Effective treatment and management of prostate cancer care requires accurate assessment of a tumor’s aggressiveness. Prostate Serum Antigen (PSA) measurement and patient history can help guide possible therapy but does not sufficiently predict disease progression. Invasive biopsies discomfort the patient, suffer from significant inter-observer variability and possible sampling error, and can lead to side effects. A number of imaging techniques, such as MRI, are non-invasive and display the entire prostate and tumor. Using only a single image or single biochemical test (such as PSA), generates useful information but is not always reliable predictions of disease status. However, combining the evaluation of multiple images and/or results of tests improves the accuracy for assessing prostate cancer.
Current conventional multi-parametric MRI, such as Prostate Imaging Reporting and Data System (PI-RADS) involves qualitative assessment of multiple MRI images and has added valuable. PI-RADS includes structural, diffusion, and dynamic contrast enhancement images require experienced and specially trained radiologists. The quality of MRI interpretations depends on the skill and experience of the reader leading to possible inconsistent and unreliable of the tumor’s aggressiveness. Current artificial intelligence (AI) focuses on extracting and exploiting spatial features in a single modality and does not use information from other modalities.
Recently, pilot studies analyzed spatially-registered multi-parametric MRI by treating each voxel as a multi-component vector, instead of a scalar. These studies showed promise as a way to predict Gleason score, a reliable predictor of tumor aggressiveness and disease progression. Analysis of the spatially registered MP-MRI find the Gleason score, tumor volume, eccentricity, and signal to clutter ratio to help assess the patient condition. The processing reproducibly and consistently combines all MRI modalities and can be used in a variety of clinical settings without significant training, such as in AI. These studies transfer and modify algorithms applied to hyperspectral imagers mounted in airborne platforms to aid remote sensing, environmental studies, and defense.
It is therefore crucial to develop a non-invasive way to quantitatively determine the tumor’s aggressiveness that can accommodate for differing clinical conditions and therapists. Combining multiple non-invasive assessments must effectively, consistently, quantitatively, and efficiently use all information that is available.
This Research Topic examines examines how combining non-invasive multiple images and/or biochemical tests can generate a better predictor of prostate cancer aggressiveness relative using a single modality or test. We invite manuscripts on, but not limited to, the following topics:
- Adding PSA (or other biochemical tests) and/or other imaging and/or multi-parametric MRI to improve prostate tumor evaluation.
- Studies of additional hyperspectral related algorithms such as change detection, anomaly detection applied to spatially registered multi parametric MRI to assess prostate tumors and response to treatment.
- Studies that employ higher number of patient samples to examine measure tumor volume, eccentricity (tumor shape), and signal to clutter ratio applied to spatially-registered MRI hypercubes.
- Studies of spatially registered MP-MRI to detect and assess metastases and extension of disease to nearby seminal lymph nodes.
- Examine the possible advantage of supplementing and complementing artificial intelligence studies by adding extra modalities such as textures to the multispectral hypercube.
- Studies of methods to spatially register the MRI that go beyond rigid transformations.
- Analogous to human color vision, apply red, green, and blue colors to registered MP-MRI in order to display prostate tumors, normal prostate in color and provide a nuanced view of tumors as heterogenous lesions.
- Modifying Magnetic Resonance Spectroscopy (MRS) so that it has higher spatial resolution but lower spectral resolution and apply the supervised target algorithms
- Demonstrating that these techniques can be applied in a variety of clinical settings, such as different magnetic fields and pulse sequences without retraining and clinics.
Please note: manuscripts consisting solely of bioinformatics or computational analysis of public genomic or transcriptomic databases which are not accompanied by validation (independent cohort or biological validation in vitro or in vivo) are out of scope for this section and will not be accepted as part of this Research Topic.
Effective treatment and management of prostate cancer care requires accurate assessment of a tumor’s aggressiveness. Prostate Serum Antigen (PSA) measurement and patient history can help guide possible therapy but does not sufficiently predict disease progression. Invasive biopsies discomfort the patient, suffer from significant inter-observer variability and possible sampling error, and can lead to side effects. A number of imaging techniques, such as MRI, are non-invasive and display the entire prostate and tumor. Using only a single image or single biochemical test (such as PSA), generates useful information but is not always reliable predictions of disease status. However, combining the evaluation of multiple images and/or results of tests improves the accuracy for assessing prostate cancer.
Current conventional multi-parametric MRI, such as Prostate Imaging Reporting and Data System (PI-RADS) involves qualitative assessment of multiple MRI images and has added valuable. PI-RADS includes structural, diffusion, and dynamic contrast enhancement images require experienced and specially trained radiologists. The quality of MRI interpretations depends on the skill and experience of the reader leading to possible inconsistent and unreliable of the tumor’s aggressiveness. Current artificial intelligence (AI) focuses on extracting and exploiting spatial features in a single modality and does not use information from other modalities.
Recently, pilot studies analyzed spatially-registered multi-parametric MRI by treating each voxel as a multi-component vector, instead of a scalar. These studies showed promise as a way to predict Gleason score, a reliable predictor of tumor aggressiveness and disease progression. Analysis of the spatially registered MP-MRI find the Gleason score, tumor volume, eccentricity, and signal to clutter ratio to help assess the patient condition. The processing reproducibly and consistently combines all MRI modalities and can be used in a variety of clinical settings without significant training, such as in AI. These studies transfer and modify algorithms applied to hyperspectral imagers mounted in airborne platforms to aid remote sensing, environmental studies, and defense.
It is therefore crucial to develop a non-invasive way to quantitatively determine the tumor’s aggressiveness that can accommodate for differing clinical conditions and therapists. Combining multiple non-invasive assessments must effectively, consistently, quantitatively, and efficiently use all information that is available.
This Research Topic examines examines how combining non-invasive multiple images and/or biochemical tests can generate a better predictor of prostate cancer aggressiveness relative using a single modality or test. We invite manuscripts on, but not limited to, the following topics:
- Adding PSA (or other biochemical tests) and/or other imaging and/or multi-parametric MRI to improve prostate tumor evaluation.
- Studies of additional hyperspectral related algorithms such as change detection, anomaly detection applied to spatially registered multi parametric MRI to assess prostate tumors and response to treatment.
- Studies that employ higher number of patient samples to examine measure tumor volume, eccentricity (tumor shape), and signal to clutter ratio applied to spatially-registered MRI hypercubes.
- Studies of spatially registered MP-MRI to detect and assess metastases and extension of disease to nearby seminal lymph nodes.
- Examine the possible advantage of supplementing and complementing artificial intelligence studies by adding extra modalities such as textures to the multispectral hypercube.
- Studies of methods to spatially register the MRI that go beyond rigid transformations.
- Analogous to human color vision, apply red, green, and blue colors to registered MP-MRI in order to display prostate tumors, normal prostate in color and provide a nuanced view of tumors as heterogenous lesions.
- Modifying Magnetic Resonance Spectroscopy (MRS) so that it has higher spatial resolution but lower spectral resolution and apply the supervised target algorithms
- Demonstrating that these techniques can be applied in a variety of clinical settings, such as different magnetic fields and pulse sequences without retraining and clinics.
Please note: manuscripts consisting solely of bioinformatics or computational analysis of public genomic or transcriptomic databases which are not accompanied by validation (independent cohort or biological validation in vitro or in vivo) are out of scope for this section and will not be accepted as part of this Research Topic.