Congratulations to Md Ashikuzzaman! His first-author paper titled “Ultrasound Displacement Tracking Techniques for Post-Stroke Myofascial Shear Strain Quantification” was accepted for publication in IEEE Transactions on Biomedical Engineering.

This study is the first to assess and compare four ultrasound displacement tracking algorithms employed to quantify myofascial shear strain in patients with post-stroke shoulder pain. Leveraging simulated and in vivo data, we additionally validated a novel clinical hypothesis that the mobility of myofascial shear strain is reduced in the painful stroke-affected (i.e., paretic) shoulders, relative to each patient’s non-paretic shoulders. One of the tracking algorithms, named L1-SOUL-Search, offered smooth displacement gradients and sharp visualization of fascial layers, advancing the potential of shear strain imaging as a biomarker for myofascial dysfunction. In addition, a deep learning approach named RAFT was demonstrated as an option to address computational bottlenecks, showing promise for real-time clinical deployment. This work lays the foundation necessary to develop reliable image-based biomarkers to guide diagnosis and treatment in musculoskeletal health applications.

Citation: Md Ashikuzzaman, Jonny Huang, Steve Bonwit, Azin Etemadimanesh, Ali Ghasemi, Patrick Debs, Robert Nickl, Jamie Enslein, Laura M. Fayad, Preeti Raghavan, and Muyinatu A. Lediju Bell, “Ultrasound Displacement Tracking Techniques for Post-Stroke Myofascial Shear Strain Quantification”, IEEE Transactions on Biomedical Engineering (accepted June 13, 2025) [pdf]

Congratulations to Mardava Gubbi! His first-author paper entitled “In Vivo Demonstration of Deep Learning-Based Photoacoustic Visual Servoing System” was accepted for publication in IEEE Transactions on Biomedical Engineering.

This paper is the first to present a real-time deep learning-based photoacoustic visual servoing system utilizing both object detection and instance segmentation to estimate catheter tip positions in three spatial dimensions from a single frame of two-dimensional of raw photoacoustic channel data. This system was designed with two features not present in previous amplitude-based or deep learning-based photoacoustic visual servoing systems. First, we estimated the location of the catheter tip along the elevation dimension of the transducer with the input being a single frame of two-dimensional of raw channel data. Second, the integration of force control into our visual servoing system improved the tracking of targets across uneven imaging surfaces. These features improve the potential of our novel visual servoing system for clinical translation in cardiac catheterizations procedures (and other minimally invasive interventional procedures).

Citation: M. R. Gubbi, A. Kolandaivelu, N. Venkatayogi, J. Zhang, P. Warbal, G. C. Keene, M. Khairalseed, J. Chrispin, and M. A. Lediju Bell, “In Vivo Demonstration of Deep Learning-Based Photoacoustic Visual Servoing System”, IEEE Transactions on Biomedical Engineering (accepted June 23, 2025) [pdf]

Congratulations to Mardava Gubbi! His first-author paper entitled “Deep learning to localize photoacoustic sources in three dimensions: Theory and implementation” was accepted for publication in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control.

This paper introduces two novel deep learning-based three-dimensional (3D) photoacoustic point source localization systems. One system uses an object detection-based approach with elevation-encoded classification (System A). The second system uses our novel theoretical framework relating the three-dimensional location of photoacoustic point sources, sound speeds, and waveform shapes in photoacoustic raw channel data (System B). System B leverages our theoretical framework to implement a novel instance segmentation and least squares optimization-based photoacoustic point source location and sound speed estimation approach. This instance segmentation-based system improves localization and sound speed estimation performance in simulated, phantom, and ex vivo data compared to traditional methods. Overall, the results enhance the potential applicability of photoacoustic guidance to surgical and interventional scenarios.

Citation: M. R. Gubbi and M. A. Lediju Bell, Deep Learning to Localize Photoacoustic Sources in Three Dimensions: Theory and Implementation, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control (accepted April 15, 2025) [pdf]

Congratulations to Nethra and Junhao! Their first-author papers were recently published in Biomedical Optics Express and Ultrasound in Medicine and Biology, respectively.

