1969
September 1, 2022Knowledge Representation
October 1, 2022Innovation to Transform
AI is rapidly advancing our ability to extend health care delivery service experiences in so many directions from making appointments, to monitoring heart rhythms...
New synthetic biomaterial developed for wound repair in mechanically dynamic tissue
Researchers and biochemical engineers from McGill University in Montreal, Canada, have developed a novel injectable hydrogel for repairing mechanically dynamic tissues. The experimental gel formula with pore-forming double-network hydrogel withstood high-frequency biomechanical stimulations, unlike existing hydrogels, while allowing direct medium perfusion and cell encapsulation and delivery, spreading, and proliferation. These characteristics are necessarily valuable for fabrics such as hearts and other dynamic tissues. The team conducted the work at McGill University and has developed this porous double networks hydrogel that could also have further applications in tissue engineering, fabrication, and drug delivery. (1)
Researchers and biochemical engineers from McGill University in Montreal, Canada, have developed a novel injectable hydrogel for repairing mechanically dynamic tissues. The experimental gel formula with pore-forming double-network hydrogel withstood high-frequency biomechanical stimulations, unlike existing hydrogels, while allowing direct medium perfusion and cell encapsulation and delivery, spreading, and proliferation.
These characteristics are necessarily valuable for fabrics such as hearts and other dynamic tissues. The team conducted the work at McGill University and has developed this porous double networks hydrogel that could also have further applications in tissue engineering, fabrication, and drug delivery.(1)
Printed' skin' hastens to heal, lessens scarring
Chemical engineers and researchers from the University of Birmingham and the University of Huddersfield in the UK have developed an approach to print skin layer equivalents, which may play an important future role in wound healing, especially chronic wounds. They formulated the bio-ink with chemical and cellular properties to emulate the three layers of skin: the hypodermis, the dermis, and the epidermis. Additionally, they created a gel-like material to support the skin equivalent even after twisting and altering the structure of the substance as it formed. The method is the first of its kind. During printing, the skin layers are deposited within the gel, which holds everything correctly. After printing, the team washed away the support material and gel, leaving behind the layered skin equivalent. This method is higher in printing resolution than its predecessors. After this process, the authors tested the skin substitute by cutting a hole in pig tissue and printing a skin equivalent to fill the gap. After culturing the model system for 14 days, signs of wound repair were evident.(2)
Chemical engineers and researchers from the University of Birmingham and the University of Huddersfield in the UK have developed an approach to print skin layer equivalents, which may play an important future role in wound healing, especially chronic wounds. They formulated the bio-ink with chemical and cellular properties to emulate the three layers of skin: the hypodermis, the dermis, and the epidermis. Additionally, they created a gel-like material to support the skin equivalent even after twisting and altering the structure of the substance as it formed.
The method is the first of its kind. During printing, the skin layers are deposited within the gel, which holds everything correctly. After printing, the team washed away the support material and gel, leaving behind the layered skin equivalent. This method is higher in printing resolution than its predecessors. After this process, the authors tested the skin substitute by cutting a hole in pig tissue and printing a skin equivalent to fill the gap. After culturing the model system for 14 days, signs of wound repair were evident.(2)
AI Dosimetry for Molecular Radiotherapy
MIM Software Inc. is a Cleveland-based private company and a leading global provider of medical imaging software. The company announced critical innovations to MIM SurePlan™ MRT, its software package for Molecular Radiotherapy. There are several challenges when implementing clinical dosimetry. One of them, for example, “is using time and activity curve information from multiple SPECT/CTs after the first cycle to calculate the dose for the remaining cycles.” For the following cycles, additional imaging is required to account for the total dose of absorbed radiation. The new AI segmentation models simplify the process for dosimetry through algorithms that “use a single SPECT/CT acquired around a carefully chosen time after each cycle of therapy to calculate the absorbed dose.”
AI segmentation provides significant time savings and enhanced results compared to manual calculations. Also, another advantage is that MIM Software’s AI segmentation models can be deployed locally or through the cloud. This cloud-based accessibility increases the reaching range of this initiative to any place with an internet connection.(3)
MIM Software Inc. is a Cleveland-based private company and a leading global provider of medical imaging software. The company announced critical innovations to MIM SurePlan™ MRT, its software package for Molecular Radiotherapy. There are several challenges when implementing clinical dosimetry. One of them, for example, “is using time and activity curve information from multiple SPECT/CTs after the first cycle to calculate the dose for the remaining cycles.
