@misc{SchroederSemmelmannSiegmundetal., author = {Schroeder, Josef A. and Semmelmann, Matthias and Siegmund, Heiko and Grafe, Claudia and Evert, Matthias and Palm, Christoph}, title = {Improved interactive computer-assisted approach for evaluation of ultrastructural cilia abnormalities}, series = {Ultrastructural Pathology}, volume = {41}, journal = {Ultrastructural Pathology}, number = {1}, doi = {10.1080/01913123.2016.1270978}, pages = {112 -- 113}, subject = {Zilie}, language = {en} } @inproceedings{RueckertRiederFeussneretal., author = {R{\"u}ckert, Tobias and Rieder, Maximilian and Feussner, Hubertus and Wilhelm, Dirk and R{\"u}ckert, Daniel and Palm, Christoph}, title = {Smoke Classification in Laparoscopic Cholecystectomy Videos Incorporating Spatio-temporal Information}, series = {Bildverarbeitung f{\"u}r die Medizin 2024: Proceedings, German Workshop on Medical Image Computing, March 10-12, 2024, Erlangen}, booktitle = {Bildverarbeitung f{\"u}r die Medizin 2024: Proceedings, German Workshop on Medical Image Computing, March 10-12, 2024, Erlangen}, editor = {Maier, Andreas and Deserno, Thomas M. and Handels, Heinz and Maier-Hein, Klaus H. and Palm, Christoph and Tolxdorff, Thomas}, publisher = {Springeer}, address = {Wiesbaden}, doi = {10.1007/978-3-658-44037-4_78}, pages = {298 -- 303}, abstract = {Heavy smoke development represents an important challenge for operating physicians during laparoscopic procedures and can potentially affect the success of an intervention due to reduced visibility and orientation. Reliable and accurate recognition of smoke is therefore a prerequisite for the use of downstream systems such as automated smoke evacuation systems. Current approaches distinguish between non-smoked and smoked frames but often ignore the temporal context inherent in endoscopic video data. In this work, we therefore present a method that utilizes the pixel-wise displacement from randomly sampled images to the preceding frames determined using the optical flow algorithm by providing the transformed magnitude of the displacement as an additional input to the network. Further, we incorporate the temporal context at evaluation time by applying an exponential moving average on the estimated class probabilities of the model output to obtain more stable and robust results over time. We evaluate our method on two convolutional-based and one state-of-the-art transformer architecture and show improvements in the classification results over a baseline approach, regardless of the network used.}, language = {en} } @unpublished{MendelRueckertWilhelmetal., author = {Mendel, Robert and R{\"u}ckert, Tobias and Wilhelm, Dirk and R{\"u}ckert, Daniel and Palm, Christoph}, title = {Motion-Corrected Moving Average: Including Post-Hoc Temporal Information for Improved Video Segmentation}, doi = {10.48550/arXiv.2403.03120}, pages = {9}, abstract = {Real-time computational speed and a high degree of precision are requirements for computer-assisted interventions. Applying a segmentation network to a medical video processing task can introduce significant inter-frame prediction noise. Existing approaches can reduce inconsistencies by including temporal information but often impose requirements on the architecture or dataset. This paper proposes a method to include temporal information in any segmentation model and, thus, a technique to improve video segmentation performance without alterations during training or additional labeling. With Motion-Corrected Moving Average, we refine the exponential moving average between the current and previous predictions. Using optical flow to estimate the movement between consecutive frames, we can shift the prior term in the moving-average calculation to align with the geometry of the current frame. The optical flow calculation does not require the output of the model and can therefore be performed in parallel, leading to no significant runtime penalty for our approach. We evaluate our approach on two publicly available segmentation datasets and two proprietary endoscopic datasets and show improvements over a baseline approach.