-
Swin UNETR segmentation with automated geometry filtering for biomechanical modeling of knee joint cartilage
Authors:
Reza Kakavand,
Peyman Tahghighi,
Reza Ahmadi,
W. Brent Edwards,
Amin Komeili
Abstract:
Simulation studies, such as finite element (FE) modeling, offer insights into knee joint biomechanics, which may not be achieved through experimental methods without direct involvement of patients. While generic FE models have been used to predict tissue biomechanics, they overlook variations in population-specific geometry, loading, and material properties. In contrast, subject-specific models ac…
▽ More
Simulation studies, such as finite element (FE) modeling, offer insights into knee joint biomechanics, which may not be achieved through experimental methods without direct involvement of patients. While generic FE models have been used to predict tissue biomechanics, they overlook variations in population-specific geometry, loading, and material properties. In contrast, subject-specific models account for these factors, delivering enhanced predictive precision but requiring significant effort and time for development. This study aimed to facilitate subject-specific knee joint FE modeling by integrating an automated cartilage segmentation algorithm using a 3D Swin UNETR. This algorithm provided initial segmentation of knee cartilage, followed by automated geometry filtering to refine surface roughness and continuity. In addition to the standard metrics of image segmentation performance, such as Dice similarity coefficient (DSC) and Hausdorff distance, the method's effectiveness was also assessed in FE simulation. Nine pairs of knee cartilage FE models, using manual and automated segmentation methods, were developed to compare the predicted stress and strain responses during gait. The automated segmentation achieved high Dice similarity coefficients of 89.4% for femoral and 85.1% for tibial cartilage, with a Hausdorff distance of 2.3 mm between the automated and manual segmentation. Mechanical results including maximum principal stress and strain, fluid pressure, fibril strain, and contact area showed no significant differences between the manual and automated FE models. These findings demonstrate the effectiveness of the proposed automated segmentation method in creating accurate knee joint FE models.
△ Less
Submitted 8 July, 2024;
originally announced July 2024.
-
Federated Learning Enables Big Data for Rare Cancer Boundary Detection
Authors:
Sarthak Pati,
Ujjwal Baid,
Brandon Edwards,
Micah Sheller,
Shih-Han Wang,
G Anthony Reina,
Patrick Foley,
Alexey Gruzdev,
Deepthi Karkada,
Christos Davatzikos,
Chiharu Sako,
Satyam Ghodasara,
Michel Bilello,
Suyash Mohan,
Philipp Vollmuth,
Gianluca Brugnara,
Chandrakanth J Preetha,
Felix Sahm,
Klaus Maier-Hein,
Maximilian Zenk,
Martin Bendszus,
Wolfgang Wick,
Evan Calabrese,
Jeffrey Rudie,
Javier Villanueva-Meyer
, et al. (254 additional authors not shown)
Abstract:
Although machine learning (ML) has shown promise in numerous domains, there are concerns about generalizability to out-of-sample data. This is currently addressed by centrally sharing ample, and importantly diverse, data from multiple sites. However, such centralization is challenging to scale (or even not feasible) due to various limitations. Federated ML (FL) provides an alternative to train acc…
▽ More
Although machine learning (ML) has shown promise in numerous domains, there are concerns about generalizability to out-of-sample data. This is currently addressed by centrally sharing ample, and importantly diverse, data from multiple sites. However, such centralization is challenging to scale (or even not feasible) due to various limitations. Federated ML (FL) provides an alternative to train accurate and generalizable ML models, by only sharing numerical model updates. Here we present findings from the largest FL study to-date, involving data from 71 healthcare institutions across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, utilizing the largest dataset of such patients ever used in the literature (25,256 MRI scans from 6,314 patients). We demonstrate a 33% improvement over a publicly trained model to delineate the surgically targetable tumor, and 23% improvement over the tumor's entire extent. We anticipate our study to: 1) enable more studies in healthcare informed by large and diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further quantitative analyses for glioblastoma via performance optimization of our consensus model for eventual public release, and 3) demonstrate the effectiveness of FL at such scale and task complexity as a paradigm shift for multi-site collaborations, alleviating the need for data sharing.
△ Less
Submitted 25 April, 2022; v1 submitted 22 April, 2022;
originally announced April 2022.
