Sweat Sensor Skin-like Microfluidic Device
July 1, 2022Article of the Month – July 2022
July 1, 2022Radiology
AI has been running rampant in radiology circles. Sometimes referred to as machine learning, AI can and will optimize radiologists workflows...
Imaging Technology Could Deliver Finely Detailed Images of the Human Brain
The Royal College London and UCL analysts have advanced high-fidelity clinical imaging of the human brain that may well change existing practices. Ultrasound cannot effectively penetrate through the bone, a problem they overcame with their new helmet-like gadget.
The relatively small imaging device is portable enough to carry and use during ambulance transfers enabling immediate investigation before arrival at the hospital. The royal analysts have adjusted seismic information and a computational procedure called total waveform reversal (FWI) to restorative imaging.
The investigators created a helmet lined with an array of acoustic transducers to send sound waves through the skull. The ultrasound energy propagating through the head is recorded via the helmet on a computer. FWI is then used to analyze the repercussions of the sound throughout the skull, constructing a 3D image of the interior.(1)
Imaging Technology Could Deliver Finely Detailed Images of the Human Brain
The Royal College London and UCL analysts have advanced high-fidelity clinical imaging of the human brain that may well change existing practices. Ultrasound cannot effectively penetrate through the bone, a problem they overcame with their new helmet-like gadget. The relatively small imaging device is portable enough to carry and use during ambulance transfers enabling immediate investigation before arrival at the hospital.
The royal analysts have adjusted seismic information and a computational procedure called total waveform reversal (FWI) to restorative imaging. The investigators created a helmet lined with an array of acoustic transducers to send sound waves through the skull. The ultrasound energy propagating through the head is recorded via the helmet on a computer. FWI is then used to analyze the repercussions of the sound throughout the skull, constructing a 3D image of the interior.(1)
Artificial Intelligence for pediatric elbow radiographs
The radiology program at Baylor College of Medicine in Houston developed a means of using AI to overcome diagnostic challenges by automating the analysis of pediatric elbow radiographs. The model leverages a convolutional neural network (CNN) and a recurrent neural network (RNN) to analyze and match multiple images together.
The CNN is a deep learning algorithm that receives an input image, assigns learnable weights and biases to various aspects or elements in the picture, and differentiates one object from another. The RNN then identifies common sequences of data and images, decomposing the latter into smaller fragments to further analyze their distribution across the picture.
The researchers tested the method on 21,456 X-rays containing 58,817 images of pediatric elbows and associated radiology reports, all captured at Texas Children’s Hospital in Houston. The studied dataset’s accuracy was 88%, with a sensitivity of 91% and specificity of 84%.(2)
Artificial Intelligence for pediatric elbow radiographs
The radiology program at Baylor College of Medicine in Houston developed a means of using AI to overcome diagnostic challenges by automating the analysis of pediatric elbow radiographs. The model leverages a convolutional neural network (CNN) and a recurrent neural network (RNN) to analyze and match multiple images together.The CNN is a deep learning algorithm that receives an input image, assigns learnable weights and biases to various aspects or elements in the picture, and differentiates one object from another.
The RNN then identifies common sequences of data and images, decomposing the latter into smaller fragments to further analyze their distribution across the picture. The researchers tested the method on 21,456 X-rays containing 58,817 images of pediatric elbows and associated radiology reports, all captured at Texas Children’s Hospital in Houston. The studied dataset’s accuracy was 88%, with a sensitivity of 91% and specificity of 84%.(2)
Could the Emerging Artificial Intelligence Platforms Improve Cancer Detection?
Saige-Dx enhances the detection of suspicious lesions on mammograms and reduces recalls and false positives rates, according to RadNet, an American radiology firm. Based on advanced deep learning algorithms, Quantib Prostate 2.0 provides automated segmentation of prostate zones and glands and lesion localization on the PI-RADS (Prostate Imaging Reporting & Data System) sector map. The company noted that the software program enhances the quality and speed of reporting with prostate MRI exams.(3)
Could the Emerging Artificial Intelligence Platforms Improve Cancer Detection?