}, subject = {Deep Learning}, language = {en} } @misc{ScheppachMendelProbstetal., author = {Scheppach, Markus W. and Mendel, Robert and Probst, Andreas and Meinikheim, Michael and Palm, Christoph and Messmann, Helmut and Ebigbo, Alanna}, title = {Artificial Intelligence (AI) - assisted vessel and tissue recognition during third space endoscopy (Smart ESD)}, series = {Zeitschrift f{\"u}r Gastroenterologie}, volume = {60}, journal = {Zeitschrift f{\"u}r Gastroenterologie}, number = {08}, publisher = {Georg Thieme Verlag}, address = {Stuttgart}, doi = {10.1055/s-0042-1755110}, abstract = {Clinical setting Third space procedures such as endoscopic submucosal dissection (ESD) and peroral endoscopic myotomy (POEM) are complex minimally invasive techniques with an elevated risk for operator-dependent adverse events such as bleeding and perforation. This risk arises from accidental dissection into the muscle layer or through submucosal blood vessels as the submucosal cutting plane within the expanding resection site is not always apparent. Deep learning algorithms have shown considerable potential for the detection and characterization of gastrointestinal lesions. So-called AI - clinical decision support solutions (AI-CDSS) are commercially available for polyp detection during colonoscopy. Until now, these computer programs have concentrated on diagnostics whereas an AI-CDSS for interventional endoscopy has not yet been introduced. We aimed to develop an AI-CDSS („Smart ESD") for real-time intra-procedural detection and delineation of blood vessels, tissue structures and endoscopic instruments during third-space endoscopic procedures. Characteristics of Smart ESD An AI-CDSS was invented that delineates blood vessels, tissue structures and endoscopic instruments during third-space endoscopy in real-time. The output can be displayed by an overlay over the endoscopic image with different modes of visualization, such as a color-coded semitransparent area overlay, or border tracing (demonstration video). Hereby the optimal layer for dissection can be visualized, which is close above or directly at the muscle layer, depending on the applied technique (ESD or POEM). Furthermore, relevant blood vessels (thickness> 1mm) are delineated. Spatial proximity between the electrosurgical knife and a blood vessel triggers a warning signal. By this guidance system, inadvertent dissection through blood vessels could be averted. Technical specifications A DeepLabv3+ neural network architecture with KSAC and a 101-layer ResNeSt backbone was used for the development of Smart ESD. It was trained and validated with 2565 annotated still images from 27 full length third-space endoscopic videos. The annotation classes were blood vessel, submucosal layer, muscle layer, electrosurgical knife and endoscopic instrument shaft. A test on a separate data set yielded an intersection over union (IoU) of 68\%, a Dice Score of 80\% and a pixel accuracy of 87\%, demonstrating a high overlap between expert and AI segmentation. Further experiments on standardized video clips showed a mean vessel detection rate (VDR) of 85\% with values of 92\%, 70\% and 95\% for POEM, rectal ESD and esophageal ESD respectively. False positive measurements occurred 0.75 times per minute. 7 out of 9 vessels which caused intraprocedural bleeding were caught by the algorithm, as well as both vessels which required hemostasis via hemostatic forceps. Future perspectives Smart ESD performed well for vessel and tissue detection and delineation on still images, as well as on video clips. During a live demonstration in the endoscopy suite, clinical applicability of the innovation was examined. The lag time for processing of the live endoscopic image was too short to be visually detectable for the interventionist. Even though the algorithm could not be applied during actual dissection by the interventionist, Smart ESD appeared readily deployable during visual assessment by ESD experts. Therefore, we plan to conduct a clinical trial in order to obtain CE-certification of the algorithm. This new technology may improve procedural safety and speed, as well as training of modern minimally invasive endoscopic resection techniques.