-
The Federated Tumor Segmentation (FeTS) Challenge
Authors:
Sarthak Pati,
Ujjwal Baid,
Maximilian Zenk,
Brandon Edwards,
Micah Sheller,
G. Anthony Reina,
Patrick Foley,
Alexey Gruzdev,
Jason Martin,
Shadi Albarqouni,
Yong Chen,
Russell Taki Shinohara,
Annika Reinke,
David Zimmerer,
John B. Freymann,
Justin S. Kirby,
Christos Davatzikos,
Rivka R. Colen,
Aikaterini Kotrotsou,
Daniel Marcus,
Mikhail Milchenko,
Arash Nazeri,
Hassan Fathallah-Shaykh,
Roland Wiest,
Andras Jakab
, et al. (7 additional authors not shown)
Abstract:
This manuscript describes the first challenge on Federated Learning, namely the Federated Tumor Segmentation (FeTS) challenge 2021. International challenges have become the standard for validation of biomedical image analysis methods. However, the actual performance of participating (even the winning) algorithms on "real-world" clinical data often remains unclear, as the data included in challenge…
▽ More
This manuscript describes the first challenge on Federated Learning, namely the Federated Tumor Segmentation (FeTS) challenge 2021. International challenges have become the standard for validation of biomedical image analysis methods. However, the actual performance of participating (even the winning) algorithms on "real-world" clinical data often remains unclear, as the data included in challenges are usually acquired in very controlled settings at few institutions. The seemingly obvious solution of just collecting increasingly more data from more institutions in such challenges does not scale well due to privacy and ownership hurdles. Towards alleviating these concerns, we are proposing the FeTS challenge 2021 to cater towards both the development and the evaluation of models for the segmentation of intrinsically heterogeneous (in appearance, shape, and histology) brain tumors, namely gliomas. Specifically, the FeTS 2021 challenge uses clinically acquired, multi-institutional magnetic resonance imaging (MRI) scans from the BraTS 2020 challenge, as well as from various remote independent institutions included in the collaborative network of a real-world federation (https://www.fets.ai/). The goals of the FeTS challenge are directly represented by the two included tasks: 1) the identification of the optimal weight aggregation approach towards the training of a consensus model that has gained knowledge via federated learning from multiple geographically distinct institutions, while their data are always retained within each institution, and 2) the federated evaluation of the generalizability of brain tumor segmentation models "in the wild", i.e. on data from institutional distributions that were not part of the training datasets.
△ Less
Submitted 13 May, 2021; v1 submitted 12 May, 2021;
originally announced May 2021.
-
A single inverse-designed photonic structure that performs parallel computing
Authors:
Miguel Camacho,
Brian Edwards,
Nader Engheta
Abstract:
In the search for improved computational capabilities, conventional microelectronic computers are facing various problems arising from the miniaturization and concentration of active electronics devices (1-2). Therefore, researchers have been exploring several paths for the next generation of computing platforms, which could exploit various physical phenomena for solving mathematical problems at h…
▽ More
In the search for improved computational capabilities, conventional microelectronic computers are facing various problems arising from the miniaturization and concentration of active electronics devices (1-2). Therefore, researchers have been exploring several paths for the next generation of computing platforms, which could exploit various physical phenomena for solving mathematical problems at higher speeds and larger capacities. Among others, physical systems described by waves, such as photonic and quantum devices, have been utilized to compute the solution of mathematical problems (1-18). However, previous devices have not fully exploited the linearity of the wave equation, which as we show here, allows for the simultaneous parallel solution of several independent mathematical problems within the same device. In this Letter, we demonstrate, theoretically and experimentally, that a transmissive cavity filled with a judiciously tailored dielectric distribution and embedded in a multi-frequency feedback loop can calculate the solutions of an arbitrary number of mathematical problems simultaneously. We design, build, and test a computing structure at microwave frequencies that solves two independent integral equations with any two arbitrary inputs. We offer another design that can invert four arbitrary 5x5 matrices, confirming its functionality with numerical simulations. We believe our results presented here can provide "coincident computing" and pave the way for the design of low-power, ultrafast, parallel photonic analog computing devices for sensing and signal processing in embedded computing applications.
△ Less
Submitted 2 September, 2020;
originally announced September 2020.