“For the following cycles, additional imaging is required to account for the total dose of absorbed radiation. The new AI segmentation models simplify the process for dosimetry through algorithms that “use a single SPECT/CT acquired around a carefully chosen time after each cycle of therapy to calculate the absorbed dose.” AI segmentation provides significant time savings and enhanced results compared to manual calculations. Also, another advantage is that MIM Software’s AI segmentation models can be deployed locally or through the cloud. This cloud-based accessibility increases the reaching range of this initiative to any place with an internet connection.(3)
Soft exoskeleton device for stroke survivors will be for use at home and in the community
ReWalk Robotics, Ltd. develops, manufactures, and markets portable robotic exoskeletons worldwide. Its models, designed specifically for people with lower extremity disabilities resulting from spinal cord injury or stroke, have announced that one of its best devices, the ReBoot, have received innovative device designation from the Food and Drug Administration (FDA). The ReBoot is a very lightweight, battery-powered orthopedic exo-suit with artificial intelligence and integrated microprocessors intended to assist the ambulatory functions of people with reduced ankle and foot function related to neurological injuries, such as stroke. The ReBoot operates in conjunction with the muscles of the affected leg to help people not only maintain a secure foot position but also get off the ground, meaning they can improve their gait; the device is customized and assessed according to each person’s needs.(4)
ReWalk Robotics, Ltd. develops, manufactures, and markets portable robotic exoskeletons worldwide. Its models, designed specifically for people with lower extremity disabilities resulting from spinal cord injury or stroke, have announced that one of its best devices, the ReBoot, have received innovative device designation from the Food and Drug Administration (FDA). The ReBoot is a very lightweight, battery-powered orthopedic exo-suit with artificial intelligence and integrated microprocessors intended to assist the ambulatory functions of people with reduced ankle and foot function related to neurological injuries, such as stroke.
The ReBoot operates in conjunction with the muscles of the affected leg to help people not only maintain a secure foot position but also get off the ground, meaning they can improve their gait; the device is customized and assessed according to each person’s needs.(4)
Novel camera images objects behind barriers
Florian Willomitzer, Ph.D. of Northwestern University in Illinois, and fellow electrical engineers have created a prototype holographic camera capable of revealing structures hidden by obstructions called Synthetic Wavelength Holography. The process involves firing laser beams with slightly different wavelengths through obstacles, either a wall or a translucent material, to reach a hidden target. The reflected wavelengths are captured and superimposed to produce an interference pattern that reveals the distances of objects hidden from view and then reconstructed using artificial intelligence to create a 3D image. The company suggests combining this technique with ultrasound imaging could allow clinicians to see around bones or tiny blood vessels under the skin. However, both researchers say more work and testing are needed to make this vision a reality.(5)
Florian Willomitzer, Ph.D. of Northwestern University in Illinois, and fellow electrical engineers have created a prototype holographic camera capable of revealing structures hidden by obstructions called Synthetic Wavelength Holography. The process involves firing laser beams with slightly different wavelengths through obstacles, either a wall or a translucent material, to reach a hidden target.
The reflected wavelengths are captured and superimposed to produce an interference pattern that reveals the distances of objects hidden from view and then reconstructed using artificial intelligence to create a 3D image. The company suggests combining this technique with ultrasound imaging could allow clinicians to see around bones or tiny blood vessels under the skin. However, both researchers say more work and testing are needed to make this vision a reality.(5)
Behavioral health AI predicts effectiveness of talk therapy
Artificial intelligence can help measure the effectiveness of cognitive-behavioral therapy, according to new research from the University of Southern California. Engineering scientists at the University of Southern California, USC’s Signal Analysis and Interpretation Laboratory, set out to explore whether AI could do this job using transcripts of more than 1,100 real conversations with patients. Early results were satisfactory, with the AI achieving 73% accuracy of what a human evaluator could achieve. Nikolaos Flemotomos, a Ph.D. student in engineering at USC, believe this is the most extensive study to date of automated assessment of this type of therapy and the first using 100 percent real people and conversations.
One of the biggest challenges for the AI in the process was capturing multiple interlocutors and discerning the meaning of the text-based conversation. However, their creation proved effective in assessing therapists’ interpersonal skills, how they structure sessions, and their capacity to establish rapport, among other factors. USC specialists believe their AI tool could be scaled up to help bolster staff and address the growing demand for these services. Their main goal is not to replace human supervision. However, only to increase the effectiveness of raters and assist in self-assessment. Moving from text-based analysis to audible conversation is an avenue for further exploration.(6)
Artificial intelligence can help measure the effectiveness of cognitive-behavioral therapy, according to new research from the University of Southern California. Engineering scientists at the University of Southern California, USC’s Signal Analysis and Interpretation Laboratory, set out to explore whether AI could do this job using transcripts of more than 1,100 real conversations with patients.