}, subject = {Bildgebendes Verfahren}, language = {en} } @misc{MeinikheimMendelProbstetal., author = {Meinikheim, Michael and Mendel, Robert and Probst, Andreas and Scheppach, Markus W. and Messmann, Helmut and Palm, Christoph and Ebigbo, Alanna}, title = {Barrett-Ampel}, series = {Zeitschrift f{\"u}r Gastroenterologie}, volume = {60}, journal = {Zeitschrift f{\"u}r Gastroenterologie}, number = {08}, publisher = {Georg Thieme Verlag}, address = {Stuttgart}, doi = {10.1055/s-0042-1755109}, abstract = {Hintergrund Adenokarzinome des {\"O}sophagus sind bis heute mit einer infausten Prognose vergesellschaftet (1). Obwohl Endoskopiker mit Barrett-{\"O}sophagus als Pr{\"a}kanzerose konfrontiert werden, ist vor allem f{\"u}r nicht-Experten die Differenzierung zwischen Barrett-{\"O}sophagus ohne Dysplasie und assoziierten Neoplasien mitunter schwierig. Existierende Biopsieprotokolle (z.B. Seattle Protokoll) sind oftmals unzuverl{\"a}ssig (2). Eine fr{\"u}hzeitige Diagnose des Adenokarzinoms ist allerdings von fundamentaler Bedeutung f{\"u}r die Prognose des Patienten. Forschungsansatz Auf der Grundlage dieser Problematik, entwickelten wir in Kooperation mit dem Forschungslabor „Regensburg Medical Image Computing (ReMIC)" der OTH Regensburg ein auf k{\"u}nstlicher Intelligenz (KI) basiertes Entscheidungsunterst{\"u}tzungssystem (CDSS). Das auf einer DeepLabv3+ neuronalen Netzwerkarchitektur basierende CDSS differenziert mittels Mustererkennung Barrett- {\"O}sophagus ohne Dysplasie von Barrett-{\"O}sophagus mit Dysplasie bzw. Neoplasie („Klassifizierung"). Hierbei werden gemittelte Ausgabewahrscheinlichkeiten mit einem vom Benutzer definierten Schwellenwert verglichen. F{\"u}r Vorhersagen, die den Schwellenwert {\"u}berschreiten, berechnen wir die Kontur der Region und die Fl{\"a}che. Sobald die vorhergesagte L{\"a}sion eine bestimmte Gr{\"o}ße in der Eingabe {\"u}berschreitet, heben wir sie und ihren Umriss hervor. So erm{\"o}glicht eine farbkodierte Visualisierung eine Abgrenzung zwischen Dysplasie bzw. Neoplasie und normalem Barrett-Epithel („Segmentierung"). In einer Studie an Bildern in „Weißlicht" (WL) und „Narrow Band Imaging" (NBI) demonstrierten wir eine Sensitivit{\"a}t von mehr als 90\% und eine Spezifit{\"a}t von mehr als 80\% (3). In einem n{\"a}chsten Schritt, differenzierte unser KI-Algorithmus Barrett- Metaplasien von assoziierten Neoplasien anhand von zuf{\"a}llig abgegriffenen Bildern in Echtzeit mit einer Accuracy von 89.9\% (4). Darauf folgend, entwickelten wir unser System dahingehend weiter, dass unser Algorithmus nun auch dazu in der Lage ist, Untersuchungsvideos in WL, NBI und „Texture and Color Enhancement Imaging" (TXI) in Echtzeit zu analysieren (5). Aktuell f{\"u}hren wir eine Studie in einem randomisiert-kontrollierten Ansatz an unver{\"a}nderten Untersuchungsvideos in WL, NBI und TXI durch. Ausblick Um Patienten mit aus Barrett-Metaplasien resultierenden Neoplasien fr{\"u}hestm{\"o}glich an „High-Volume"-Zentren {\"u}berweisen zu k{\"o}nnen, soll unser KI-Algorithmus zuk{\"u}nftig vor allem Endoskopiker ohne extensive Erfahrung bei der Beurteilung von Barrett- {\"O}sophagus in der Krebsfr{\"u}herkennung unterst{\"u}tzen.}, subject = {Speiser{\"o}hrenkrebs}, language = {de} } @misc{ScheppachMendelProbstetal., author = {Scheppach, Markus W. and Mendel, Robert and Probst, Andreas and Meinikheim, Michael and Palm, Christoph and Messmann, Helmut and Ebigbo, Alanna}, title = {Intraprozedurale Strukturerkennung bei Third-Space Endoskopie mithilfe eines Deep-Learning Algorithmus}, series = {Zeitschrift f{\"u}r Gastroenterologie}, volume = {60}, journal = {Zeitschrift f{\"u}r Gastroenterologie}, number = {04}, publisher = {Thieme}, address = {Stuttgart}, doi = {10.1055/s-0042-1745652}, pages = {e250-e251}, abstract = {Einleitung Third-Space Interventionen wie die endoskopische Submukosadissektion (ESD) und die perorale endoskopische Myotomie (POEM) sind technisch anspruchsvoll und mit einem erh{\"o}hten Risiko f{\"u}r intraprozedurale Komplikationen wie Blutung oder Perforation assoziiert. Moderne Computerprogramme zur Unterst{\"u}tzung bei diagnostischen Entscheidungen werden unter Einsatz von k{\"u}nstlicher Intelligenz (KI) in der Endoskopie bereits erfolgreich eingesetzt. Ziel der vorliegenden Arbeit war es, relevante anatomische Strukturen mithilfe eines Deep-Learning Algorithmus zu detektieren und segmentieren, um die Sicherheit und Anwendbarkeit von ESD und POEM zu erh{\"o}hen. Methoden Zw{\"o}lf Videoaufnahmen in voller L{\"a}nge von Third-Space Endoskopien wurden aus der Datenbank des Universit{\"a}tsklinikums Augsburg extrahiert. 