Saige-Dx enhances the detection of suspicious lesions on mammograms and reduces recalls and false positives rates, according to RadNet, an American radiology firm. Based on advanced deep learning algorithms, Quantib Prostate 2.0 provides automated segmentation of prostate zones and glands and lesion localization on the PI-RADS (Prostate Imaging Reporting & Data System) sector map.
The company noted that the software program enhances the quality and speed of reporting with prostate MRI exams.(3)
AI Can Diagnose Osteoporosis on Hip X-ray
Statistics from the International Osteoporosis Foundation reveal that one in three women and one in five men worldwide over 50 years will experience osteoporotic fractures at some point. To assess bone mineral density, osteoporosis screening with dual-energy X-ray absorptiometry (DXA) is essential for timely interventions that reduce fracture risk. However, their low availability and high cost have limited their use for screening and post-treatment follow-up.
The new method combines imaging information with AI to diagnose osteoporosis from hip X-rays. Dr. Chae from Korea’s National Health Service used a database with the imaging data of the Seoul University Hospital to develop a model that can automatically diagnose osteoporosis from hip X-rays. The method combines radiomics, a series of image processing, and analysis methods to obtain information from the image with deep learning.
The researchers developed the deep-radionics model using almost 5,000 hip X-rays from 4,308 patients received over ten years. They created the models with various deep, clinical, and texture features and then tested them externally on 444 hip X-rays from another institution.(4)
AI Can Diagnose Osteoporosis on Hip X-ray
Statistics from the International Osteoporosis Foundation reveal that one in three women and one in five men worldwide over 50 years will experience osteoporotic fractures at some point.
To assess bone mineral density, osteoporosis screening with dual-energy X-ray absorptiometry (DXA) is essential for timely interventions that reduce fracture risk. However, their low availability and high cost have limited their use for screening and post-treatment follow-up.
The new method combines imaging information with AI to diagnose osteoporosis from hip X-rays. Dr. Chae from Korea’s National Health Service used a database with the imaging data of the Seoul University Hospital to develop a model that can automatically diagnose osteoporosis from hip X-rays. The method combines radiomics, a series of image processing, and analysis methods to obtain information from the image with deep learning. The researchers developed the deep-radionics model using almost 5,000 hip X-rays from 4,308 patients received over ten years. They created the models with various deep, clinical, and texture features and then tested them externally on 444 hip X-rays from another institution.(4)
AI Can Predict Cancer Risk of Lung Nodules
Pulmonary nodules present as small spots on the lungs on chest imaging. They have become a much more common finding as CT gained favor over X-rays for chest imaging. Dr. Vachani and colleagues from the Oncology Department at the Perelman School of Medicine, University of Pennsylvania in Philadelphia, evaluated an AI-based computer-aided diagnosis tool developed by Optellum Ltd. in Oxford, England. The objective is simple: assist clinicians in assessing pulmonary nodules on chest CT scans. While CT scans show many aspects of the nodule, such as size and border characteristics, AI can delve even more profoundly. A total of 300 chest CTs of indeterminate pulmonary nodules were used in the study. Analysis showed that using the AI tool improved the malignancy risk prediction of nodules on chest CT. It also enhanced the readers’ agreement with risk stratification and management recommendations.(5)
AI Can Predict Cancer Risk of Lung Nodules
Pulmonary nodules present as small spots on the lungs on chest imaging. They have become a much more common finding as CT gained favor over X-rays for chest imaging. Dr. Vachani and colleagues from the Oncology Department at the Perelman School of Medicine, University of Pennsylvania in Philadelphia, evaluated an AI-based computer-aided diagnosis tool developed by Optellum Ltd. in Oxford, England.
The objective is simple: assist clinicians in assessing pulmonary nodules on chest CT scans. While CT scans show many aspects of the nodule, such as size and border characteristics, AI can delve even more profoundly. A total of 300 chest CTs of indeterminate pulmonary nodules were used in the study. Analysis showed that using the AI tool improved the malignancy risk prediction of nodules on chest CT. It also enhanced the readers’ agreement with risk stratification and management recommendations.(5)
AI-powered Solution Supports the Evaluation of COVID-19 Patients
RADLogics is an initiative that develops AI-Powered solutions that support image analysis to improve radiologists’ productivity. AI-driven technology automatically detects and measures abnormalities, enhancing efficiency and expediting care.
