiation require compliance with European directives and national regulations in order to protect patients. Each exposure must be indicated, individually adapted, and documented. Unacceptable dose exceedances must be detected and reported. These tasks are time-consuming and require meticulous diligence.</p><p><strong>Standard radiological methods: </strong>Computed tomography (CT) is the most important contributor to medical radiation exposure. Optimizing the patient's dose is therefore mandatory. Use of modern technology and reconstruction algorithms already reduces exposure. Checking the indication, planning, and performing the examination are further important process steps with regard to radiation protection. Patient exposure is usually monitored by dose management systems (DMS). In special cases, a risk assessment is required by calculating the organ doses.</p><p><strong>Methodological innovations: </strong>Artificial intelligence (AI)-assisted techniques are increasingly used in various steps of the process: they support examination planning, improve patient positioning, and enable automated scan length adjustments. They also provide real-time estimates of individual organ doses.</p><p><strong>Evaluation: </strong>The integration of AI into medical imaging is proving successful in terms of dose optimization in various areas of the radiological workflow, from reconstruction to examination planning and performing exams. However, the use of AI in conjunction with DMS has not yet been considered on a large scale.</p><p><strong>Practical recommendation: </strong>AI processes offer promising tools to support dose management. However, their implementation in the clinical setting requires further research, extensive validation, and continuous monitoring.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"787-792"},"PeriodicalIF":0.0,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141322147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Diagnostics and treatment of pulmonary artery embolisms].","authors":"Alexander Böhner, Anton Faron, Daniel Kütting","doi":"10.1007/s00117-024-01364-3","DOIUrl":"10.1007/s00117-024-01364-3","url":null,"abstract":"<p><p>In recent years the diagnostics of pulmonary artery embolisms (PE) has gained significance, with confirmation occurring in only about 15-25 % of suspected cases. Despite technological advances, radiological methods remain problematic due to radiation and contrast medium exposure. Clinical scores play a crucial role in the risk assessment of PE. High-risk situations call for specific measures, while negative D‑dimers can help avoid overdiagnosis. Computed tomographic pulmonary angiography (CTPA) remains the gold standard with high sensitivity and specificity. Treatment requires an interdisciplinary team (pulmonary embolism response team, PERT). Anticoagulation is an option for stable patients, while in unstable or unsuccessful courses, thrombolysis or interventional procedures can be considered. Side effects, especially the risk of bleeding, need to be considered for both forms of treatment.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"801-810"},"PeriodicalIF":0.0,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142121294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Accelerated musculoskeletal magnetic resonance imaging with deep learning-based image reconstruction at 0.55 T-3 T].","authors":"Jan Vosshenrich, Jan Fritz","doi":"10.1007/s00117-024-01325-w","DOIUrl":"10.1007/s00117-024-01325-w","url":null,"abstract":"<p><strong>Clinical/methodical issue: </strong>Magnetic resonance imaging (MRI) is a central component of musculoskeletal imaging. However, long image acquisition times can pose practical barriers in clinical practice.</p><p><strong>Standard radiological methods: </strong>MRI is the established modality of choice in the diagnostic workup of injuries and diseases of the musculoskeletal system due to its high spatial resolution, excellent signal-to-noise ratio (SNR), and unparalleled soft tissue contrast.</p><p><strong>Methodological innovations: </strong>Continuous advances in hardware and software technology over the last few decades have enabled four-fold acceleration of 2D turbo-spin-echo (TSE) without compromising image quality or diagnostic performance. The recent clinical introduction of deep learning (DL)-based image reconstruction algorithms helps to minimize further the interdependency between SNR, spatial resolution and image acquisition time and allows the use of higher acceleration factors.</p><p><strong>Performance: </strong>The combined use of advanced acceleration techniques and DL-based image reconstruction holds enormous potential to maximize efficiency, patient comfort, access, and value of musculoskeletal MRI while maintaining excellent diagnostic accuracy.</p><p><strong>Achievements: </strong>Accelerated MRI with DL-based image reconstruction has rapidly found its way into clinical practice and proven to be of added value. Furthermore, recent investigations suggest that the potential of this technology does not yet appear to be fully harvested.</p><p><strong>Practical recommendations: </strong>Deep learning-reconstructed fast musculoskeletal MRI examinations can be reliably used for diagnostic work-up and follow-up of musculoskeletal pathologies in clinical practice.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"758-765"},"PeriodicalIF":0.0,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11422270/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141307594","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Mitteilungen des Berufsverbandes der Deutschen Radiologie.","authors":"","doi":"10.1007/s00117-024-01377-y","DOIUrl":"10.1007/s00117-024-01377-y","url":null,"abstract":"","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":"64 10","pages":"811-819"},"PeriodicalIF":0.0,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142309321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Importance and implementation of radiological modalities in the diagnostics of soft tissue tumors].","authors":"Caroline Grund, Dorothea Kerner, Deborah Driulini, Günther Schneider","doi":"10.1007/s00117-024-01348-3","DOIUrl":"10.1007/s00117-024-01348-3","url":null,"abstract":"<p><p>Malignant soft tissue tumors, in particular, require a multimodal treatment concept involving interdisciplinary cooperation between radiologists, pathologists, surgeons and oncologists at special tumor centers. The foundations of the treatment decision are the imaging diagnostics and the diagnosis confirmation based on tissue samples. The (local) extent and growth behavior of a tumor are among the most important findings of imaging as they have a direct influence on the surgical procedure. The most important diagnostic procedure here is magnetic resonance imaging (MRI). The T1-weighted and fat-suppressed sequences after i.v. contrast administration are used to visualize the extent of the tumor. In synopsis with diffusion-weighted and T2-weighted sequences, a differentiation between vital tumor tissue and tumor necrosis is additionally possible. This also enables targeted sampling from vital tumor parts so that the patient can be assigned to the appropriate treatment concept as quickly as possible.