1686 Einzelbilder wurden f{\"u}r die Kategorien Submukosa, Blutgef{\"a}ß, Dissektionsmesser und endoskopisches Instrument annotiert und segmentiert. Mit diesem Datensatz wurde ein DeepLabv3+neuronales Netzwerk auf der Basis eines ResNet mit 101 Schichten trainiert und intern anhand der Parameter Intersection over Union (IoU), Dice Score und Pixel Accuracy validiert. Die F{\"a}higkeit des Algorithmus zur Gef{\"a}ßdetektion wurde anhand von 24 Videoclips mit einer Spieldauer von 7 bis 46 Sekunden mit 33 vordefinierten Gef{\"a}ßen evaluiert. Anhand dieses Tests wurde auch die Gef{\"a}ßdetektionsrate eines Experten in der Third-Space Endoskopie ermittelt. Ergebnisse Der Algorithmus zeigte eine Gef{\"a}ßdetektionsrate von 93,94\% mit einer mittleren Rate an falsch positiven Signalen von 1,87 pro Minute. Die Gef{\"a}ßdetektionsrate des Experten lag bei 90,1\% ohne falsch positive Ergebnisse. In der internen Validierung an Einzelbildern wurde eine IoU von 63,47\%, ein mittlerer Dice Score von 76,18\% und eine Pixel Accuracy von 86,61\% ermittelt. Zusammenfassung Dies ist der erste KI-Algorithmus, der f{\"u}r den Einsatz in der therapeutischen Endoskopie entwickelt wurde. Pr{\"a}limin{\"a}re Ergebnisse deuten auf eine mit Experten vergleichbare Detektion von Gef{\"a}ßen w{\"a}hrend der Untersuchung hin. Weitere Untersuchungen sind n{\"o}tig, um die Leistung des Algorithmus im Vergleich zum Experten genauer zu eruieren sowie einen m{\"o}glichen klinischen Nutzen zu ermitteln.}, language = {de} } @article{SouzaJrPachecoPassosetal., author = {Souza Jr., Luis Antonio de and Pacheco, Andr{\´e} G.C. and Passos, Leandro A. and Santana, Marcos Cleison S. and Mendel, Robert and Ebigbo, Alanna and Probst, Andreas and Messmann, Helmut and Palm, Christoph and Papa, Jo{\~a}o Paulo}, title = {DeepCraftFuse: visual and deeply-learnable features work better together for esophageal cancer detection in patients with Barrett's esophagus}, series = {Neural Computing and Applications}, volume = {36}, journal = {Neural Computing and Applications}, publisher = {Springer}, address = {London}, doi = {10.1007/s00521-024-09615-z}, pages = {10445 -- 10459}, abstract = {Limitations in computer-assisted diagnosis include lack of labeled data and inability to model the relation between what experts see and what computers learn. Even though artificial intelligence and machine learning have demonstrated remarkable performances in medical image computing, their accountability and transparency level must be improved to transfer this success into clinical practice. The reliability of machine learning decisions must be explained and interpreted, especially for supporting the medical diagnosis. While deep learning techniques are broad so that unseen information might help learn patterns of interest, human insights to describe objects of interest help in decision-making. This paper proposes a novel approach, DeepCraftFuse, to address the challenge of combining information provided by deep networks with visual-based features to significantly enhance the correct identification of cancerous tissues in patients affected with Barrett's esophagus (BE). We demonstrate that DeepCraftFuse outperforms state-of-the-art techniques on private and public datasets, reaching results of around 95\% when distinguishing patients affected by BE that is either positive or negative to esophageal cancer.}, subject = {Deep Learning}, language = {en} } @article{SouzaPachecodeSouzaetal., author = {Souza, Luis A. and Pacheco, Andr{\´e} G.C. and de Souza, Alberto F. and Oliveira-Santos, Thiago and Badue, Claudine and Palm, Christoph and Papa, Jo{\~a}o Paulo}, title = {TransConv: a lightweight architecture based on transformers and convolutional neural networks for adenocarcinoma and Barrett's esophagus identification}, series = {Neural Computing and Applications}, journal = {Neural Computing and Applications}, number = {37}, publisher = {Springer}, doi = {10.1007/s00521-025-11299-y}, pages = {15535 -- 15546}, abstract = {Barrett's esophagus, also known as BE, is commonly associated with repeated exposure to stomach acid. If not treated properly, it may evolve into esophageal adenocarcinoma, aka