The FDA has authorized their applications to assist in detecting and quantifying findings associated with COVID-19. They include case triage and disease extent measurements using imaging findings on CT and X-ray scans. These solutions are being integrated into worklist and PowerScribe reporting workflows via the Nuance AI Marketplace and PowerShare Network, connecting over 7,500 healthcare facilities in the U.S.(6)
AI-powered Solution Supports the Evaluation of COVID-19 Patients
RADLogics is an initiative that develops AI-Powered solutions that support image analysis to improve radiologists’ productivity. AI-driven technology automatically detects and measures abnormalities, enhancing efficiency and expediting care. The FDA has authorized their applications to assist in detecting and quantifying findings associated with COVID-19. They include case triage and disease extent measurements using imaging findings on CT and X-ray scans.
These solutions are being integrated into worklist and PowerScribe reporting workflows via the Nuance AI Marketplace and PowerShare Network, connecting over 7,500 healthcare facilities in the U.S.(6)
PET scans with artificial intelligence, Alzheimer’s disease may be predicted years before actual diagnosis
Radiologist Jae Ho Sohn, MD, of UC-San Francisco, and colleagues created a calculation that analyzes Positron outflow tomography filters of patients whose memory is now deteriorating. Based on this examination, the analysis gives what Sohn considers a “highly exact forecast that can boost the certainty of Alzheimer’s disease determination or rule it out.” The analysts tried the method on two novel datasets after training consisting of 1,921 images. One dataset from the same ADNI database contained 188 pictures unfamiliar to the algorithm. The other dataset was a novel set of looks from 40 UCSF Memory and Maturing Unit patients, all of whom had shown conceivable cognitive disability.
The calculation accurately recognized 92% of patients from the former and 98% from the test latter sets who, in the long run, developed Alzheimer’s disease.(7)
PET scans with artificial intelligence, Alzheimer’s disease may be predicted years before actual diagnosis
Radiologist Jae Ho Sohn, MD, of UC-San Francisco, and colleagues created a calculation that analyzes Positron outflow tomography filters of patients whose memory is now deteriorating. Based on this examination, the analysis gives what Sohn considers a “highly exact forecast that can boost the certainty of Alzheimer’s disease determination or rule it out.”
The analysts tried the method on two novel datasets after training consisting of 1,921 images. One dataset from the same ADNI database contained 188 pictures unfamiliar to the algorithm. The other dataset was a novel set of looks from 40 UCSF Memory and Maturing Unit patients, all of whom had shown conceivable cognitive disability. The calculation accurately recognized 92% of patients from the former and 98% from the test latter sets who, in the long run, developed Alzheimer’s disease.(7)
An AI-Driven "Virtual Biopsy" Approach for Identification of Genetic Alterations in Glioblastomas
A modern ponder from researchers at UCSF Radiology employs insights to foresee the nearness of specific genetic changes in patient’s tumors utilizing non-invasive brain MRI. Using this AI-driven “virtual biopsy” approach, they could precisely distinguish some clinically relevant hereditary modifications counting a few that are beneath examination as potential treatment targets. The work speaks to an imperative step toward a completely-robotized strategy for non-invasive, imaging-based distinguishing proof glioblastomas with IDH (Isocitrate dehydrogenase protein) changes and specific other atomic biomarkers significant for directing treatment and deciding forecast.(8)
An AI-Driven "Virtual Biopsy" Approach for Identification of Genetic Alterations in Glioblastomas
A modern ponder from researchers at UCSF Radiology employs insights to foresee the nearness of specific genetic changes in patient’s tumors utilizing non-invasive brain MRI. Using this AI-driven “virtual biopsy” approach, they could precisely distinguish some clinically relevant hereditary modifications counting a few that are beneath examination as potential treatment targets.
The work speaks to an imperative step toward a completely-robotized strategy for non-invasive, imaging-based distinguishing proof glioblastomas with IDH (Isocitrate dehydrogenase protein) changes and specific other atomic biomarkers significant for directing treatment and deciding forecast.(8)