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"728-738"},"PeriodicalIF":0.0,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141899141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Guideline for the management of symptomatic and asymptomatic internal carotid artery stenosis].","authors":"Michelle Natalie Haffner","doi":"10.1007/s00117-024-01309-w","DOIUrl":"10.1007/s00117-024-01309-w","url":null,"abstract":"<p><strong>Background: </strong>Stenoses of the internal carotid artery (ICA) with a ≥ 50% degree of stenosis are common in the German population with a prevalence of approx. 4.2%. Furthermore, about 15% of ischemic strokes are due to stenosis or occlusion of the ICA. There are currently three approaches to the treatment of ICA stenoses: conventional therapy, endovascular treatment and surgical treatment.</p><p><strong>Methods: </strong>Summary of the S3 guideline on the diagnosis, treatment and follow-up of extracranial carotid stenosis from 3 February 2020.</p><p><strong>Results: </strong>Color-coded duplex sonography, computed tomography (CT) angiography and magnetic resonance (MR) angiography and, in exceptional cases, diagnostic subtraction angiography can be used to diagnose ICA stenoses, including the stenosis grade. The classification is into symptomatic and asymptomatic ICA stenoses. ICA stenoses can be treated conservatively, endovascularly with stenting or surgically by carotid endarterectomy. Invasive treatment should be carried out under inpatient conditions, whereby the hospital stay should be kept as short as possible. Color-coded duplex sonography should be performed regularly before discharge and at regular intervals thereafter to detect recurrent stenosis at an early stage and, if necessary, to initiate further invasive therapy.</p><p><strong>Conclusions: </strong>When choosing the treatment modality, particular attention must be paid to whether the stenosis is symptomatic or asymptomatic and the degree of stenosis. The S3 guideline on the diagnosis, treatment and follow-up of extracranial carotid stenosis from 3 February 2020 provides guidance here.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"710-715"},"PeriodicalIF":0.0,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140924180","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Open surgical treatment for carotid stenoses].","authors":"Roushanak Shayesteh-Kheslat, Mario Lescan","doi":"10.1007/s00117-024-01307-y","DOIUrl":"10.1007/s00117-024-01307-y","url":null,"abstract":"<p><strong>Background: </strong>Globally, stroke is considered the second most common cause of death. According to the German Federal Statistical Office, 33.6% of mortality was due to cardiovascular diseases, making them the most prevalent cause of death.</p><p><strong>Problem: </strong>Specifically, cerebral infarctions were recorded as the cause in over 16,000 cases. These figures underscore the significant role that cerebrovascular diseases play in Germany's mortality statistics. Notably, about 80% of strokes are ischemic. Moreover, one-fifth of all strokes result from extracranial carotid stenosis. The increase in stroke risk with advancing age, especially among men, is particularly striking. This trend highlights the growing importance of efficient stroke prevention and treatment in an aging society.</p><p><strong>Conclusion: </strong>Open surgical therapy, as a crucial treatment method, is increasingly important, especially as an emergency intervention. Its role in modern medicine emphasizes the urgency of prioritizing this life-saving procedure in healthcare and making it widely available.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"699-704"},"PeriodicalIF":0.0,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140869022","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
L Schappe, C Klein, J Stögbauer, J Federspiel, P Lochner
{"title":"[Role of duplex ultrasonography of the ophthalmic artery in internal carotid artery stenosis].","authors":"L Schappe, C Klein, J Stögbauer, J Federspiel, P Lochner","doi":"10.1007/s00117-024-01331-y","DOIUrl":"10.1007/s00117-024-01331-y","url":null,"abstract":"<p><strong>Background: </strong>Duplex sonographic visualization of a retrogradely perfused ophthalmic artery (AO) as an expression of an existing collateral supply in high-grade stenosis of the internal carotid artery (ACI) is a widely used and validated tool. After revascularization there may be another reversal of flow. Recently, the question of whether knowledge of the flow direction of the AO before and after implantation of a stent can be used as an outcome predictor has been posed more frequently.</p><p><strong>Methods and results: </strong>In this article, the method of duplex sonographic assessment of the AO is explained more elaborately and we present a case of a patient with 75% restenosis of the left ACI with contralateral chronic occlusion of the right ACI. We focus on the special aspect that the ipsilateral AO was initially perfused retrogradely and that postinterventionally there was a flow reversal to a physiological anterograde flow. The case report is used to illustrate the value of duplex sonographic visualization of the AO.</p><p><strong>Conclusion: </strong>Our case report is able to illustrate two aspects: On the one hand, the AO can be perfused retrogradely in the specific case of chronic ACI occlusion of the opposite side, even when the ACI stenosis of the considered side is less than 80%. On the other hand, the AO can offer added diagnostic value as a follow-up parameter for re- or in-stent stenoses.</p>","PeriodicalId":74635,"journal":{"name":"Radiologie (Heidelberg, Germany)","volume":" ","pages":"687-693"},"PeriodicalIF":0.0,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141433495","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}