esophageal cancer. This paper proposes TransConv, a hybrid architecture that benefits from features learned by pre-trained vision transformers (ViTs) and convolutional neural networks (CNNs), followed by a shallow neural network composed of three normalizations, ReLU activations, and fully connected layers, and a SoftMax head to distinguish between BE and esophageal cancer. TransConv is designed to be training-lightweight, and for the ViT and CNN backbone models, weights are kept frozen during training, i.e., the primary goal of TransConv is to learn the weights of the fully connected layer from both backbones only, avoiding the burden of updating their weights but still learning their final descriptions for the lightweight convolutional model. We report promising results with low computational training costs in two datasets, one public and another private. From our achievements, TransConv was able to deliver balanced accuracy results around 85\% and 86\% for each evaluated dataset, respectively, in a design that required only 50 epochs of model training, a very reduced number compared to state-of-the-art conducted studies in the same domain.}, language = {en} } @unpublished{RueckertRueckertPalm, author = {R{\"u}ckert, Tobias and R{\"u}ckert, Daniel and Palm, Christoph}, title = {Methods and datasets for segmentation of minimally invasive surgical instruments in endoscopic images and videos: A review of the state of the art}, doi = {10.48550/arXiv.2304.13014}, pages = {25}, abstract = {In the field of computer- and robot-assisted minimally invasive surgery, enormous progress has been made in recent years based on the recognition of surgical instruments in endoscopic images. Especially the determination of the position and type of the instruments is of great interest here. Current work involves both spatial and temporal information with the idea, that the prediction of movement of surgical tools over time may improve the quality of final segmentations. The provision of publicly available datasets has recently encouraged the development of new methods, mainly based on deep learning. In this review, we identify datasets used for method development and evaluation, as well as quantify their frequency of use in the literature. We further present an overview of the current state of research regarding the segmentation and tracking of minimally invasive surgical instruments in endoscopic images. The paper focuses on methods that work purely visually without attached markers of any kind on the instruments, taking into account both single-frame segmentation approaches as well as those involving temporal information. A discussion of the reviewed literature is provided, highlighting existing shortcomings and emphasizing available potential for future developments. The publications considered were identified through the platforms Google Scholar, Web of Science, and PubMed. The search terms used were "instrument segmentation", "instrument tracking", "surgical tool segmentation", and "surgical tool tracking" and result in 408 articles published between 2015 and 2022 from which 109 were included using systematic selection criteria.}, language = {en} } @inproceedings{GutbrodRauberWeberNunesetal., author = {Gutbrod, Max and Rauber, David and Weber Nunes, Danilo and Palm, Christoph}, title = {OpenMIBOOD: Open Medical Imaging Benchmarks for Out-Of-Distribution Detection}, series = {2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 10.-17. June 2025, Nashville}, booktitle = {2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 10.-17. June 2025, Nashville}, publisher = {IEEE}, isbn = {979-8-3315-4364-8}, doi = {10.1109/CVPR52734.2025.02410}, pages = {25874 -- 25886}, abstract = {The growing reliance on Artificial Intelligence (AI) in critical domains such as healthcare demands robust mechanisms to ensure the trustworthiness of these systems, especially when faced with unexpected or anomalous inputs. This paper introduces the Open Medical Imaging Benchmarks for Out-Of-Distribution Detection (OpenMIBOOD), a comprehensive framework for evaluating out-of-distribution (OOD) detection methods specifically in medical imaging contexts. OpenMIBOOD includes three benchmarks from diverse medical domains, encompassing 14 datasets divided into