Early results were satisfactory, with the AI achieving 73% accuracy of what a human evaluator could achieve. Nikolaos Flemotomos, a Ph.D. student in engineering at USC, believe this is the most extensive study to date of automated assessment of this type of therapy and the first using 100 percent real people and conversations. One of the biggest challenges for the AI in the process was capturing multiple interlocutors and discerning the meaning of the text-based conversation. However, their creation proved effective in assessing therapists’ interpersonal skills, how they structure sessions, and their capacity to establish rapport, among other factors. USC specialists believe their AI tool could be scaled up to help bolster staff and address the growing demand for these services. Their main goal is not to replace human supervision. However, only to increase the effectiveness of raters and assist in self-assessment. Moving from text-based analysis to audible conversation is an avenue for further exploration.(6)
AI scores 1 against a knee injury common among athletes
Physicians at the Hospital for Special Surgery in New York have used machine learning and deep learning to predict outcomes after arthroscopic knee reconstruction of the anterior cruciate ligament (ACL) after a traumatic sports accident. This prediction can help distinguish ACL-injured individuals who are likely to benefit significantly from surgery from those who may have to adjust their post-surgical expectations or even forgo surgery and the use of rehabilitation. The team led by David Altchek, MD, medical director for the New York Mets and consulting physician for the NBA, trained six different AI algorithms on retrospective data from 442 patients from a registry fed by 27 experienced knee surgeons at their institution. This yielded data that could be used in the future to improve the management of acute injuries in athletes. (7)
Physicians at the Hospital for Special Surgery in New York have used machine learning and deep learning to predict outcomes after arthroscopic knee reconstruction of the anterior cruciate ligament (ACL) after a traumatic sports accident. This prediction can help distinguish ACL-injured individuals who are likely to benefit significantly from surgery from those who may have to adjust their post-surgical expectations or even forgo surgery and the use of rehabilitation.
The team led by David Altchek, MD, medical director for the New York Mets and consulting physician for the NBA, trained six different AI algorithms on retrospective data from 442 patients from a registry fed by 27 experienced knee surgeons at their institution. This yielded data that could be used in the future to improve the management of acute injuries in athletes.(7)
AI predicts tumor metastasis from the colon to lymph nodes
A convolutional neural network, or CNN, has demonstrated its ability to predict the spread of colorectal cancer to lymph nodes by automated analysis of histologic slices. In the study on which the finding is based, the tool behaved similarly to a manual clinical classifier using patient data such as age, sex, and tumor characteristics such as stage and location. The research was conducted at the German Cancer Research Center Heidelberg and is currently published in the European Journal of Cancer. Titus Brinker, author, and head of the study, points out that lymph node involvement is crucial for making timely treatment decisions for patients with colorectal cancer. They describe their work training and in-house validation of a CNN on whole slide images from more than 2,400 patients in a database at their center.
They then developed an external test set using data from 582 patients from the public Cancer Genome Atlas. The analysis found that their SBAIP (short for slide-based AI predictor) achieved an area under receiver operating characteristics of 71.0%. In comparison, the clinical classifier achieved 67.0%. Combining both classifiers resulted in a peak performance of 74.1%.(4)
A convolutional neural network, or CNN, has demonstrated its ability to predict the spread of colorectal cancer to lymph nodes by automated analysis of histologic slices. In the study on which the finding is based, the tool behaved similarly to a manual clinical classifier using patient data such as age, sex, and tumor characteristics such as stage and location. The research was conducted at the German Cancer Research Center Heidelberg and is currently published in the European Journal of Cancer.
They then developed an external test set using data from 582 patients from the public Cancer Genome Atlas. The analysis found that their SBAIP (short for slide-based AI predictor) achieved an area under receiver operating characteristics of 71.0%. In comparison, the clinical classifier achieved 67.0%. Combining both classifiers resulted in a peak performance of 74.1%.(4)
Titus Brinker, author, and head of the study, points out that lymph node involvement is crucial for making timely treatment decisions for patients with colorectal cancer. They describe their work training and in-house validation of a CNN on whole slide images from more than 2,400 patients in a database at their center.