Multispectral photoacoustic imaging of breast cancer tissue with histopathology validation describes the first known work to use multispectral photoacoustic imaging with optical wavelengths up to 2000 nm to provide pathological information about breast tissue without traditional staining staining. We leveraged machine learning to differentiate between healthy and cancerous tissues based on photoacoustic spectra. Using our custom imaging platform, this method generates a virtual stain of a microscope tissue slide in under 30 minutes, compared to the more than 24 hours required for conventional H&E staining. The co-registered ultrasound images provide complementary anatomical context to the pathological information derived from multispectral photoacoustic images. Our approach has the potential to enable real-time tumor margin determination during biopsy or surgery.

Congratulations to Md Ashikuzzaman! His first-author paper entitled MixTURE: L1-Norm-Based Mixed Second-Order Continuity in Strain Tensor Ultrasound Elastography was accepted for publication in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control.

This work is the first to introduce an L1-norm-based second-order regularizer in a mechanically-inspired total strain tensor imaging framework named — L1-norm Mixed derivative for Total UltRasound Elastography — or, L1-MixTURE when abbreviated. Displacement tracking of ultrasound images can be implemented by optimizing a cost function consisting of a data term, a mechanical congruency term, and first- and second-order continuity terms. This approach recently provided a promising solution to two-dimensional axial and lateral displacement tracking in ultrasound strain elastography. However, the associated second-order regularizer only considers the unmixed second derivatives and disregards the mixed derivatives, thereby providing suboptimal noise suppression and limiting possibilities for total strain tensor imaging. We improved axial, lateral, axial shear, and lateral shear strain estimation quality by formulating and optimizing a novel L1-norm-based second-order regularizer that penalizes both mixed and unmixed displacement derivatives. Results are promising to advance the state-of-the-art in elastography-guided diagnostic and interventional decisions.

Citation: Ashikuzzaman M, Sharma A, Venkatayogi N, Oluyemi E, Myers K, Ambinder E, Rivaz H, Bell MAL, MixTURE: L1-Norm-Based Mixed Second-Order Continuity in Strain Tensor Ultrasound Elastography, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control (accepted August 13, 2024) [pdf]

Congratulations to Junior Arroyo! His first-author paper entitled Predictive model for laser-induced tissue necrosis with immunohistochemistry validation was accepted for publication in the newest SPIE journal, Biophotonics Discovery.

This work is the first to demonstrate a unified theoretical, computational, and experimental approach with quantitative immunohistochemistry (IHC) validation to determine laser safety for biological tissues other than skin or eyes. Although photoacoustic imaging has the potential to provide critical guidance in surgical interventions, its widespread use is challenged by the absence of applicable safety guidelines across diverse target tissues. Maximum permissible exposure (MPE) guidelines currently focus solely on skin and eyes. Results are promising to provide tissue-specific MPE guidelines to maintain healthy liver tissue during laser-based optical and photoacoustic surgeries and interventions. The presented approach and associated outcomes are promising for the introduction of tissue-specific safety guidelines for photoacoustic imaging and other optics-based imaging technologies that are designed to maximize signal-to-noise ratios while being designated as safe for patient use. In addition, the presented simulation framework and corresponding experimental protocols may be applied to other internal organs to achieve similar benefits.

Citation: Arroyo J, Sharma A, Zhang J, Bell MAL, Predictive model for laser-induced tissue necrosis with immunohistochemistry validation, Biophotonics Discovery 1(2):025003, 2024 [pdf]

Congratulations to Mardava Gubbi! His first-author paper entitled Theoretical Framework to Predict Generalized Contrast-to-Noise Ratios of Photoacoustic Images With Applications to Computer Vision was accepted for publication in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control.