covariate-shifted in-distribution, nearOOD, and far-OOD categories. We evaluate 24 post-hoc methods across these benchmarks, providing a standardized reference to advance the development and fair comparison of OODdetection methods. Results reveal that findings from broad-scale OOD benchmarks in natural image domains do not translate to medical applications, underscoring the critical need for such benchmarks in the medical field. By mitigating the risk of exposing AI models to inputs outside their training distribution, OpenMIBOOD aims to support the advancement of reliable and trustworthy AI systems in healthcare. The repository is available at https://github.com/remic-othr/OpenMIBOOD.}, language = {en} } @unpublished{RueckertRauberMaerkletal., author = {R{\"u}ckert, Tobias and Rauber, David and Maerkl, Raphaela and Klausmann, Leonard and Yildiran, Suemeyye R. and Gutbrod, Max and Nunes, Danilo Weber and Moreno, Alvaro Fernandez and Luengo, Imanol and Stoyanov, Danail and Toussaint, Nicolas and Cho, Enki and Kim, Hyeon Bae and Choo, Oh Sung and Kim, Ka Young and Kim, Seong Tae and Arantes, Gon{\c{c}}alo and Song, Kehan and Zhu, Jianjun and Xiong, Junchen and Lin, Tingyi and Kikuchi, Shunsuke and Matsuzaki, Hiroki and Kouno, Atsushi and Manesco, Jo{\~a}o Renato Ribeiro and Papa, Jo{\~a}o Paulo and Choi, Tae-Min and Jeong, Tae Kyeong and Park, Juyoun and Alabi, Oluwatosin and Wei, Meng and Vercauteren, Tom and Wu, Runzhi and Xu, Mengya and an Wang, and Bai, Long and Ren, Hongliang and Yamlahi, Amine and Hennighausen, Jakob and Maier-Hein, Lena and Kondo, Satoshi and Kasai, Satoshi and Hirasawa, Kousuke and Yang, Shu and Wang, Yihui and Chen, Hao and Rodr{\´i}guez, Santiago and Aparicio, Nicol{\´a}s and Manrique, Leonardo and Lyons, Juan Camilo and Hosie, Olivia and Ayobi, Nicol{\´a}s and Arbel{\´a}ez, Pablo and Li, Yiping and Khalil, Yasmina Al and Nasirihaghighi, Sahar and Speidel, Stefanie and R{\"u}ckert, Daniel and Feussner, Hubertus and Wilhelm, Dirk and Palm, Christoph}, title = {Comparative validation of surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation in endoscopy: Results of the PhaKIR 2024 challenge}, pages = {36}, abstract = {Reliable recognition and localization of surgical instruments in endoscopic video recordings are foundational for a wide range of applications in computer- and robot-assisted minimally invasive surgery (RAMIS), including surgical training, skill assessment, and autonomous assistance. However, robust performance under real-world conditions remains a significant challenge. Incorporating surgical context - such as the current procedural phase - has emerged as a promising strategy to improve robustness and interpretability. To address these challenges, we organized the Surgical Procedure Phase, Keypoint, and Instrument Recognition (PhaKIR) sub-challenge as part of the Endoscopic Vision (EndoVis) challenge at MICCAI 2024. We introduced a novel, multi-center dataset comprising thirteen full-length laparoscopic cholecystectomy videos collected from three distinct medical institutions, with unified annotations for three interrelated tasks: surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation. Unlike existing datasets, ours enables joint investigation of instrument localization and procedural context within the same data while supporting the integration of temporal information across entire procedures. We report results and findings in accordance with the BIAS guidelines for biomedical image analysis challenges. The PhaKIR sub-challenge advances the field by providing a unique benchmark for developing temporally aware, context-driven methods in RAMIS and offers a high-quality resource to support future research in surgical scene understanding.}, language = {en} } @misc{RueckertRueckertPalm, author = {R{\"u}ckert, Tobias and R{\"u}ckert, Daniel and Palm, Christoph}, title = {Corrigendum to "Methods and datasets for segmentation of minimally invasive surgical instruments in endoscopic images and videos: A review of the state of the art" [Comput. Biol. Med. 169 (2024) 107929]}, series = {Computers in Biology and Medicine}, journal = {Computers in Biology and Medicine}, publisher = {Elsevier}, doi = {10.1016/j.compbiomed.2024.108027}, url = {http://nbn-resolving.de/urn:nbn:de:bvb:898-opus4-70337}, pages = {1}, abstract = {The authors regret that the SAR-RARP50 dataset is missing from the description of publicly available datasets presented in Chapter 4.