Recently, artificial intelligence has emerged as an indispensable tool for the reconstruction and analysis of various biomedical signals. Development of powerful machine learning technologies, especially deep learning, has paved the way for the rapid development of biological and biomedical signal processing solutions in various areas such as AI-powered imaging, surface enhanced Raman spectroscopy (SERS), and EEG, etc.
Signal processing is an important step for the meaningful reconstruction and analysis of biosignals which are highly perturbed in nature. Conventional signal processing methods provide a reasonable solution, however, designing a sophisticated method for optimal performance often requires multiple hits and trials, large computation and in depth mathematical analysis for optimal parameter selection. In this regard deep learning allows for a quick and reliable solution by directly mimicking the desired output. With the recent advancements in learning strategies the desired output can be obtained with the help of supervised, self-supervised and unsupervised methods. Nowadays, AI is playing an increasingly significant role in both biological and biomedical sciences, with technologies including bio-sensors, bio/medical imaging, and bioinformatics.
One of the major challenges in bio signal processing is high-dimensional data with fewer numbers of samples, commonly referred to as “curse of dimensionality”. The ability of machine learning methods to deal with ill-posed nature has already attracted the attention of the research community. However, the ever growing data in biomedical science demands novel solutions to improve reproducibility, robustness, and efficiency in biosignal acquisition and processing. Investigating AI-based solutions to handle rich information about biological systems and molecules at various resolutions, all the way from atomic-level to physiological-level is critical, not only for diagnosis, but also for prognosis and treatment.
In this Research Topic we welcome manuscripts on any aspect of biomedical signal processing using reinforcement, supervised, semi-supervised or unsupervised learning approaches. Here, biomedical processing includes all kinds of biomedical signals captured using different modalities such as bioinformatics data including but not limited to:
- genomics or proteomics
- wearable and laboratory-based biosensors data from SERS, EEG,ECG, etc
- medical imaging data acquired using optical microscopy, fluorescence microscopy, electron tomography, nuclear magnetic resonance, single particle cryo-EM, and X-ray crystallography, ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), optical coherence tomography (OCT), etc
It covers reconstruction, pre/post processing and analysis of these signals for improved performance and we welcome submissions related to the following sub-topics:
- Denoising, super-resolution, registration and pre-processing of biological or biomedical images
- Feature extraction from measurement data
- Classification of biosensors and biomedical signals with applications in diagnosis and prognosis
- Segmentation and quantification from images
- Integration of data acquired using different modalities
- Image-omics and radiomics in disease diagnosis and therapy
- Digital pathology through bioinformatics and bioimaging
Topic Editor Shujaat Khan is employed by Siemens Medical Solutions USA, Inc. and holds patents US11145028B2 and US11368349; Topic Editor Imran Naseem is employed by Love for Data, Paskistan. All other Topic Editors declare no competing interests with regards to the Research Topic subject.
Recently, artificial intelligence has emerged as an indispensable tool for the reconstruction and analysis of various biomedical signals. Development of powerful machine learning technologies, especially deep learning, has paved the way for the rapid development of biological and biomedical signal processing solutions in various areas such as AI-powered imaging, surface enhanced Raman spectroscopy (SERS), and EEG, etc.
Signal processing is an important step for the meaningful reconstruction and analysis of biosignals which are highly perturbed in nature. Conventional signal processing methods provide a reasonable solution, however, designing a sophisticated method for optimal performance often requires multiple hits and trials, large computation and in depth mathematical analysis for optimal parameter selection. In this regard deep learning allows for a quick and reliable solution by directly mimicking the desired output. With the recent advancements in learning strategies the desired output can be obtained with the help of supervised, self-supervised and unsupervised methods. Nowadays, AI is playing an increasingly significant role in both biological and biomedical sciences, with technologies including bio-sensors, bio/medical imaging, and bioinformatics.
One of the major challenges in bio signal processing is high-dimensional data with fewer numbers of samples, commonly referred to as “curse of dimensionality”. The ability of machine learning methods to deal with ill-posed nature has already attracted the attention of the research community. However, the ever growing data in biomedical science demands novel solutions to improve reproducibility, robustness, and efficiency in biosignal acquisition and processing. Investigating AI-based solutions to handle rich information about biological systems and molecules at various resolutions, all the way from atomic-level to physiological-level is critical, not only for diagnosis, but also for prognosis and treatment.
In this Research Topic we welcome manuscripts on any aspect of biomedical signal processing using reinforcement, supervised, semi-supervised or unsupervised learning approaches. Here, biomedical processing includes all kinds of biomedical signals captured using different modalities such as bioinformatics data including but not limited to:
- genomics or proteomics
- wearable and laboratory-based biosensors data from SERS, EEG,ECG, etc
- medical imaging data acquired using optical microscopy, fluorescence microscopy, electron tomography, nuclear magnetic resonance, single particle cryo-EM, and X-ray crystallography, ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), optical coherence tomography (OCT), etc
It covers reconstruction, pre/post processing and analysis of these signals for improved performance and we welcome submissions related to the following sub-topics:
- Denoising, super-resolution, registration and pre-processing of biological or biomedical images
- Feature extraction from measurement data
- Classification of biosensors and biomedical signals with applications in diagnosis and prognosis
- Segmentation and quantification from images
- Integration of data acquired using different modalities
- Image-omics and radiomics in disease diagnosis and therapy
- Digital pathology through bioinformatics and bioimaging
Topic Editor Shujaat Khan is employed by Siemens Medical Solutions USA, Inc. and holds patents US11145028B2 and US11368349; Topic Editor Imran Naseem is employed by Love for Data, Paskistan. All other Topic Editors declare no competing interests with regards to the Research Topic subject.