This work is the first to the first to present a novel framework to establish relationships among photoacoustic imaging system parameters, image quality, and computer vision-based task performance. Our framework leverages gCNR to quantify the relationships between system parameters (e.g., channel SNR, laser energy) and photoacoustic image quality. Within this framework, we present a theoretical derivation of gCNR predictions based on using the statistics of the target and background signal powers, then validate these predictions on simulated, experimental, and in vivo data. This framework was then leveraged to quantify the accuracy of a photoacoustic target segmentation algorithm as a function of gCNR and demonstrate the robustness of gCNR to thresholding, with possible extensions to other computer vision-based tasks (e.g., target tracking and image classification) and to improve the overall photoacoustic imaging system design process.

Citation: Gubbi MR, Gonzalez EA, Bell MAL,Theoretical Framework to Predict Generalized Contrast-to-Noise Ratios of Photoacoustic Images With Applications to Computer Vision,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 69(6):2098-2114, 2022 [pdf]

Prof. Bell and Lingyi Zhao co-authored an invited review entitled, A Review of Deep Learning Applications in Lung Ultrasound Imaging of COVID-19 Patients, which was recently published in BME Frontiers.

This review is focused on deep learning applications in lung ultrasound imaging of COVID-19 and provides a comprehensive overview of ultrasound systems utilized for data acquisition, associated datasets, deep learning models, and comparative performance..

Citation: Lingyi Zhao, Muyinatu A. Lediju Bell, “A Review of Deep Learning Applications in Lung Ultrasound Imaging of COVID-19 Patients“, BME Frontiers, vol. 2022, Article ID 9780173, 17 pages, 2022. https://doi.org/10.34133/2022/9780173

Congratulations to Eduardo González! His first-author paper entitled Acoustic Frequency-Based Approach for Identification of Photoacoustic Surgical Biomarkers was accepted for publication in the journal Frontiers in Photonics.

This paper demonstrates a novel approach to accurately identify biological markers by analyzing the acoustic frequency response from either a single-wavelength emission (i.e., single-wavelength atlas method) or two consecutive wavelength emissions (i.e., dual-wavelength atlas method). The proposed approach relies on training sets to identify photoacoustic-sensitive materials, is robust against changes in fluence levels, and has comparable sensitivity, specificity, and accuracy to those obtained with conventional and enhanced spectral unmixing methods. This paper is part of the journal’s research topic Biophotonics Technologies for Clinical Translation.

Citation: González EA, Graham CA, Bell MAL, Acoustic frequency-based approach for identification of photoacoustic surgical biomarkers, Frontiers in Photonics, 2, 2021 [pdf]

Congratulations to Alycen Wiacek! Her first-author paper entitled Photoacoustic-guided laparoscopic and open hysterectomy procedures demonstrated with human cadavers was accepted for publication in IEEE Transactions on Medical Imaging.

This work is the first to demonstrate a novel method for photoacoustic image-guided hysterectomies within the realistic imaging environment of a human cadaver during both open and laparoscopic procedures. With a contrast agent injected into the ureter, two laser wavelengths can be used to create a simultaneous display of the ureter and the uterine artery. This dual-wavelength approach was then integrated to create a novel surgical guidance system by estimating the tool-to-ureter distance and mapping that distance to an audible signal, similar to the parking sensor on a modern automobile. This auditory signal is intended to alert surgeons who are operating too closely to the ureter, which can lead to multiple life-threatening complications caused by accidental injury to the ureter during surgery.

Citation: Wiacek A, Wang KC, Wu H, Bell MAL, Photoacoustic-guided laparoscopic and open hysterectomy procedures demonstrated with human cadavers, IEEE Transactions on Medical Imaging (accepted May 13, 2021) [pdf]

Prof. Bell and Alycen Wiacek co-authored an invited review entitled, Photoacoustic-Guided Surgery from Head to Toe, which was recently published in Biomedical Optics Express.

This review covers multiple aspects of photoacoustic imaging to guide surgical & related non-surgical interventions, including a discussion of complete systems and tools needed to maximize success.

Citation: Alycen Wiacek and Muyinatu A. Lediju Bell, “Photoacoustic-guided surgery from head to toe [Invited],” Biomed. Opt. Express 12, 2079-2117 (2021) [bibtex]