}, language = {en} } @article{HammerNunesHammeretal., author = {Hammer, Simone and Nunes, Danilo Weber and Hammer, Michael and Zeman, Florian and Akers, Michael and G{\"o}tz, Andrea and Balla, Annika and Doppler, Michael Christian and Fellner, Claudia and Da Platz Batista Silva, Natascha and Thurn, Sylvia and Verloh, Niklas and Stroszczynski, Christian and Wohlgemuth, Walter Alexander and Palm, Christoph and Uller, Wibke}, title = {Deep learning-based differentiation of peripheral high-flow and low-flow vascular malformations in T2-weighted short tau inversion recovery MRI}, series = {Clinical hemorheology and microcirculation}, journal = {Clinical hemorheology and microcirculation}, edition = {Pre-press}, publisher = {IOP Press}, doi = {10.3233/CH-232071}, pages = {1 -- 15}, abstract = {BACKGROUND Differentiation of high-flow from low-flow vascular malformations (VMs) is crucial for therapeutic management of this orphan disease. OBJECTIVE A convolutional neural network (CNN) was evaluated for differentiation of peripheral vascular malformations (VMs) on T2-weighted short tau inversion recovery (STIR) MRI. METHODS 527 MRIs (386 low-flow and 141 high-flow VMs) were randomly divided into training, validation and test set for this single-center study. 1) Results of the CNN's diagnostic performance were compared with that of two expert and four junior radiologists. 2) The influence of CNN's prediction on the radiologists' performance and diagnostic certainty was evaluated. 3) Junior radiologists' performance after self-training was compared with that of the CNN. RESULTS Compared with the expert radiologists the CNN achieved similar accuracy (92\% vs. 97\%, p = 0.11), sensitivity (80\% vs. 93\%, p = 0.16) and specificity (97\% vs. 100\%, p = 0.50). In comparison to the junior radiologists, the CNN had a higher specificity and accuracy (97\% vs. 80\%, p <  0.001; 92\% vs. 77\%, p <  0.001). CNN assistance had no significant influence on their diagnostic performance and certainty. After self-training, the junior radiologists' specificity and accuracy improved and were comparable to that of the CNN. CONCLUSIONS Diagnostic performance of the CNN for differentiating high-flow from low-flow VM was comparable to that of expert radiologists. CNN did not significantly improve the simulated daily practice of junior radiologists, self-training was more effective.}, language = {en} } @misc{ScheppachMendelProbstetal., author = {Scheppach, Markus W. and Mendel, Robert and Probst, Andreas and Nagl, Sandra and Meinikheim, Michael and Yip, Hon Chi and Lau, Louis Ho Shing and Chiu, Philip Wai Yan and Palm, Christoph and Messmann, Helmut and Ebigbo, Alanna}, title = {Effekt eines K{\"u}nstliche Intelligenz (KI) - Algorithmus auf die Gef{\"a}ßdetektion bei third space Endoskopien}, series = {Zeitschrift f{\"u}r Gastroenterologie}, volume = {61}, journal = {Zeitschrift f{\"u}r Gastroenterologie}, number = {08}, publisher = {Thieme}, address = {Stuttgart}, doi = {10.1055/s-0043-1771980}, pages = {e528-e529}, abstract = {Einleitung Third space Endoskopieprozeduren wie die endoskopische Submukosadissektion (ESD) und die perorale endoskopische Myotomie (POEM) sind technisch anspruchsvoll und gehen mit untersucherabh{\"a}ngigen Komplikationen wie Blutungen und Perforationen einher. Grund hierf{\"u}r ist die unabsichtliche Durchschneidung von submukosalen Blutgef{\"a}ßen ohne pr{\"a}emptive Koagulation. Ziele Die Forschungsfrage, ob ein KI-Algorithmus die intraprozedurale Gef{\"a}ßerkennung bei ESD und POEM unterst{\"u}tzen und damit Komplikationen wie Blutungen verhindern k{\"o}nnte, erscheint in Anbetracht des erfolgreichen Einsatzes von KI bei der Erkennung von Kolonpolypen interessant. Methoden Auf 5470 Einzelbildern von 59 third space Endoscopievideos wurden submukosale Blutgef{\"a}ße annotiert. Zusammen mit weiteren 179.681 nicht-annotierten Bildern wurde ein DeepLabv3+neuronales Netzwerk mit dem ECMT-Verfahren f{\"u}r semi-supervised learning trainiert, um Blutgef{\"a}ße in Echtzeit erkennen zu k{\"o}nnen. F{\"u}r die Evaluation wurde ein Videotest mit 101 Videoclips aus 15 