The gastroenterology and hepatology CRO is a growing industry with several key trends. For example, the rise of radiomics and AI are two trends that are likely to impact the industry in the coming years. Another trend is a focus on increasing research output. This has resulted in several new K-awards and NIH awards.
The National Institute of Health (NIH) offers several K-awards, which support researchers, postdoctoral fellows, and senior postdoctoral fellows. These awards are designed to prepare candidates for significant grant support. An NIH K-award for gastroenterology and hepatology CRO can be used as an opportunity to pursue independent research. However, it is not intended to imply a commitment to a tenure-track position. For this purpose, the NIH awarding office may terminate the award if it determines that the awardee needs to prepare to assume such a role. Alternatively, a K-awardee may continue with the award for one year after the initial funding period. Research experience is required in a supervised research setting. Knowledge may be obtained in a laboratory within the intramural NIH system or elsewhere. Regardless of the site, the experience must be under the guidance of a faculty mentor. K-award recipients are notified of their grant status at regular intervals. This includes periodic updates on their research grant, employment history, and honors. They may also be contacted for additional information. If a recipient is preparing to accept a position at a tenure-track institution, they should contact the program officer immediately.
An increasing number of studies applying AI to gastroenterology
The increasing use of artificial intelligence (AI) in gastroenterology and hepatology has produced many promising results. AI technologies have advanced radiology image analysis, ultrasound, and computed tomography image analysis. Artificial intelligence has been applied to various applications, including detecting lesions, predicting treatment responses, predicting the likelihood of gene mutations, predicting prognosis, and facilitating treatment. These advancements have enabled precise chemotherapy and improved the diagnosis and prognosis of patients with digestive system malignancies. However, technology has also created some challenges in clinical practice. One major challenge in gastroenterology and hepatology is the need for trained pathologists. Currently, only a limited number of pathologists are available worldwide. This can negatively affect the accuracy of the pathological analysis.
Developing a collaborative approach among doctors and researchers is essential to address the problems. Another challenge is the availability of appropriate databases. Although many studies have demonstrated promising results, it is essential to consider the limitations of retrospective studies. The model may overestimate its accuracy if the training data matches the patient data. Other challenges include the need for high-quality testing and validation datasets. Furthermore, the use of AI has legal and ethical issues. As such, it is crucial to test the model for legal and moral liabilities before implementing it. AI-aided endoscopy is one promising solution to address these challenges. This technology uses the EGD to classify images of the stomach. It also provides a three-dimensional image reconstruction. Several studies have found that this method is more accurate than conventional methods. AI-assisted pathology tools have been applied in detecting steatosis and quantifying lobular inflammation. They can also identify steatosis in hepatitis B virus infection and metabolic-associated fatty liver disease. Although these techniques show impressive results, the research community has a long way to go before approving them. More randomized controlled studies will be necessary. Until then, doctors and researchers should work together to address the challenges of AI. It is essential to be aware of the limitations of AI in gastroenterology and hepatology.
Rise of radiomics
Radiomics is a computational imaging technique that exploits high-throughput extraction of standardized and morphological imaging characteristics from diagnostic imaging images. This technique offers new directions for the diagnosis, staging, and treatment response assessment of liver diseases. The clinical application of radiomics still needs to be more extensive, however. Several studies have demonstrated promising results. However, there is a need for further clinical validation. Also, researchers should standardize and optimize the implementation of the algorithms. In clinical trials, the presence of chemo-resistant HCC is a limitation to the effectiveness of chemotherapy. Early detection of HCC is essential for improved clinical care. If detected early, local tumor therapy may be the preferred option. Multimodal imaging has demonstrated successful prediction of VMI in HCC patients. Specifically, arterial-phase CT imaging has shown satisfactory performance in detecting newly emerging HCCs, with acceptable sensitivity and specificity. An additional advantage of this approach is its ability to see preselected regions of interest in cirrhotic livers.
Radiomics has gained increasing attention as a method of computerized image analysis. This process can be performed at various institutions and is typically characterized by a workflow that includes a series of preprocessing steps, including image coregistration, intensity normalization, and tumor identification and segmentation. Ultimately, the results can be used to develop mathematical models supporting clinical decision-making. Several studies have examined the potential of radiomics in neuro-oncology. Some of these studies have incorporated deep learning. Deep learning is a method for computerized image analysis that can perform predictive tasks. Compared to conventional visual image analysis, this method can classify feature images using texture information from the initial convolution layer. Although radiomics and deep learning techniques have shown promising results in several cancers, some restrictions exist. For instance, radiomics and deep learning can only progress to clinical practice if they are validated and their implementation is standardized. These limitations are expected to be addressed in future developments. Another possible challenge is the need for more sufficient data. Radiomics can be used in various imaging modes, but different acquisition and processing parameters affect the outcome. It can also be challenging to interpret the biological significance of the features derived from the high-level analysis. Thus, further studies must correlate radiomics features with tumor genomics and proteomics.