vom Trainingsdatensatz separaten Prozeduren mit 200 vordefinierten Gef{\"a}ßen erstellt. Die Gef{\"a}ßdetektionsrate, -zeit und -dauer, definiert als der Prozentsatz an Einzelbildern eines Videos bezogen auf den Goldstandard, auf denen ein definiertes Gef{\"a}ß erkannt wurde, wurden erhoben. Acht erfahrene Endoskopiker wurden mithilfe dieses Videotests im Hinblick auf Gef{\"a}ßdetektion getestet, wobei eine H{\"a}lfte der Videos nativ, die andere H{\"a}lfte nach Markierung durch den KI-Algorithmus angesehen wurde. Ergebnisse Der mittlere Dice Score des Algorithmus f{\"u}r Blutgef{\"a}ße war 68\%. Die mittlere Gef{\"a}ßdetektionsrate im Videotest lag bei 94\% (96\% f{\"u}r ESD; 74\% f{\"u}r POEM). Die mediane Gef{\"a}ßdetektionszeit des Algorithmus lag bei 0,32 Sekunden (0,3 Sekunden f{\"u}r ESD; 0,62 Sekunden f{\"u}r POEM). Die mittlere Gef{\"a}ßdetektionsdauer lag bei 59,1\% (60,6\% f{\"u}r ESD; 44,8\% f{\"u}r POEM) des Goldstandards. Alle Endoskopiker hatten mit KI-Unterst{\"u}tzung eine h{\"o}here Gef{\"a}ßdetektionsrate als ohne KI. Die mittlere Gef{\"a}ßdetektionsrate ohne KI lag bei 56,4\%, mit KI bei 71,2\% (p<0.001). Schlussfolgerung KI-Unterst{\"u}tzung war mit einer statistisch signifikant h{\"o}heren Gef{\"a}ßdetektionsrate vergesellschaftet. Die mediane Gef{\"a}ßdetektionszeit von deutlich unter einer Sekunde sowie eine Gef{\"a}ßdetektionsdauer von gr{\"o}ßer 50\% des Goldstandards wurden f{\"u}r den klinischen Einsatz als ausreichend erachtet. In prospektiven Anwendungsstudien sollte der KI-Algorithmus auf klinische Relevanz getestet werden.}, language = {de} } @misc{RoserMeinikheimMendeletal., author = {Roser, David and Meinikheim, Michael and Mendel, Robert and Palm, Christoph and Muzalyova, Anna and Rauber, David and R{\"u}ckert, Tobias and Parsa, Nasim and Byrne, Michael F. and Messmann, Helmut and Ebigbo, Alanna}, title = {Mensch-Maschine-Interaktion: Einfluss k{\"u}nstlicher Intelligenz auf das diagnostische Vertrauen von Endoskopikern bei der Beurteilung des Barrett-{\"O}sophagus}, series = {Zeitschrift f{\"u}r Gastroenterologie}, volume = {62}, journal = {Zeitschrift f{\"u}r Gastroenterologie}, number = {09}, publisher = {Georg Thieme Verlag KG}, doi = {10.1055/s-0044-1789656}, pages = {e575 -- e576}, abstract = {Ziele: Das Ziel der Studie war es, den Einfluss von KI auf die diagnostische Sicherheit (Konfidenzniveau) von Endoskopikern anhand von B{\"O}-Videos zu untersuchen und m{\"o}gliche Korrelationen mit der Untersuchungsqualit{\"a}t zu erforschen. Methodik: 22 Endoskopiker aus zw{\"o}lf Zentren mit unterschiedlicher Barrett-Erfahrung untersuchten 96 standardisierte Endoskopievideos. Die Untersucher wurden in Experten und Nicht-Experten eingeteilt und nach dem Zufallsprinzip f{\"u}r die Bewertung der Videos mit oder ohne KI eingeteilt. Die Teilnehmer wurden in zwei Gruppen aufgeteilt: Arm A bewertete zun{\"a}chst Videos ohne KI und dann mit KI, w{\"a}hrend Arm B die umgekehrte Reihenfolge einhielt. Die Untersucher hatten die Aufgabe, B{\"O}-assoziierte Neoplasien zu erkennen und ihr Konfidenzniveau sowohl mit als auch ohne KI auf einer Skala von 0 bis 9 anzugeben. Ergebnis: In Arm A erh{\"o}hte der Einsatz von KI das Konfidenzniveau bei beiden signifikant (p<0.001). Bemerkenswert ist, dass jedoch nur Nicht-Experten durch die KI eine signifikante Verbesserung der Sensitivit{\"a}t und Spezifit{\"a}t (p<0.001 bzw. p<0.05) erfuhren. W{\"a}hrend Experten ohne KI im Vergleich zu Nicht-Experten mit KI ein h{\"o}heres Konfidenzniveau aufwiesen, gab es keinen signifikanten Unterschied in der Genauigkeit. In Arm B zeigten beide Gruppen eine signifikante Abnahme des Konfidenzniveaus (p<0.001) bei gleichbleibender Genauigkeit. Dar{\"u}ber hinaus wurden in 9\% der Entscheidungen trotz korrekter KI eine falsche Wahl getroffen. Schlussfolgerung: Der Einsatz k{\"u}nstlicher Intelligenz steigerte das Konfidenzniveau sowohl bei Experten als auch bei Nicht-Experten signifikant - ein Effekt, der im Studienmodell reversibel war. Dar{\"u}ber hinaus wiesen Experten mit oder ohne KI durchweg h{\"o}here Konfidenzniveaus auf als Nicht-Experten mit KI, trotz vergleichbarer Ergebnisse. Zudem konnte beobachtet werden, dass die Untersucher in 9\% der F{\"a}lle die KI zuungunsten des Patienten ignorierten.}, language = {de} }